The SHWoodwind F.O. Guide To Octave Key Problems
Saxophone octave key mechanisms
are responsible for more playing problems than any other mechanical
aspect of the instrument - and while they can be among the most
difficult for a player to diagnose, they can also be among the easiest
By far the biggest problem is that so few players
really understand how octave key mechanisms works - and so this
article aims to explain exactly what they do, how they do it and
how they can fail.
The modern octave key mech isn't really all that complicated
mechanically - but the older a horn is, the more complex the mechanism
is likely to be...to the point where some look almost surreal in
their design. These
can be a real challenge, even for a repairer, but once you understand
the operating principles behind the octave key mechanism it should
be possible to mentally reduce its design and functionality down
to something more comprehensible.
This guide is by no means complete. It's taken over four years
just to get to this point, and even though it runs to over 75 pages
of text, photos and diagrams - and for the time being tends to focus
mainly on modern mechs as found on altos, tenors and soprano saxes
- it still doesn't cover every problem you might encounter. That
in itself ought to give you some idea of the complexity of the subject
- but even incomplete it stands as the most comprehensive article
on the subject published to date.
As and when time permits I intend to add to it and widen its scope
(see Version History), as well as cover many of the more common
vintage mechs, but I felt it would be of more use publishing it
now rather than waiting until it was finished...if such a point
even exists. If nothing else, reading this article should furnish
you with the ability to diagnose most octave key mech problems.
And even if you still can't figure out what's wrong, you might just
be able to work out what isn't wrong...and that's got to
be better than staring at your horn and wondering why it won't play.
This first part is devoted to describing the octave key mech,
its various parts and showing how and why it works - and if you're
going to tinker with the mech you'd, frankly, be bonkers not to
at least skim through it beforehand.
The second part - troubleshooting
the octave mech - takes you through diagnosing and (hopefully)
correcting many of the common problems you might encounter. If you're
familiar with the mechanics of the octave key mech or just want
to find out why you can't hit certain notes, you might want to skip
Part One: Understanding The Octave Key Mech
Before we get cracking on sorting out your octave key woes, we need
to sort out a bit of terminology, and take a close look at how an
octave key mech works.
This shot shows a typical modern octave key mech - as fitted to
a TJ Signature Custom
RAW XS alto.
In its most basic form the mech consists of three major parts: the
touchpiece or thumb key, which is the bit you press to make the
mech work; the body octave key, which opens for the notes D to G;
and the crook (or neck) key, which opens for the notes A and above.
The very first saxes had little more than this - in fact they had
completely separate octave keys, so when you wanted to move from
octave G to A, you had to release the first octave key and switch
to the second. This was a royal pain in the...thumb - and it wasn't
long before 'automatic' octave mechs were developed, that did the
switching for you.
These early automatic mechs were often pretty crude, fussy and
overcomplicated - and usually relied on levers that slid over yet
more levers to provide the automation. All very steampunk.
And then, in around 1930, Henri Selmer came up with the swivel arm
octave mech - and it's this design (and slight variations thereof)
that still dominates today. The most important difference is that
the swivel mech is (more or less) powered by leverage - whereas
earlier mechs often relied on springs within the mech to do the
I've colour-coded the mech to make it easier to see which bit is
which - but the main feature is the swivel arm (coloured silver).
It pivots on a pin (the swivel pin) fixed to the body key (blue)
and connects to the surrounding keywork by means of a pair of rounded
pins that fit into corresponding sockets on the keys.
When you press the octave key down (red), power is transmitted to
the mech via the connector (light red) at the bottom of the swivel
- and depending on what note you're going for, either the body key
will open or, via the (green) pin or lifter/shift lever key, the
crook key will (yellow).
The beauty of this mech is that it's elegant and simple, reasonably
robust and usually quite reliable. That said, it has its problems
- most of which are typically due to wear and tear or damage. It
also requires a degree of precision in order to work properly, which
is why many poorly-made swivel mechs (as found on nearly all Ultra-Cheap
horns) can be very problematic. In the worst cases I've seen, almost
all of the power transmitted by the octave key is lost in taking
up excess free play in the swivel arm (and the connecting pins)
- to the point where the body key barely opens, and the crook key
not at all (we'll discuss fixes for this later).
Not every manufacturer adopted the Selmer design, and even through
to the 1960s there were still a few 'freaky' designs to be seen.
Even today it's possible to find horns that don't use a swivel bar
mech (Yamaha student horns in particular), though it should be noted
that these modern non-swivelling mechs are usually just as efficient
(perhaps even more so in some instances).
have a look at the mech in action.
In this sequence you can see the rise and fall of the crook/neck
key, as would take place were you to play any note above octave
G. Because the G key foot is down over the body octave key cup,
the power transmitted from the depression of the octave key touchpiece
is unable to lower the swivel pin - which in turn would normally
raise the body octave key.
This causes the power to be sent along the swivel arm which, as
its name suggests, swivels on the pin. The touchpiece end goes down
and the crook key end goes up - which raises the crook key. When
the touchpiece is released it rises under the power of its own spring
and in so doing it raises the touchpiece end of the swivel arm which
in turn lowers the crook key end. This allows the crook key to close,
helped by the closing force of its own spring.
You can see that there's actually not a lot of movement
going on. This represents an ideal setup; there's very little free
play (from loose/worn or badly regulated action) and almost all
of the motion from the touchpiece is transmitted to the crook key.
I say almost because there will always be a slight amount of free
play in the system otherwise it might perform sluggishly - and there
should always be a slight gap between the pin and the ring at the
lower end of the crook key when the mechanism is at rest (this is
to allow for variable positioning of the crook, as we'll see later).
The total amount of motion available to the mech is entirely dependent
on how far the thumb key is able to move. If you can only press
it down a couple of millimetres, then that's pretty much all the
movement you'll get at either the crook key pad or the body key
pad. If you increase the amount the thumb key can move, you increase
the amount the octave key pads can open...assuming the mech is in
good order. The distance the thumb key moves is called the 'throw'
of the mech.
this next sequence you can see the body octave key cup in action.
This only comes into play between the middle D and octave G (fake
fingerings aside) - because it's only on these notes that the G
key is pressed down, which in turn raises the key's foot over the
body octave key cup.
Without the downward force from the G key's foot, the power from
depressing the octave key touchpiece is now able to bring the swivel
pin down which in turn raises the body octave key. Note the action
of the swivel pin compared to the first sequence - previously it
was static, with just the swivel arm moving; this time the pin rises
Why this happens is down to a simple balance of power.
When a downward force is applied to the touchpiece end of the swivel
arm, the force seeks out the path of least resistance. While the
G key foot is down it's applying a downward force to the body octave
key cup - and as the spring that powers the G key is usually quite
hefty it's more force than the octave mech can overcome...so the
power from the touchpiece takes the easier route, which is via the
swivel arm to the crook key. But remove that downward force on the
body octave key and the power will flow back from the crook key
as it closes and the body key will rise with ease.
If you remove all external forces to the octave mech (take the crook
off and press the G key down) and then press the octave key down,
you may see the body key rise. If it does so it'll be down to friction
within the mech or drag from the lubricants on the pivots - but
it means that the only key that's free to move in these circumstances
is the body key. This, then, is the mech's 'default' action.
final sequence shows the mech moving between octave A and G.
Because the touchpiece is held down all the movement takes place
from the body key upwards. As the G key foot falls it brings down
the body key, which in turn raises the swivel pin. The force cannot
travel back to raise the touchpiece because it's held down firmly
by your thumb - so it has no choice but to travel upwards to the
pin...which now rises, and so opens the crook key.
When the G key is pressed down, its foot rises and
offers a path of less resistance to the energy now being provided
by the downward force of the crook key's spring. The crook key closes,
and as it does so brings the swivel arm down, which in turn lowers
the swivel pin and raises the body key.
This kind of octave mech is pretty tolerant and will
keep on working even when quite worn and out of adjustment - and
for the most part all you really need to worry about is that the
G key spring is strong enough to do its job - but it still has some
Its functionality relies on the transmission of relatively light
forces, so anything that impedes the flow of energy up and down
the mech will lead to problems. This is why a little wear and tear
isn't usually an issue - when things are a little worn they tend
to move quite freely. Any hint of friction, though, and it's a very
different matter. A slight knock to the mech can really throw things
out, too - as you can see in the three sequences, there's really
not a great deal of movement going on and if one or more of the
component keys gets bent out of alignment it may well mean that
the mech won't be able to move enough to raise one or other of the
By the same token, there's a need for such mechs to be reasonably
accurate. Granted, they needn't be all that precise in engineering
terms, but a poorly-built mech can throw up all manner of unexpected
problems - and while it's the balance of leverages that makes this
type of mech work, in a poor-quality mech those leverages can be
precisely what stops it from working.
You might well find that your octave mech looks nothing like the
one above, but the mechanical principle will be the same. It's essentially
a balanced mechanism with three main areas of force or power that
have to be taken into account. The first is the force provided by
the closing and releasing of the octave key touchpiece. As it's
your thumb that presses this key down this is the strongest force
applied to the mech. Upon releasing the key, however, the force
will only be as strong as the spring which powers it.
The second is the G key foot. When it's up (when you're playing
octave G to middle D) it applies no force to the mech - but when
it's down it will apply as much force as the spring which holds
the G key open is able to provide.
Lastly there's the crook key spring, and although this is a very
lightly spring key it will still apply some closing force as well
as some resistance to opening.
When troubleshooting an octave key mech you can pretty much disregard
the force applied by the touchpiece. Granted, the key must fall
and rise with ease otherwise it will cause problems, but in simple
terms you can think of this key as being a switch. Press it down
and it switches the octave mech on, release it and it switches it
It's seldom the root of problems in a modern mech, though in older
designs the return force of the touchpiece's spring may have to
be set so as to overcome resistance from other springs in the mech.
mentioned earlier that not every manufacturer adopted the Selmer
mech design, and here's an excellent example. It's the mech as fitted
to the ubiquitous Yamaha 23 tenor. Yamaha clearly have no problem
making use of the Selmer design - they use it on almost all of their
horns - so perhaps the 23 was fitted with a different mech because
it was either cheaper to make or more robust.
I rather like this mech because it's simple, reliable, sturdy and
really rather slick in use. It's also quite tolerant of wear or
imbalance in the mechanism - and about the only thing that will
cause it to fail is because it's had a knock...and a hefty one at
As per the Selmer mech there are three sources of power going into
the mech; the touchpiece spring, the G key and the crook key. There's
no swivel, just a rather clever extended pin key that pivots on
the touchpiece arm.
In terms of design it's quote close to a vintage mech, given that
most of them had the body key on a separate pivot. The big difference
here is that some thought has gone into the placement of this key,
and rather than being shoved off to one side of the mech, the pivot
for it is tucked underneath the touchpiece barrel.
In terms of operation, the stub that extends upwards out of the
body key (just back from the key cup) is perhaps the most important
part. The pin key has two vaguely triangular plates fixed to each
end - of which the lower is inverted and slotted to accommodate
the body key pin. It's this slot that converts the back-and-forth
action of the pin key when the G key is open into an up-and-down
motion when the G key is closed.
What I particularly like about this mech is that it shows that
there's more than one way 'to skin a cat' - as they say. Had the
Selmer mech never been invented, it's quite possible that a mech
like this would have been the de facto standard.
So far, so good - but up until now we've had it easy.
The Selmer mech might look complicated, but once you understand
its operating principles, it's not so hard to get to grips with
- and modern versions of vintage-style mechs are, arguably, even
But what about the 'real McCoy' - the proper vintage mech...?
Well, some of those older mechs can be extremely complex.
Aside from the three sources of power (the touchpiece, the G key
and the crook key), the modern mechs above have no internal springs
- it's all done with balanced levers - but before the invention
of the swivelling octave mech it was common to have two or three
additional springs within the mech to power the keys. It may help
to think in terms of the modern mech as being 'passive' and the
older mechs as being 'active' - the former relying on power that's
supplied from outside the main part of the mech and the latter relying
on power supplied by additional springs within the main part of
the mech. If you're at all uncertain as to whether the mech on your
horn is active or passive, a very simple test is to finger an octave
G and then try to move the pin key up and down. If it moves freely
with no resistance and only returns to its starting position under
its own weight, it's a passive mech. If feels slightly resistant
to movement and snaps back smartly into position when you release
it, it's an active mech (assuming all the springs are working).
Active mechs have to be very carefully balanced, with each of the
three external sources of power being balanced against each other
as well as being balanced against any internal springs. These mechs
very often depend on a very specific 'off' force from the touchpiece
a vintage octave key mech as fitted to a Buescher
curved soprano sax.
The principle of operation is exactly the same as with the modern
mech, but as you can see it's a far more complex arrangement. As
with the modern mech there are there three main sources of power
- the touchpiece, the G key and the crook key. However, there are
also two additional sources of power from the body octave key (which
is sprung to open) and the pin key (which is sprung to rise).
If you were to play an octave A, the following would happen.
As you press the touchpiece it would raise the two arms that straddle
the mech. Because the G key is open, its foot still rests on the
body key's arm...so the body key cannot open. At the moment the
pin key's lower arm is suspended above the G key's foot, and when
the touchpiece arm rises, the pin key's will be free to rise. In
order to do so, though, its spring must be strong enough to overcome
the closing force of the crook key.
When the touchpiece is released (so as to play a low A), the arms
fall under the power of the key's spring - and the upper arm will
overcome the opening force of the pin key and bring it down.
If you now play an octave G, this is what happens.
With the G key down the foot will be raised. The first thing that
happens is that the stub beneath the pin key arm rises to meet that
arm - and this prevents the pin from rising. This 'locks out' the
crook key. In so doing, the stub that rests on the body key's arm
will rise, thus allowing the body key to open. But it can't yet,
because the lower arm of the touchpiece is still holding it down
(you can just about see the end of the body key's stub beneath the
When the touchpiece is pressed down, the arms rise and allow the
body key to open.
When you release the touchpiece to play a low G, the long arm of
the touchpiece will fall under the power of the spring - and will
bring the body key arm down with it. In order to so the touchpiece's
spring must be strong enough to overcome the opening force of the
body key's spring.
When moving between octave A and G the touchpiece key is held open
and does nothing - and it's the G key foot that does all the work,
with that 'offset T' shaped stub allowing the body and crook keys
to work in turn or similarly forcing them to close. It's worth keeping
those two terms (allowing and forcing) in mind. When diagnosing
octave key problems it's extremely helpful to know what's 'allowed'
to work at any particular moment, and what's not.
another vintage octave mech, this time on a single-piece straight
It contains all the components of the Buescher mech above, but the
layout is a little different - most notably the crook key is now
housed on the main body tube (as it will be on any single-piece
straight soprano, which won't have a separate crook).
Another difference is that there's no distinct pin key - and rather
than the crook key being sprung closed, it's now sprung open. This
means there's no need for a pin to open the crook key...but there's
still a need for a pin to close it, and this function is taken up
by the G key foot, and shared with the upper arm of the touchpiece
As before, when playing an octave A the touchpiece is pressed down
and the arms are raised. The G key foot over the body key prevents
it from rising but the crook key is now free to open under the power
of its own spring.
Releasing the touchpiece brings the arms down and the upper arm
overcomes the crook key's spring and closes it.
When playing an octave G, the G key foot rises and allows the body
key to rise under the power of its own spring - but the top part
of the foot also rises underneath the crook key and prevents it
This is a relatively simple mech, and because the various key arms
act in a more-or-less up and down fashion, there's less friction
involved. However, there's still a need for the springs of the crook
and body keys to be balanced against the touchpiece and G key springs.
these vintage mechs look a bit confusing, how about one that works
The shot on the left shows the octave mech as fitted to the Conn
6M alto - and the first thing that's most notable about this
mech is that the crook octave key pip is fitted beneath the tube.
It's this feature that gives the horn its nickname - the 'Underslung'.
If you follow the crook key down from the pad you'll also see that
the pin key sits on top of a plate attached to the key arm (this
is still technically the ring). In order to open the crook key,
the pin must press down onto the plate rather than rise up, as it
does on a standard mech. Following the mech down still further and
you can see it's a virtual nightmare of tubes and arms - what's
known in the trade as 'a whole bunch of gubbins'.
If you have one of these horns and you're having problem with the
octave key mech, you'd be well advised to have it looked at by a
repairer. Fortunately it's a pretty reliable mech - but when it
goes wrong it can be a proper pain to sort it out. That said, the
troubleshooting chart (below) will still help you diagnose most
of the common problems.
you'll often see the mech on the right (as fitted to a Yanagisawa
992 alto) referred to as an underslung mech - and while some
of the keywork is situated beneath the crook tube, the octave key
pip is still fitted to the top...so it's not a true underslung.
This mech works in exactly the same way as a standard Selmer-type
mech - which is worth remembering if you ever have to bend the octave
key. It's also worth pointing out that these type of octave keys
can be rather tough, and very difficult to bend.
Regulating or balancing a vintage octave mech can be a nightmare,
even for a professional repairer. I've been at this job for decades
and I still come across mech designs that I've never encountered
before - and, frankly, some of them border on the bizarre.
Compared to the modern mech, the vintage one throws a couple more
springs into the equation - but it also raises the spectre of friction.
All those arms and stubs, sliding against each other - if they're
unable to move smoothly they'll bring the whole mech to a grinding
halt, and if they're not buffered in some fashion you'll end up
with an incredibly noisy mech. Back in the day it was common to
use very thin cork to buffer the arms - which was often greased
to reduce friction. Felt was often used too - but these days a repairer
is more likely to make use of Teflon sheet, which provides resilient,
low friction buffering.
There's also a requirement for the various arms and levers to be
set at a certain height. For example, if the upper of a pair of
arms on a touchpiece key was set too high, it would never be able
to bring the pin key down...because its travel would be limited
by that of the lower arm. Likewise, if the pin key cops a whack
(as they often do) it may become so distorted as to prevent the
touchpiece key's lower arm from doing its job properly.
Thus the ultimate nightmare for a repairer is an unfamiliar vintage
octave mech that's badly worn and that's taken a bash...and is fitted
to a baritone sax.
Speaking of which, here's a baritone octave mech - as fitted to
an East German built horn. When the touchpiece is pressed the default
action is to open the crook key (though this might not actually
be mounted on the crook...but it'll be pretty close, typically just
below the crook socket).
The critical point on this type of mech is where the lever that
comes off the crook key contacts the foot of the G key. As the crook
key opens, its lever arm comes down over the G key foot. How far
down it goes will depend on how far you're able to push the touchpiece
down - and just as with the pin key and the ring on a standard mech,
there should be a small gap between the G key foot and the crook
key lever arm when the mech is in this state. When the G key is
operated, its foot rises and lifts the crook key lever arm - which
closes the crook key pad and forces the body key open.
A really good way to see what can go wrong with this kind of mech
is to poke an obstruction (the back of an old reed etc.) between
the G key foot and the crook key lever - and the press the octave
key. Because the crook key lever isn't able to drop all the way
down, the mech fails - and both octave key pads will open. The same
will happen if you push the reed underneath the G key foot to raise
And because there's a relationship between the distance the touchpiece
goes down and the resting point of the crook key lever, a very common
fault is that insufficient buffering on the touchpiece allows the
body key to open very slightly when the crook key is fully open.
This is very easy to spot - simply press the octave key down (and
no other keys) and watch for movement at the body key cup. You should
see none at all. Once the octave key is fully pressed down, try
pressing it down just a little harder. If there's even a hint of
movement on the body key you'll need to increase the buffering beneath
is a modern bari mech, as fitted to a Chinese copy of a Yamaha baritone.
As you can see, it's a swivel mech - similar to the one fitted to
the horn in the first image in this article. However, there are
a couple notable differences.
Because of the placement of the octave key pips on a modern baritone,
the crook key pip sits more or less dead centre of the mech - while
the body key is perched up at the top end. It's very similar to
the modern sopranino mech below except that the octave pips are
reversed - the body pip(s) sit at the top of the mech and the crook
pip sits in the middle.
It's an 'active' mech - aside from the spring on the thumb key
there are two springs powering the swivel. The lower spring forces
the crook key closed and the upper spring forces the body key open.
This mech features twin body keys - a recent innovation that's supposed
to improve the tone and tuning between octave D and G - and there's
and additional (flat) spring fitted between the two key cups. This
is a balancing mechanism. It's quite fiddly to seat a connected
pair of pads without some means of adjusting the relationship between
the two key cups, and with the spring mounted on the lower key and
its tip resting on a plate connected to the upper key it provides
a crude but effective self-regulating mechanism.
There's often a cork fitted above the lower key cup to prevent any
bounce in the mechanism causing a clatter. If you have to replace
this cork, ensure it's not too thick or it will hold the top key
There's also no true pin key. Given that the rise and fall of the
pin key determines where the power goes within the mech, it's fair
to say that the link between the G key foot and the body key fits
the bill - but it's also true to say that this link is merely an
extension of the G key foot. Either way, the critical point is where
the pin key meets the body key arm.
As to how the mech works - with the G key open (and thus its foot
down), pressing the octave key will transmit power along the mech.
The pin key is up, thus preventing the top half (blue) of the mech
from moving, so the lower end of the swivel is pulled down, and
thus pulls down the crook key (yellow). When the G key is closed,
its foot rises and allows the pin key to drop. The opening force
of the body key spring now comes into play and the key opens. This
raises the top end of the swivel which forces the body key to close.
Setting up (or 'balancing') this type of mech can be very tricky.
For a start, the sheer size of the mech means that the slightest
amount of unwanted free play in the mech will be magnified - so
what seems like a little bit of play in, say, the swivel pin will
result in perhaps one or two millimetres less movement at the body/crook
key pads. Throw in a bit more free play at, say, the swivel arm
ends, and there's half a chance that the crook key pad won't open
at all when it's meant to.
There are two top tech tips that will help whenever you're fiddling
about with a bari mech. The first is to take photos of the spring
orientation. The springs are often very lightly tensioned, and if
they're dislodged from their cradles it might not always be obvious
whether they were set to open or close a key. The second is to check
that the twin body key pads are closing properly. Get this right
and you can subsequently treat them as a single key.
mech differs in that it has no pin key. It's on a sopranino - but
the same type of mech can be found on modern single-piece sopranos
(those that don't have detachable crooks). The job of raising the
crook key is carried out by the top end of the swivel bar, and as
per a standard mech as found on an alto or tenor the crook key is
There's also an extra link from the touchpiece to the upper key
stack - note that arm (in red) extending out to the right. This
connects to the auxiliary B key and comes into use when playing
an octave C#. This note has a tendency to be a bit wild when it
comes to tuning, and bringing the Aux.B pad down helps to flatten
off the sharpness. You'll sometimes find this mechanism fitted to
altos and tenors.
As far as its impact on the octave mech goes, it limits the throw
of the touchpiece. Once the Aux.B key pad is down, the touchpiece
can't go any further - and if the mech is badly worn it might mean
that there's not enough throw to make it work.
Fortunately there's nearly always some means of adjusting the link
- in this the stub that sits on the touchpiece arm can be moved
up and down its own arm, thus allowing you to change the amount
of throw in the octave mech. Ideally the two should be balanced
so that there's no gap between the touchpiece arm and the Aux.B
arm when at rest - but if there's a problem with the mech or it's
just not that well made, you might have to settle for a bit of double-action
(a gap between the connection of two keys, resulting in a slight
clunk when the mechanism is operated).
Note that some mechs on single-piece sopranos/ninos often have
no easy means of internal regulation (corks, felts etc.) and are
reliant on the thickness of the octave pads for such regulation.
In general terms this means both pads should be of identical thickness.
If one is thinner than the other, it may lead to unexpected opening
of one or other of the pads.
It would be impractical to show all the octave key mechs that have
been fitted to horns down the years, which is why it's important
to understand the mechanical principles of the mechanism. Once you
know what's supposed to happen you can examine an unfamiliar mech
and work out what's supposed to do what, and when. At the very least
it'll help you to understand why a particular mech isn't working
properly, even if it doesn't mean you'll be able to fix it.
Part Two: Troubleshooting The
Octave Key Mech
Below you'll find a simple flowchart that will help you to diagnose
some of the more common octave key mech problems. It's by no means
an exhaustive checklist, and is geared more towards modern alto/tenor
mechs rather than vintage/bari ones, but it should cover most eventualities
- and with an understanding of how octave mechs work you should
be able to apply it to non-standard designs.
the directions to the appropriate list of possible causes. Clicking
the link on each cause will take you to a more detailed description
and a possible solution - and a corresponding link at the end of
each solution will bring you back to the chart. Some solutions may
require you to skip from one to another, and you'll find related
links at the end of each section that will take you back to your
starting point. Because it's very easy to get lost I recommend
you right-click on related links within a section and open them
in a separate browser window.
To make it even more confusing it's sometimes the case that there
are several things wrong with the mech and that you'll have to fix
one problem before you can tackle another - and if you don't get
the sequence right you can end up going round in circles. It may
also be the case that a problem can have different symptoms depending
on how severe it is. If you can find a 'one shot' solution here
that solves your problem then that's great. If it takes a couple
of techniques, that's not so bad - but if it starts to look like
you might need to do three or more things to your mech to get it
working, it's probably time to take it to a repairer.
If you feel you'd like to try carrying out your own repairs (or
you have a very pressing need to) please bear in mind that many
of the techniques described require a degree of skill and experience
- and sometimes specialist tooling. I've tried to keep things as
simple as possible - and where possible have favoured the safest
method over perhaps the most efficient...but there will still be
an element of risk involved.
Finally, it's worth bearing in mind that the three most common problems
are sticking pads, bent or binding mechs and bent crook keys.
Key Mech Troubleshooting Chart
The most common octave key problem is that of the crook/neck
key not closing properly because either the crook key is
bent or the crook itself is. When playing the horn from
top C downwards, the notes will start to break up around
G - and it'll be difficult or impossible to play any notes
in the lower octave.
If this describes your problem you can go straight to the
Crook key bent
up or 'pulldown' solution.
Otherwise, follow the chart.
Fit the crook/neck to the instrument.
Does the crook key pad appear to be sitting on the crook octave
Finger an octave G (place your fingers in
the playing position, without playing) - does the body octave
key rise freely?
Worn/badly made mech.
Weak, dislodged or
broken body key spring (vintage/bari)
Weak, dislodged or
broken crook key spring
Release the octave key - does the body key
fall freely and completely
Worn or squashed
key spring (vintage/bari)
beneath pin key
Pin key bent down
When you play notes between middle D and
octave G, are they clear and easy to produce?
key not rising enough (hissing).
Weak, dislodged or
broken crook key spring
Pad skin sticking
to pip when key rises
Octave key pip blocked
Octave key pip
unsoldered or leaking
Crook key opening
Other leaks on
Finger an octave A - does the crook key
Does the pin key rise and fall when you
switch between low and octave A?
When it rises, does the pin make contact
with the crook key?
key bent down
Worn or badly made
on pin key
Pin key bent down
Release the octave key - does the crook
key fall freely and completely?
Sticking octave key pads are a major cause of problems with the
octave mech, but it's one of the easiest to diagnose and cure.
You'll have noted that one or other of the octave key pads isn't
rising when it ought to, and this is because it's glued to the octave
key pip by a mixture of dried saliva and whatever residues were
left on the pad after manufacture. Because the pads are so small,
and the surface of the pip comparatively large, even a small amount
of gunge on the pads can lead to a slight hesitation in the key
cups rising - often accompanied by a faint ticking sound as the
'glue' gives way.
You can confirm the problem by carefully lifting the pad with a
finger. This breaks the 'seal', after which the mech should work
However, you will need to address the problem - and the easiest
way to do this is to clean the pad. See the Handy Hints article
on dealing with sticky pads
for more details.
some rare cases the octave key cups will rise, but the pad skin
remains stuck to the pip. From a mechanical perspective the mech
appears to be working, but because the pad skin is blocking the
hole the octaves won't come cleanly or even at all.
It's usually a sign of a cheap pad - but even good pads, if prone
to excessive stickiness, can go 'baggy'. Cleaning the pad will solve
the problem in the short term, but the proper fix is to have the
horns are prone to sticking octave key pads, and in such cases it
would be wise to consider replacing the standard leather ones with
pads made from other materials. Cork is a popular option, but synthetic
pads are even better. You can even make them yourself; see the Handy
Hints article on making pads
out of Sugru.
Watch out, too, for burrs on the octave key pips. Any sharp edges
on the pips will 'grab' at the pad leather, causing it to stick.
It will also wear the leather away surprisingly quickly. You can
treat the pad for stickiness all you like, but the problem will
never go away until the surface of the pip has been smoothed off.
You're more likely to find such burrs on new, cheap horns - but
this example was taken from a horn that dates back to the 1950s...which
means that owner was probably plagued with a sticking pad, and perhaps
wondered why the horn needed a new body octave pad every year or
If you feel comfortable about having a go at fixing it yourself,
you just need to place a strip of fine-grade carborundum paper (minimum
800 grit) between the pad and the pip...grit side against the pip,
obviously. Press down gently on the pad as you slowly withdraw the
abrasive paper. You'll probably need to repeat the operation a number
of times - but if you're not seeing any improvement after half a
dozen goes, you'll be better off having a repairer look at it. Don't
be tempted to keep on sanding it, or switch to a coarser grade of
carborundum paper - you may end up making the top of the pip uneven.
There are many reasons why an octave mech might bind or be stiff
in action, but the most common is that it's had a bit of a knock.
It's quite a vulnerable mechanism, and even careless handling of
the horn (typically when fitting or removing the crook) can throw
it out of line.
The first thing you need to do is test it.
Without the crook fitted, finger an octave G. Release and press
the octave key a few times and note the movement of the body octave
key (as per the second animated sequence at the start of this article).
It should be smooth and swift, and there should not be any movement
of the pin key (you might see a little, but just a little).
Now keep the octave key down (still fingering an octave G) and wiggle
the pin key up and down. Again, it should move swiftly and smoothly.
Release the octave key and wiggle the pin again - the touchpiece
should rise and fall as you do so. Release the G key and wiggle
the pin. Finally, with the G key released press and hold the octave
key down and press the pin (which has now risen) down - it should
open the body key and close the G key.
If there's any hint of binding or stiffness, you should notice it
- and if it's severe enough you'll notice that some parts of the
mech will move when they're not supposed to. While you're carrying
out these checks keep an eye on the key barrels. Do they seem to
rise and fall in a wavelike motion? If so then probably means the
mech has had a knock and the rod/hinge screw is bent.
can test for bent rod screws in another way. Take a suitable screwdriver,
place it in the slot of the mech's main pivot screw and give it
a quarter turn anti-clockwise (loosening the screw), followed by
the same clockwise (tightening). Keep an eye on the keys mounted
on the rod - do they rise and fall or move from side-to-side? If
they do then that's a dead give-away that the mech has been damaged.
You might notice a little movement from one or other of the keys,
but this should be slight - and it usually means that the mech has
been lubricated with grease. If you spot significant movement of
one or more keys then this may indicate a bent key (so that its
barrel is gripping the rod) or a spot of rust. If the rod screw
fails to move at all then it's probably rusted into the thread end
- and that's not good news.
I don't generally recommend grease as a lube for octave mechs,
but some manufacturers use it. With the right kind of grease there
shouldn't be any problems - though you might notice that the mech
moves a little sluggishly. It should still move smoothly though.
Many cheap horns come with greased octave mechs, and the grease
used is bloody awful - it dries out to a horrible, sticky gunge.
The fix for this is to dismantle the mech, degrease it, lube it
properly and refit it - but there's a nifty get-you-out-of-trouble
fix, which is to pop a drop of cigarette lighter fluid at the barrel
ends of the individual keys that make up the mech. This will 'rejuvenate'
the grease and should restore functionality...for a while, at least.
It will also help if you then oil
a knock to the mech won't bend any keys, but it might throw the
mech out of alignment - and this often leads to binding. Vintage
Selmer mechs, with their C-shaped touchpiece forks, will bind if
either the touchpiece key's upper pillar or the main mech's lower
pillar are shifted closer together. As the touchpiece is pressed
down, bringing the lower pin key down with it - the pin moves further
into the fork. On a more modern mech this isn't a problem, as the
fork usually has room to spare - but this C-shaped design has very
little room for manoeuvre and the pin will jam up against the back
of the fork, usually when the touchpiece is about halfway down.
In the shot on the right the mech is at rest, and you can see the
clearance between the C fork and the pin - it's about a millimetre.
When the octave key is depressed this clearance should either remain
the same or diminish slightly. If it disappears completely and you
feel a corresponding sponginess in the octave key, it's a fair bet
that there's an issue with the alignment of the mech.
If you're lucky you might be able to spot a dent or a small crease
around the base of one of the pillars - and obviously this will
need to be removed - but there's often no visible sign of damage.
The fix for this problem is to lightly tap one of the pillars back
a tad, to create a bit of clearance between the fork and the pin.
It's not a fix I'd recommend to try yourself but if you're on a
gig and the set is about to start, you probably don't have a lot
left to lose.
a pencil, break it in half and throw the pointy bit away. Place
the flat end up against the top of one of the pillars (the main
mech's lower pillar is the best bet) and lightly - and I mean lightly
- tap the other end of the pencil with something suitably solid.
Friction is another common cause of binding, and this can be down
to poor choice of buffering material, corks or felts dropping off,
misalignment of keys or manufacturing defects (some of which may
be more serious than others). One of the less serious defects is
roughness on the interconnections. Here you can see a touchpiece
key - specifically the fork at the far end of the key, in which
the lower pin key sits.
Note those small machining ridges on the upper arm of the fork.
Although these aren't ideal they've at least been buffed and lacquered
- so they're not so rough that they'll cause the mech to bind. In
fact about the only effect they'll have is to make the mech feel
a little 'scratchy' - but if they were much worse than this they
might lead to binding.
The fix is fairly obvious - remove the ridges with a fine file,
then either burnish the inner surface of the fork or smooth it off
with fine-grade emery paper.
A short-term fix for friction problems is to pop a drop of oil
on the offending part. This won't be a problem on an interconnect
like this fork, which typically sits on a pin that's buffered with
a plastic tube - but where cork or felt has been used to buffer
the connection, the oil may soon dissolve the glue holding it on.
One 'gotcha' to watch out for is a mech that's binding because
the pivot screws are too tight. The keywork itself might be in good
order, but if a rod/hinge screw has a drunken thread (one that hasn't
been cut square on the shaft) it may well distort when the screw
is driven home. Point or pivot screws can also be the source of
binding - if they're driven too far into the pillars they can, effectively,
clamp the key in place. A poorly made rod screw really needs sorting
out - typically a new screw will be required - because although
it's possible to remedy the situation by backing the screw out a
little, there's a good chance that it will work its way back in
at a later date, and the mech will bind again. Over-tight point
screws will generally only need backing out slightly, but this may
make them loose in their pillars - in which case a drop of suitable
threadlock is the answer. Loctite 222 or 243 will do the job, with
the latter being slightly stronger and more widely available. Don't
use anything stronger.
If you think binding screws might be the problem, check out the
article on testing the action
for more details about rod and point screws - and read through the
latter part of the section below on worn
mechs, which gives some details about adjusting point screws.
The bottom line on a binding mech is that it usually requires some
dismantling, it often requires some specialist tools and techniques
and it nearly always means the mech has to be re-balanced once any
specific issues have been fixed. In other words, if your mech is
binding, let a repairer sort it out.
Bent pin key
Worn or badly made mech
the moving parts on a horn are prone to wear and tear, but the action
on a horn will often keep working way past the time it should have
The octave mech, however, tends to degrade rather more rapidly and
can sometimes reach a point where the combined wear in the component
parts lead to a complete failure.
Things would have to be pretty bad to get to this point though,
and a more likely symptom of a worn mech is that it's noisy in use
and that it's no longer able to open the octave key pads high enough
to provide sufficient venting.
An easy way to test for wear is to use a similar test as the one
use to diagnose a binding mech.
With the crook off the horn, grip the pin key 'twixt forefinger
and thumb and lift it up. It will tilt the swivel arm, which in
turn will bring down the touchpiece - but before any of that happens,
were you able to lift the pin any distance before either the swivel
arm moved or the touchpiece dropped? If so, that's what's known
as 'lost motion' or 'double action' - the distance a linked key
(one that's connected to another) moves before the key it's linked
to moves. On an octave mech a single key may well be linked to three
or four other keys...and if all of these are worn then by the time
you get to the last key in the chain, most of movement you put into
the first key has been lost along the way.
the difference between stirring your cup of tea with a spoon or
a strip of paper.
Now place a finger over the pin key to hold it down and then press
the octave touchpiece down - this will force the body key to rise
and will close the G key. But before any of that happens, how far
could you move the touchpiece before the body key began to rise?
grip the touchpiece and hold it halfway down (it needs to be gripped,
not pressed). Give the pin key a bit of a wiggle - and take note
of what moves and by how much.
Finally, grip each component key in turn and give it a wiggle -
back and forth, side-to-side.
You're bound to see some movement going on - there will always
be a little free play in the mech otherwise it's likely to be too
stiff in use - and you'll likely see the most amount of movement
in the swivel when you hold the touchpiece halfway down and wiggle
the pin key, but if any of the other tests show more that a millimetre
or two of free play then it's probably safe to say you've got some
wear and tear in the mech. Or it might have been built that way.
Cheaper horns, especially ultra-cheap ones, can have very variable
mechs. Sometimes they're fine, sometimes not so great - and sometimes
they can be completely awful.
Here's a typical example of a not-so-great octave mech. In this
instance the mech was built this way - but it's also how an otherwise
well-built but badly worn mech might appear. There are at least
five points of free play in the mech; going from right to left it
has free play at the fork and pin connector off the thumb key, wear
in the upper socket of the swivel pin, wear in the central pivot
pin, wear in the lower socket of the swivel pin and wear in the
key barrels of the swivel mech itself (though this cannot be seen
Watch the action at the far right and notice how the thumb key fork
appears to bounce at the top of its travel. This is the free play
being taken up by the mech. It's barely a millimetre or so, but
as we'll soon see it has a knock-on effect.
next point of wear is the upper socket of the swivel pin - though
it's quite hard to see it...so skip to the central pivot. If you
watch the centre pin you'll see a gap around it that travels from
top to bottom as the swivel arm moves up and down. Again, it's only
a small gap...but its effects will be added to that of the free
play on the thumb key.
Now watch the lower end of the swivel arm, on the left. You can
see that as the swivel arm rises there's a delay before it raises
the pin key. You can see it even more clearly as the arm comes down
- the pin key comes to rest, but the swivel arm continues to drop.
It's still not by much, but it's cumulative.
Watching the whole mech in action you can see that the fork off
the thumb key drops three or four millimetres before the pin key
starts to rise - and at the touchpiece this will translate into
about a third of the total throw of the key. That's an awful lot
of lost motion, and aside from it making the octave mech feel clunky
and unresponsive, it'll also mean that neither octave key pad will
open quite as far as it ought to...and in some cases perhaps not
So what can be done about it? From the home tweaker's perspective,
not a lot. By far the easiest issue to deal with is the free play
between the thumb key fork and the corresponding pin on the swivel
mech. Simply closing up the jaws of the fork a tad will remove the
free play - but only a tad...if you overdo it, it'll bind. But check
the condition of any buffers/tubes first (see below), because the
fix might be something as simple as replacing them. The play in
the swivel arm ends can sometimes be tweaked by closing up the sockets.
If you look at them closely you'll see they have a slot cut into
them. By carefully pinching the sockets you can reduce the diameter
of the hole - but it's not a terribly neat or long-lasting solution
because it makes the sockets tapered.
variation on the theme is slotted tips on the swivel bar. These
typically fit into a hole cut directly into a key arm. The purpose
of the slot is to allow for the tips to be expanded to take up any
wear. It's a neat idea, but in practice it tends to be less than
satisfactory because the tips are quite weak and the force that
gets put through the mech means the holes apply a sort of crimping
action to the tips...which closes the slots and requires them to
be opened out again.
When they've gone through this cycle a few times they get progressively
weaker...and then half of the tip snaps off during the tightening
process (typically performed by inserting a tapered blade into the
slot). Depending on how and where it breaks it might not affect
the working of the mech too much (other than making it a bit loose)...at
least until the point where the other half of the tip breaks off.
I really wouldn't recommend attempting to tighten these tips yourself
- leave it to a repairer. If they break them, they'll have to fix
them on their own time.
A more long-term fix is to to ream the holes out slightly larger
and fit Teflon tubes over the swivel arm tips. Incidentally - if
you see one of these tips which looks like the slot has been filled
with a silvery-grey material, you can bet your last Rico Royal that
a clumsy repairer has busted the tip and stuck it back on with soft
solder. I see an awful lot of old Selmers that sport this particular
'fix'. It works though - it's just not an especially neat repair.
Dealing with play in the central pin is generally much harder.
In the shot above you can see that there's a slot cut into the pin
- just like the slots on the swivel arm tips. It works on exactly
the same principle - you insert a tool into the slot and splay out
the two sides of the pin. However, unlike the ball-like shape of
the swivel arm tips, this pin is cylindrical...so splaying the sides
out makes it tapered. As such it's a bit like stuffing a beermat
under the leg of a wobbly table...it works, after a fashion, but
it's not a reliable long-term fix. And, just as with the swivel
arm tips, there's a chance the pin might break in two during the
operation. In short, it's a just best left to a repairer.
In the above instance a proper repair would typically involve
swedging the key barrels to remove any free play in the main pivot,
and Teflon sleeves fitted to the central pin and the swivel arm
ends. This will mean having to either ream out the socket holes
to fit, or reducing the diameter of the pins...both of which methods
will require specialist tools. Thereafter, any wear at these points
can be dealt with by replacing the sleeves. For a horn where originality
is required to be maintained, the repair is likely to be rather
Keep an eye open for any missing or worn/damaged buffers (strips
of cork or felt, or plastic tube fitted over pins). A missing or
worn buffer will translate into free play in the mech, and rather
than signalling an appointment with a repairer to have the mech
serviced, it may just need a bit of cork or a piece of tubing.
take note of the shape of any forks. The arms should run parallel
to each other and the pin that sits between them should, with its
buffer fitted, be at least a reasonably close fit no matter where
it sits along the length of the fork. If the arms aren't parallel
then the fit of the pin will vary depending on where it is in the
fork, and this will lead to lost motion at certain points during
the touchpiece key's operation and similarly less movement of the
Here you can see three forks as fitted to a typical thumb key.
The top fork is spot on - the arms are parallel, and no matter where
the pin sits it's always the same distance from the arms (the faded
pin showing its position when the touchpiece is down).
The middle fork has been opened out and the arms are no longer parallel.
When the touchpiece is in the down position you can see that the
pin is a good fit - but when the touchpiece is raised there's now
a larger gap between the pin and arms. This will lead to lost motion
when the touchpiece is pressed...it will need to take up that gap
before the upper arm of the fork contacts the pin and starts to
bring it down.
The lower fork has been compressed at its end, and the effects are
the reverse of the opened out fork. In the up position the pin fits
nicely between the arms, but in the down position it's a much looser
fit. This will lead to lost motion on the upstroke of the thumb
And while we're discussing the thumb key, a good place to look
for wear and tear is where the key is suspended between its pillars.
A spot of wear here is unlikely to cause problems with note production,
but it will make the octave mech feel imprecise and clunky - and
it may also make it noisy. It's very easy to test for wear; simply
grip one end of the key and give it a wiggle. Try to move the key
from side-to-side as well as back and forth - and repeat the test
at the other end of the key for good measure. There should be no
movement at all - a key properly mounted between point screws should
only be free to move up and down.
are exceptions though. Some modern Selmers use a sprung piston inside
the key barrel - the idea being that the springs automatically take
up any free play. And they do, but the design means the key can
never be as snug a fit as one mounted on proper point screws. That
said, the movement you might be able to detect should still only
be barely detectable. Those of you with older Selmers (MkVI etc.)
will be delighted to hear that you have proper point screws fitted.
Dealing with any wear can be difficult for the home tweaker as
it typically requires specialist tools and techniques, but a quick
peek at the screw(s) holding the key in place will show you whether
there's anything you can tweak.
There are two ways this key can be fitted to a horn - either with
a single rod or hinge screw that runs right through the key's barrel,
or by a pair of opposing point or pivot screws. If you're not sure
how your thumb key is mounted, take a look at each of the pillars
in turns. If they both have slotted screw heads in them then the
key is mounted on point screws - if only one pillar shows a slotted
screw head, the key is mounted on a rod screw.
In the latter case you're a bit stuffed...taking up wear in a rod
screw mounted key is a job for a professional, but if point screws
are used there's half a chance you might be able to take up the
play by carefully tightening up each of the screws in turn.
Whether or not this'll work depends on how much wear there is, what
type of point screw is fitted and whether or not they'll actually
move. There's certainly no harm in trying provided you follow a
few simple rules.
a screwdriver that fits the slot - not too wide, not too short.
Note the position of the slot(s) before you start - if things go
wrong you might need to return the screws to this position. Turn
one screw just a quarter of a turn clockwise to tighten it and check
that the thumb key still moves freely before turning the other screw.
If the key stiffens up or binds, back off the screw you just tightened
until it frees up again.
You can be firm with these screws, but not forceful - if you really
have to go at it to make them move, you're much more likely to do
This horn is a Mauriat, which uses pseudo point screws. Tightening
pseudo points up will usually make little difference to the tightness
of the key. If the screw is loose, it'll certainly help to tighten
it up - and it'll make a small improvement, but it simply doesn't
have the built-in adjustability of a properly tapered point screw.
In this instance I was able to reduce some of the side-to-side play
but not the end-to-end play. Fixing this will require rather more
If you're lucky enough to have a horn fitted with proper point
screws, and there's some adjustment left in them, you may find that
when you tighten the screw up until the key binds and then slacken
it off, the screw isn't tight enough in the pillar to stay put...or
it turns as you operate the key.
solution for this is to lock the screw in place. Ideally a drop
of proper threadlock is what you need. Loctite 222 (low strength)
is the stuff to get - but it's not widely available and can be a
little temperamental if there's any oil around (which there usually
is). Loctite 243 (medium strength) is more commonly available, and
will tolerate a bit of oil and grease. Simply back the screw out
to expose the threads, pop a tiny drop of threadlock on the tip
of a smaller screwdriver or a cocktail stick, then carefully apply
it to the threads.
you can see, you really don't need much - just enough to wet the
threads nearest the head. If you overdo it and flood the screw,
soak up the excess with a bit of tissue paper. Tighten up the screw
until the key binds, back it off a touch and leave the threadlock
to set (ten minutes ought to do it). You can then make a further
small adjustment if necessary.
If you don't have any Loctite handy you can try a drop of nail varnish
- or you can wrap a little cotton thread around the screw threads.
Whatever you do, though, don't be tempted to use any other kind
of threadlock than those mentioned. Some of them will lock such
a small screw permanently in place.
can, to some extent, tighten up the rod screws too. These sometimes
work loose over time, and this can introduce a little free play
into the mech. Follow the same rules - though this time the screw
must be turned until it locks into place. You'll feel it when it
happens, at which point you can give the screw juuuuust a tiny pinch
up to drive it fully home.
There's a caveat, though. If there's some wear in the key, or the
rod's a little bit bent, you might find that tightening the screw
up causes the key to bind...in which case you have no choice but
to back the screw out to a position where the key moves freely again.
This'll need fixing properly at some point.
You can find more details about screws, key barrels and the action
in general here.
Wear in vintage mechs is slightly easier to diagnose because although
the mechs might be complex in structure (with various key arms flying
about here and there), the actual components parts are relatively
crude. You'll have fewer swivelling gadgets and fork/pin connectors
to contend with and, essentially, most of the checks will be concerned
with looking for wear in individual keys. The tests you'll use are
exactly the same as those used to check the rest of a horn's action
for wear and tear - as in the link above.
This relative crudeness is both the mech's advantage and its Achilles
heel because you can set up a worn vintage mech in such a fashion
that it will work just fine (if a little clumsily) - it's just a
matter of making allowances for a bit of looseness - but unlike
a modern mech it has very little tolerance for wear that occurs
after the mech has been set up (unless all the keys wear more or
on pin key
Insufficient travel on pin key
You'll have seen, hopefully, from the preamble to the troubleshooting
section that there's a direct relationship between the distance
the pin key rises and the amount the crook key opens. This assumes
that there's very little free play between the pin key at rest and
the crook key ring, and that the motion transmitted by pressing
the thumb key isn't being lost within the mech because of wear and
tear or poor build quality.
the first port of call has to be the section on checking for a worn/badly
Assuming your mech passes that test, the next thing to check is
that the ring key is correctly
positioned. When the mech is at rest there should be a small
amount of clearance between the pin key and the crook key ring...ideally
around a millimetre or so. In some instances (perhaps due to a worn
or poorly-built mech) a greater distance may be required - but once
you get much beyond 3mm it's a fair bet that you have a problem.
With this much clearance a great deal of the pin's movement is
going to be taken up by moving from its resting position to a point
where it contacts the crook key ring...after which point any movement
that's left in the mech is all that'll be available for raising
the crook key. If it's only a couple of millimetres or so it's likely
to mean that you'll hear hissing
from the upper octave notes, and there's a chance that the body
octave key won't be opening as much as it ought to. If the mech
passes that test too, then your problems are likely to be down to
how much 'throw' is available in the mech.
If your mech was working fine previously then it means that something
has changed - and this could be down to damage (pillars being knocked
out of alignment or part of the mech binding) or something obstructing
the movement of the various parts. Perhaps a cork or felt has fallen
off and got itself trapped in the mech - or, more commonly, half
of it has come free and has folded up on itself...thus doubling
its effective thickness. Check out the section on over-thick
pin key corks.
Worn/badly made mech
Body/crook key not rising
Over-thick buffer beneath
Over-strong crook key spring
The crook key spring has two functions; the first is to ensure
that, at rest, the octave key pad is held securely against the pip
- and it must do so with just enough force to maintain a
good seal against the air pressure present in the bore of the horn
when you play it. The second is to be strong enough to provide power
back down the octave key pin to assist in opening the body octave
key - and to resist any 'stray' power coming up the pin when the
body key is in use. This shouldn't happen on a modern swivel
mech, but what with friction, wear and poor manufacture, it's not
uncommon to see the pin key move a little when the body key is in
play. Provided the spring is of sufficient strength it'll easily
be able to resist this relatively weak force. On an older mech,
however, the power supplied by the closing force of the crook key
spring may be vital to the proper operation of the entire mech.
As a general rule, an overly strong crook key spring won't prevent
a modern swivel mech from working. Provided the mech's in good condition
all it will do is make the mech feel stiff and sluggish when playing
from top A upwards. But if the mech's worn or badly made there'll
be a certain amount of free play that will need to be 'taken up'
before the pin key moves...and this can be problematic. Let's say,
for example, that when you press the octave touchpiece key down,
it moves 5 millimetres - but due to free play in the mech this only
results in the pin key moving half as much (2.5 millimetres). In
other words you've lost half the motion you put in.
the crook key spring is strong enough it might have enough power
to resist what little power remains in the mech and push back against
the pin key...and in so doing it will take up the free play in the
opposite direction - and as it amounts to half the throw of the
octave key, it'll effectively cancel out all the motion. If you're
unsure as to whether your octave mech has excessive free play in
it, check out the section on worn
or badly made octave mechs.
It's rare that things ever get quite this bad, and what usually
happens is that a strong crook key spring will have enough power
to resist some - but not all - of the power. In this case the result
is that the crook doesn't open quite as much as it would were its
spring a little weaker.
Things are often far more serious with an older mech, where the
power from the crook key spring may have to be balanced against
other springs in the mech - so that in certain conditions one spring
is able to overcome the power of another.
it's reasonably easy to diagnose this problem it's much harder to
fix it. It raises the question of whether the crook key spring is
too strong...or the other springs within the mech are too weak.
Sure, it often only involves tweaking a spring to reduce the power
it provides - but because you can run into as many problems with
an overly weak spring, there's a need to get the spring tension
just right...and on a vintage mech you may well have to adjust a
couple of other springs as well.
It's the sort of job even an experienced repairer always approaches
with least a little caution, and more than half an eye on what effects
any tweaks will have on the rest of the mech. In other words it's
a job best left to a professional. Consider that a friendly warning
- and if you still wish to proceed, be sure to read the caveats
at the end of this section...
To adjust the crook key spring you're going to have to remove the
crook key. To be honest, it is possible to adjust the spring
without removing the key - but it's a tricky job, and unless you
really know what you're doing (and what you can get away with) the
chances of breaking the spring are very high. In most cases it breaks
off at the screw hole (as shown above), but it's entirely possible
to snap a spring clean in half.
the key is pretty simple - it's usually just a matter of removing
the rod screw that the key pivots on. In some cases there may be
a pair of point screws; if so, remove one screw and see whether
the key can be lifted off. If not, back the other screw out a bit
and try again. It's unlikely you'll need to remove both screws.
In both cases, though, set the screw aside in a safe place....if
you lose it, you'll ruin your day.
You should now be able to see the spring on the underside of the
key. Note the (very) tiny screw that holds it in place. If, when
bending the spring, you allow this screw to take all of the strain,
it may well be pulled right out of the key...stripping the threads
in the key as it does so. It's a small risk, admittedly, but still
a risk nonetheless.
My advice is to undo this little screw and remove the spring. Put
the screw in a safe place. A really safe place.
see that the spring is curved when at rest. The more curved it is,
the stronger the force it will provide when fitted to the key -
and thus the less curved, the less force. As it's too strong it'll
need straightening out slightly. If the spring isn't too short and
thick you may well be able to bend it with your fingers - simply
grip each end and carefully bend it backwards against the curvature.
Chances are it'll be too small and fiddly, so you'll need some
a couple of pairs of pliers. Ideal these should have smooth (not
serrated) jaws - but if you don't have any, just wrap some tape
over the jaws.
The first pair is placed over the end with the screw hole in it.
I like to place it just forward of the hole to ensure that no stress
is placed on this very weak part of the spring. With the spring
gripped firmly you can now use the other pair of pliers to adjust
Don't simply clamp the pliers down on the tip of the spring. More
often that not the tip has been turned back on itself slightly so
as to provide a smooth surface against the spring cradle. If you
damage it, the spring might bind in use.
Your best bet is to grip the spring just a little further forward
of its centre, towards the tip. Now, very carefully bend the spring
backwards. And that's it. You've now weakened the spring. Whether
it's by enough or too much remains to be seen - and you won't know
that until you refit the key.
the spring can be a damn fiddly job. Holding the spring and the
screw in position while trying to get the screw to engage on the
threads in the key is a bit of a parlour game - but there are a
couple of tricks. You can pop the screw through the hole in the
spring and use the spring to guide the screw into place. One little
shake of the hand, though, and out pops the screw - so do this job
over a sheet of white paper on top of an uncluttered table. It'll
help if you're able to clamp the key down in some fashion (it'll
be one less thing that can move). I'm using a lump of adhesive tack,
and with the key secure I can lift the spring (with its screw fitted)
over the top of the screw socket. At this point you want to carefully
insert the tip of the screwdriver into the slot and use it, and
the spring, to guide the screw down into the socket.
you can squish some adhesive tack onto the end of your screwdriver,
then push the screw into it...ensuring that the tip of the blade
sits in the slot on the screw. You can also use a magnetised screwdriver...but
this often ends up making things worse.
What's vitally important is that you don't force the screw home.
It should require no force at all to get the threads started - and
at all times you must ensure the screw is held straight. If you
'cross' the thread you'll end up in a lot of trouble...and you may
even break the screw.
So that's the 'proper' method. You can probably see that if you
need to make more than one adjustment to the spring, all that taking
it off and putting it back on again is going to get very tiresome...so
here's how I do it.
Slacken the screw off just a tad - just enough to allow you to push
the spring to one side. You should now have enough room to grip
the spring near its base (assuming your pliers are small enough).
As before, use another pair of pliers to make the adjustment...then
push the spring back into line and tighten down the screw.
you've a fair few jobs like this you soon learn how to adjust the
spring using only a single pair of pliers - but I'm not going to
explain how to do this on the basis that even a slight miscalculation
will break the spring or rip the screw out of its socket.
If you've overdone the bending and the spring is now too weak,
strengthening it up is less of a faff. As you'll be increasing the
curvature of the spring, the force you'll need to apply will push
the base of the spring down onto the key...which means there's no
risk of pulling the screw out. Just grip the spring in the centre
and give it a bit of a tweak. Just a bit, not too much.
refitting the key, take the time to clean out the key barrel (use
a pipe cleaner dampened with a drop of cigarette lighter fluid/naphtha)
and pop a drop of fresh oil on it. Give the spring cradle a wipe
too, then pop a drop of oil (or grease) on it. This will help the
spring tip to slide in the cradle. Lower the key onto the crook,
locate the tip of the spring in its cradle then push the key down
until the barrel sits between the pillars...then push the rod screw
in. If you're dealing with point screws, locate the barrel on the
tip of the screw you left in place, then line up the spring in its
cradle before pushing the barrel down between the pillars.
with many of the techniques described in this guide, there are caveats
and gotchas to take into account - the chief of which is that of
the spring profile (the shape of its curve).
There's rarely a great deal of room between the underside of the
crook key and the crook tube, and you'll nearly always find that
the spring has to span the key's barrel...which may reduce the available
space even further. In some cases it's not possible to fit a gracefully
curved spring into this space, and rather sharper curves and kinks
are needed. When dealing with such springs it's highly likely that
any adjustment you make to these specific bends will result in the
tweaked spring touching the key barrel before the key fully opens.
It may also mean that that the foot of the spring (which sits in
the spring cradle) effectively moves back towards the centre of
the spring (think of it as walking on your heels rather than tiptoeing),
or the middle of the spring hits the underside of the crook key.
All of these situations will lead to unpredictable results - and
in your efforts to correct them the chances of breaking the spring
the home tweaker your best bet is to think of the bent/sharply curved
section of the spring as being a no-man's-land. Don't go there,
don't touch it. Any adjustments you need to make to the spring will
have to be made before the bend or after it. And yes, it does rather
limit what you can do. Replacing the spring with a thinner/thicker
one is often the best solution to adjusting the spring tension.
If nothing else, it'll certainly pay dividends to start by spending
some time opening and closing the crook key while you watch how
the spring moves - and note which parts are likely to foul against
the barrel/body if you bend the spring in a particular direction.
Owners of cheap curved sopranos should be particularly aware of
this problem, as it's very common for the crook key spring to be
set heavy in order to disguise manufacturing problems in the mech.
Typically the mech will work, but feels awful - but if you weaken
the spring, the mech is likely to stop working altogether. There's
also the issue of the spring's profile (the way in which it's bent
so that it fits under the key barrel). I've covered this, briefly,
in this article - but suffice
to say that getting the profile right on such springs is less of
a science and more of an art.
As for what strength the spring should be set at - well, that's
a 'how long is a piece of string' conundrum. There's certainly a
minimum strength - and that's described in the section on weak,
dislodged or broken springs - but as far as I'm aware there
aren't any convenient standards for setting the maximum practical
strength of the spring other than 'experience tells me that feels
about right'...which is a fat lot of good to you. So I've made one
I made up a set of balls of various sizes/weights by rolling up
a load of adhesive tack and stuck them on to a variety of crook
keys. By juggling my balls around (what? what??) I found that when
a ball weighing 25 grammes (about 30mm in diameter) is stuck to
the top of the octave key cup and the crook turned upside down,
the key would be 'in balance'...which is to say that it opened slightly
and then came to rest about a centimetre above the octave key pip.
Of course, it's heavily dependent on the size and design of the
key - but it at least represents a practical maximum.
For tenors it should be just fine - and eminently workable for most
altos. Sopranos and baritones can go a bit lighter.
If anyone can come up with a better method, please feel free to
let me know.
Worn or badly made mechs
Weak, dislodged or broken
crook key spring
Weak, dislodged or broken crook key spring
As much as I like to encourage players to carry out their own maintenance
and basic tweakery, I think it's fair to point out that fiddling
around with the springs on an octave mech unless you really, really
know what you're doing is liable to end in tears, fits of uncontrolled
rage, acute despair, regret, remorse and feelings of inadequacy
that may only be quelled by consuming vast quantities of alcohol.
By itself, tweaking a spring isn't a particularly difficult job.
Sure, it carries with it the risk that the spring might break but
aside from that it's really just a matter of knowing how much to
tweak the spring and for what purpose. However, when you're dealing
with a number of springs (anything from 3 to 5 of them, sometimes
even more) that must be carefully set so that they're all balanced
against each other, the chances of your getting it right on the
basis of having read an article about it are pretty slim.
So the emphasis for this section is "Look, diagnose and understand
what's gone wrong - but think very carefully before attempting a
If the problem is down to a broken spring, it's going to mean a
trip to the repairer.
There are a couple of workarounds though, the most well-known being
the judicial placement of elastic bands to act as temporary springs.
How successful this is depends on finding an elastic band of suitable
size and strength, and whether or not there's any way of attaching
it to the keywork in such a fashion that it duplicates the action
of the broken spring (i.e. opening or closing a key).
Keep in mind that all a spring does is provide a certain force in
a certain direction, and that you can often achieve the same effect
by the use of weights. For example, wrapping a small elastic band
over the crook key will duplicate the action of the spring that's
usually fitted to this key - but exactly the same effect can be
had by sticking a lump of adhesive tack on the key cup.
OK, so it looks bloody silly, and while playing your eyes will be
inexorably drawn towards it - thus making you appear cross-eyed
to onlookers...but hey, the horn will work, and that's what counts.
springs are rather easier to deal with, flat springs particularly
so. These are fixed to the key at one end, usually by means of a
small screw, and the other rests on the body on the instrument...typically
in a spring cradle in the form of a raised block or a slot cut into
a plate. It's not common for a flat spring's tip to jump out of
its cradle, so any fix should involve trying to work out how the
spring became dislodged in the first place. Perhaps the screw that
holds it onto the key is loose? Because the spring is always under
tension, it would take a fair old knock to make it jump out of it
cradle...unless the spring had become weaker for some reason. This
is an unlikely possibility (but a possibility nonetheless) and what's
more likely is that the spring has a crack in it and isn't far of
Another cause might be a spot of overenthusiastic cleaning. If you
stuffed a cloth beneath the crook key to clean off a stubborn mark,
you might easily have pushed the flat spring out of its cradle.
reseat a flat spring, carefully insert the tip of a small screwdriver
underneath the spring as close to the tip as can be managed, and
gently lift it up and over into the cradle. Try not to simply push
the spring over...it'll likely mark the body. If there's neither
a channel or a cradle, and the spring tip simply sits on the body,
just position in line with the centre of the key.
And bear in mind that this is likely to be very temporary fix if
the cause of the problem is a loose spring. If there's no spring
cradle, the spring will probably slide over again the moment the
key is operated. You could try sticking a blob of adhesive tack
either side of the spring until such times as the screw can be properly
Weak crook key springs are quite rare. They either turn up on (cheap)
brand new horns or they're a short-lived prelude to the spring breaking
altogether. Any other cause is usually man-made (poor tweakery).
They can be rather tricky to diagnose, because a weak spring will
do half of its intended job (to bring the crook key pad down onto
the octave key pip) but won't have sufficient strength to provide
any power to operate the rest of the octave key mech - and it may
well allow the crook key pad to leak.
So how light is too light? That's quite a tricky question to answer,
because so much depends on the design and size of the crook key
- and the mech it connects to - but there are a couple of tests
you can carry out that will help.
The first of these is a simple blow test. Take the crook in your
hand, turn it upside down so that the crook key cup is facing the
floor then place the palm of your hand over the mouthpiece end to
seal it up, put the other end (the tenon sleeve) against your lips...and
blow hard. You don't need to make yourself go blue in the face,
but you do need to give it a firm blow. The key's spring should
be able to resist the force of the air pressure combined with the
downward weight of the key. If it can't, you'll hear/feel a leak
from the octave key pad. It's a crude test - and yes, you're a bit
stuffed if you own a soprano with a single-piece body - but it's
a start. Owners of horns with properly underslung crooks (such as
the Conn 6M) will not need to invert the crook.
All this assumes your octave key pad is sealing properly - so if
you detect a leak, flip the crook over, put your palm over the tenon
sleeve and blow down the mouthpiece end as before. If there's no
leak now, it's all down to the spring.
The second test is to check for the amount of bounce when the key
is operated smartly. Fit the crook to the horn then hit the octave
key smartly. The crook key will jump open and will most likely bounce
a little. You'll hear this as a slight, quick rattle for a fraction
of a second - and if you lift the crook key up and hit the octave
key again and you'll hear that the slight rattle has gone.
If you hear a very distinct rattle or even another clunk as the
key bounces, it's likely to mean the crook key spring is on the
final test is to increase the strength of the spring to see whether
it improves the operation of the octave key mech and solves your
problem - but rather than muck about with the spring we're simply
going to use the trick for dealing with a broken spring and add
a little extra weight to the crook key. Grab yourself some adhesive
tack and pull off a pea-sized ball. This'll weigh just over a gramme.
Stick it on top of the crook key cup. If you haven't got any adhesive
tack handy you'll have to improvise with a piece of tape and something
small and light (though at a pinch you could use a lump of cheese).
The extra weight will increase the power the spring passes back
down to the mech - and if your problem related to a weak crook key
spring, this'll cure it temporarily...or for as long as you want
to gig with a lump of adhesive tack/cheese stuck to your crook key.
Owners of properly underslung horns are out of luck with this test...unless
they're able to play the horn upside down.
If your tests indicate a weak crook key spring, it'll need to be
strengthened. The technique for doing this is described in the section
on dealing with over-strong
crook key springs - all you have to do is increase the curvature
of the spring rather than flattening it out.
Dealing with needle springs is a rather more complicated affair
inasmuch as they often work in pairs or trios in an octave mech
- which is to say the operation of the mech relies on the strength
of one spring being balanced against the strength of another. What
that means is that if you tweak the strength of one spring, you
may well have to tweak the strength of another...and another. And
you may have to keep going back and forth until you find the balance.
Even worse, if you don't really know whether the mech is working
properly in the first place you may well find yourself going round
and around in a circle until such times as you break one of the
springs. There is, however, one simple problems you can tackle with
springs are typically fixed into pillars, and are seated on cradles
suspended from the key barrels or cut into key arms. It's slightly
more common to find that they've 'jumped the cradle' - and a somewhat
lighter knock to the horn than that required to dislodge a flat
spring may be the culprit.
If this happens, however, it usually means that the spring angle
in relation to the body of the horn is wrong - so that as well as
the spring wanting to push to one side or the other, it also has
a tendency to push down towards the body. As a general rule of thumb
the spring should stick out of its pillar perpendicular to a line
running from top to bottom down the pillar...and a very slight angle
upwards is an even better bet.
reseat a needle spring it is simply pushed back underneath the cradle
and then lifted up an allowed to spring back against it - but it's
not always easy to work out which way is 'back'. Some springs are
so lightly tensioned that when they become unseated they sit more
or less directly beneath the cradle. If you're lucky there will
only be one way in which the spring can go (typically when the cradle
is cut into a key arm), but a suspended cradle often offers no clues
as to which side of it the spring should sit.
If you really insist on tweaking the springs, here's how you go
Note the original position of the spring on its cradle. Unhook it
from its cradle. The tip of the spring should now come to rest to
one side of the cradle. In this case the spring has come towards
me before coming to rest, and in order to make it stronger I need
to pull it even further towards me. If I wanted to weaken it I'd
have to push it back away from me.
I'm now pulling the spring towards me. Because it's a spring it's
going to resist the pull, so I'll need to pull it a little way past
this point of resistance so that when I let it go it'll come to
rest a little further towards me than it was before.
There's a bit of risk involved because it's during this tweakery
that the spring may break. You can avoid this by removing the key
and using a special pair of spring pliers to manipulate the spring...but
almost no-one uses the damn things because hardly anyone's got that
much time to spare.
your springs are in good shape and are of good quality they'll certainly
tolerate a bit of tweaking in this manner - and if they're stainless
springs (they'll be silver in colour) they're unlikely to break
no matter what you do to them. Just be careful, OK?
Once you've tweaked the spring to your satisfaction, push it back
beneath the cradle then lift it up to locate it in place. Test the
key to see whether the modified spring tension has had the desired
And it may well do - but as mentioned earlier, you may also find
that another part of the mech doesn't work now because it relied
on the tension of the spring you adjusted being 'just so'. So now
you'll have to adjust the tension of that other key...which may
well affect yet another key, or it might affect the key you just
adjusted. Or the mech might simply not work at all.
Fortunately there's a handy tool that will sort this problem out
in a jiffy....it's called 'decades of experience'. I know that probably
sounds glib - but trust me, it's not.
For those of you with modern swivelling mechs, the outlook isn't
quite so bleak because there are only three springs to deal with,
including the crook key spring.
This leave just two springs powering the mech, and they are the
touchpiece spring and the G key spring. You can tighten these two
springs up as much as you like and the mech will likely carry on
working. It might feel awful in use, but it won't come to a grinding
Over-strong body key spring (vintage/bari)
The body octave key on modern horns (baritone excepted) is usually
unsprung - and is raised or closed by leverage. Many vintage horns,
however, rely on a spring to open the body key. Closure of the key
is via a separate lever, which is sprung to oppose the upward force
of the body key's spring - and thus there is a relationship between
the two springs. In short, the force of the closing spring must
be sufficient to overcome that of the opening spring.
It's a relatively simple arrangement, and if you have any problems
in this area it's more likely to be down to friction (where the
closing arm contacts the body key) rather than a springing issue.
The tension of the opening spring should be quite light - merely
enough to lift the body key pad clear of the octave key pip in a
brisk fashion, and at least strong enough to overcome any slight
stickiness of the pad. If the spring is set much stronger than this
you may find that the action becomes sluggish, and it will accentuate
any friction-related problems in the mechanism. Popping a small
drop of oil where the closing arm contacts the body key will show
if you have a problem with friction - and while it's tempting to
leave the oil in place and rely on it for the smooth operation of
the mech, it'll soon fail (either the oil will run off the key or
it will dissolve the glue holding the buffer corks/felts on, which
will then drop off) and your best bet would be to replace the buffer
with something a bit less grippy (Teflon sheet, for example).
Assuming the opening spring is set correctly and that there are
no friction-related problems, it may be that the closing spring
is a bit on the weak side. This would be unusual, given that the
spring operates the thumb key...and the whole mech relies on its
closing power - but it will do little harm to increase the tension
of this spring. The worst it will do is make the thumb key action
a bit on the heavy side.
If the body key spring is in good condition and easily accessible
(some have flat springs fitted, the adjustment of which requires
removal of the key) it's possible to play around with the tension
without causing too many problems. Set it a little (and only a little)
stronger/weaker and see what happens. If it makes matters worse
it's a simple job to return the spring to its former tension.
Regulation buffer/tube missing
You can put regulation buffers and/or tubes in to two categories;
there are those that have no critical purpose other than to quieten
the noise of what would otherwise be metal-to-metal contact, and
there are those without which the mech will function unpredictably
or not at all.
In the former case the mech just carries on working - rattling away
each time the octave key is pressed. I see a lot of mechs like this,
especially on pro players' horns - because the rattling is a mere
inconvenience rather than a show-stopping fault. In the latter case
the rattling will still be there - but some parts of the mech will
now not be able to move as much as they ought to, with the result
that one or other of the key cups may not open or close.
So, a rattly mech is a sign that a buffer may have fallen off -
though it's also often a sign that the mech needs lubricating.
modern swivelling mech typically has three main regulation points;
a buffer beneath the touchpiece key, to limit how far it can be
pressed down; a tube over the lower pin, to allow contact with the
touchpiece arm forks without making too much noise and a buffer
beneath the pin key, which sets the height of the pin and the touchpiece
It's this latter buffer that's the most common cause of problems,
either because it compresses over time or simply falls off. You
can see that the pin is held away from the body by a couple of millimetres
or so - and that's usually about as far away as it ever needs to
be. If you're getting a nasty clunk and you switch from octave G
to A, this cork is likely to be the prime suspect. You can usually
replace this buffer without causing any major issues.
On older mechs, however, this buffer may also have a more important
role with regard to the internal regulation of the mech, and as
such its thickness may be directly linked to the proper opening
and closure of the body octave key (see Over-thick
buffer beneath pin key).
In an ideal world the mech should be able to function without
any of these buffers - it might be noisy and clunky, it might not
work terribly well, but it should still work.
In the real world though, many mechs aren't built well enough to
cope with the extremes of movement that this would bring, and will
lock up completely. Some might only lock up partially - and this
can lead to parts of the mech being forced to operate when they
A fair analogy of this effect is the old cartoon comedy routine
with the power drill. A hole is being drilling into something, and
at some point the drill bit catches. There's a slight pause, and
then the drill spins round...along with whoever/whatever is holding
it. Things are still going round...but they're the wrong things.
The best way to work out whether a buffer is missing is to understand
what each part of the mech does, and to have an understanding of
some basic mechanics. I know it sounds very complicated, but a great
deal of it is just common sense.
example, if you find that part of the mech is hitting the body of
the sax then it's a done deal that it will make a noise when it
does so. This, then, is a likely spot for a buffer.
If one of the keys seems to move quite some way before it contacts
another, perhaps a bit of padding between them would help.
Generally-speaking there should be no metal-to-metal contact - save
for that on the swivel arm (the central pivot pin and each end of
the swivel arm...though even these are sometimes buffered with thin
tubes) - and if you find any it's a fair bet that that's the source
of your problems.
Missing buffers will need to be replaced - but that's where things
really do get rather complicated, because some mechs rely on certain
keys moving a very particular distance. No more, no less. Too thin
or thick a buffer and everything goes to pot. As with adjusting
the spring tension within a mech, it really is a job best left to
You can, however, experiment with bits of electrical insulating
tape. Just wrap the tape around the area where you think the buffer
should go, and see how the mech reacts. If all seems well you can
remove the tape and fit a thin buffer in its place. If it looks
like the buffer needs to be thicker, add another layer of tape and
buffers that need to slide over other keys (quite common on older
mechs) you may find that a piece of plain cork generates too much
friction, and the mech will stall halfway through its operation.
This is why you'll often see synthetic buffers (Teflon sheet, tech
cork etc.) fitted to octave mechs - but as a temporary fix you can
just pop a drop of oil on the cork. It's not likely to last very
long, but it'll certainly get you out of trouble for a few weeks.
Fork and pin connectors are typically buffered with a plastic (usually
Teflon) tube, and the only requirement is that the tube must fit
on the pin and between the forks. Replacing them is simply a matter
of finding a piece of tubing that fits - and a good source of such
tubing is plain old heatshrink. It's not as sturdy as Teflon tube,
but it tends to be more readily available, it's quieter in use and
you don't have to be quite so precise about the size required. You
can either use a piece of tube that's a little large and shrink
it down with a bit of heat (from a hair dryer or a gas gun, if you
have one and are able to use it carefully) or you can find a piece
that's a snug fit. You can even stretch a tube that's too small
by poking a couple of small screwdriver blades down the tube and
prising it apart. If going for the shrink method I recommend placing
a very tiny drop of superglue on the pin beforehand. This will prevent
the tube from slipping off later...but you'll have to be rather
quick when fitting it.
You probably won't need to dismantle any keywork - just push the
tube in place and then cut it off flush with the end of the pin.
You could cut it to size beforehand, but it tends to make the job
tubes that are present, but just about to fall off (a very common
problem) - remove them, clean the pin up and degrease it, then pop
a drop of superglue on the pin and shove the tube smartly back on
to the pin. And I mean smartly... you'll likely only get one shot
might also find that while a buffer tube is present, it's nonetheless
damaged - and if the damaged portion is right at the point where
the two keys rub, it may as well be missing.
It's obviously best to replace damaged tubes, but a bodge can be
quite effective. In this instance the pin key buffer tube has torn
right at the point where it rubs against the crook key ring. You
could pop a drop of superglue on it and press it back into place,
but a better bet would be to remove the tube, turn it over and refit
it to the tube...so that the damaged part lies clear of the contact
area. And then you could glue it down for good measure.
Assuming all seems well, you
can try oiling the mech
to see if that helps - but be careful not to overdo it. Drenching
the mech in oil/grease won't help.
Over-thick buffer beneath
Over-thick buffer beneath pin key
While this is undoubtedly a fault that can scupper some octave
mechs, it's not likely to be one that occurs spontaneously. The
various corks, felts and tubes used as buffers within the mech often
get thinner over time, or fall off completely...but they seldom
grow in size. Hence this problem tends to be confined to mech that
have recently been incorrectly tweaked - in other words, someone's
done something wrong.
there is one instance in which a buffer can increase in thickness,
and that's when part of it comes unglued and folds itself under
the remaining cork or felt.
This is a shot of a pin key arm, and if you look closely at the
cork fitted to the bottom of it you'll see that the right hand end
has come unglued from the arm and has managed to tuck itself under
the rest of the corks that's still glued on. So where you previously
had a buffer cork of a couple of millimetres, you now have one that's
near enough twice as thick.
How significant a problem this is will depend of the design and
state of your mech. If there's a lot of wear in the mech, this sudden
increase in the buffer thickness may well just be absorbed by the
free play - and about the only symptom you might notice is a less
responsive mech and perhaps slight loss of tone on some notes.
As for the fix, simply lift the key, roll the cork out flat and
re-glue it in position...though you'd be better off replacing it
completely. If the rolled-over portion falls off you might find
that there's still enough cork remaining to provide buffering for
the time being...and there's always the option of carefully slicing
the remaining cork off the key, moving it up the key a little and
re-glueing it in place.
However, some mechs, particularly older ones or 'active' ones,
rely on the pin key resting in a certain position to ensure the
body key pad remains closed when the G key is pressed without the
octave key (for low G, for example), and in some cases this may
mean that the pin key needs to 'float' in mid air with some clearance
beneath it - which is to say that its resting position is entirely
dependent on other keys that act upon it (typically the G key foot).
On such mechs the pin key buffer cork is largely superfluous, and
its function (if it even has one at all) is merely to quieten any
metal-to-metal contact that may occur due to the key bouncing. An
overly-thick pin key buffer here is likely to prevent the body key
from closing, and may well have a similar effect on the G key.
The section on bent pin
keys goes into a little more detail about this, specifically
on an active mech fitted to a vintage Buescher.
on pin key
Bent pin key
Body/crook key not rising enough (hissing)
Stuffiness (often accompanied by a hissing sound) in the octaves
is often due to insufficient pad clearance over the octave key pips.
Every tonehole on the sax relies on a certain amount of 'venting'
in order to produce a clean, clear note - which is to say that there's
a relationship between the quality of the note produced and the
height of the pad above that note's tonehole. It's not a particularly
critical relationship, but a pad that sits too closely over the
tonehole may result in a stuffy tone and a slightly flattened note
- and one that sits too high may result in some instability in the
note and perhaps a touch of sharpness.
But there's plenty of room for manoeuvre; on a scale of 1 to 10,
where 1 is a pad closed against the tone hole and 10 is the most
open a key can be (i.e. with no buffer corks fitted to the key foot),
then the 'ideal' venting height would be in the region of 4 to 7.
Thus you can set the venting to suit the horn or the player's preference.
Some may favour a low action, with the slight loss of tone being
an acceptable trade-off for a faster feel to the action - and some
may favour a very open action, so as to get the maximum clarity
from each note at the expense of a little dexterity.
And so it is with the octave key pads - though because the toneholes
are so small the actual difference between positions 4 and 7 (as
described above) may be barely a couple of millimetres - but while
it's true to say that having the pad set too low may cause some
tonal problems, it's rather less likely that having it set too high
(higher than '7') will.
The ideal height is generally around 3 or 4mm. If the crook key
pad opens much more than this then there's a good chance that you
have 'wasted motion' in your octave key mechanism...which usually
means that you're having to press the thumb key down a lot further
than needs be.
Bear in mind that it's not a given that a lack of octave key venting
will cause problems - plenty of horns work just fine with less venting,
but it's very much dependent on the horn and/or the player and their
You can test if the venting is the cause of your problems by playing
a suitable note in the octave range and have an assistant gently
raise the relevant octave key cup. Even if the body key appears
to be fully open, poking a pencil under the arm of the cup (not
under the pad) will gain you another millimetre or so - which is
usually enough to confirm/discount any problems. The crook key can
simply be lifted by hand.
opening height of the body octave key is dependent on the 'throw'
of the octave thumb key (the distance it can be pushed down) and/or
the distance the G key foot rises when the G key is closed. Press
the G key down then press the octave key and note how far the body
octave key rises. If it touches the G key foot before you've completely
depressed the thumb key then it's the G key foot that's limiting
its opening. If you can depress the octave key fully before the
body octave key touches the G key foot then it's the thumb key that's
determining how far the key opens.
If the G key is the 'limiter' then the way to increase the opening
of the body octave key will be to increase the amount the G key
foot rises. This can be done by reducing the thickness of the buffer
on the G key foot (relatively easy job) or by bending the foot upwards
(not so easy, best left to a repairer). Either way it will raise
the G key action...and that may not be a good thing.
If the thumb key is the limiter it may be rather more difficult
to increase its throw.
There are typically two places where the thumb key contacts the
body of the horn; beneath the touchpiece and/or beneath the lever
arm that connects the thumb key to the octave mech.
Press the thumb key down and note where it contacts the body. If
you're lucky the only point of contact will be beneath the touchpiece.
If you're even luckier there will be a rather thick piece of cork
or felt acting as a buffer. To increase the throw of the key, sand
or shave a little off the buffer. If it's a very thin buffer it
might not be possible (or worthwhile) removing any material, and
you'll be better off bending the touchpiece up a little. As before,
this is a job best left to a repairer - and you should bear in mind
that it will mean the touchpiece will sit slightly higher than before.
Ergonomically-speaking this is usually not a good thing.
If the point of contact is beneath the lever arm then your options
are limited to adjusting the thickness of the buffer between it
and the body of the horn. It's possible to bend the lever arm, but
this is very definitely not a job to take lightly (take it to a
Another way to increase the throw of a modern swivelling mech is
to thin out the buffer beneath the pin key. This will lower the
pin key and raise the thumb key, which in turn allows the body key
to open more. However, this is messing with the internal regulation
of the mech and may result in unexpected and unwanted behaviour
- so I wouldn't recommend it as a home fix.
the opening height of the crook key cup is an easier job - simple
a matter of bending the key - but you first have to determine whether
or not it will make any difference. Just as the opening height of
the body key was dependent on the throw of the thumb key or the
G key foot, so the crook key is dependent on the throw of the octave
With the instrument assembled and the crook lined up in its usual
playing position, note the size of the gap between the crook key
ring and the octave key pin.
It should be around 1mm - if it's any more than this then there's
a chance that bending the crook key to close this gap will give
you a bit more clearance at the pip.
Now gently press the octave key and watch what happens to the gap
between the pin and the ring. As you depress the thumb key the pin
should rise, make contact with the ring and then open the crook
key. Pay particular attention to the moment where the pin contacts
the ring. Is there any lost motion in the rest of the octave key
mechanism...are you able to press the thumb key down a little way
further before the pin begins to raise the crook key?
If so this is typical of wear in the mechanism - and this lost motion
will limit the effectiveness of bending the crook key, and you might
have better results getting the mechanism tightened up.
Assuming all is well, though, and you have a 1mm gap between the
ring and the pin, increasing the throw of the crook key will be
entirely dependent on increasing the throw of the thumb key (as
above). If, however, the gap is larger than 1mm - and there's no
appreciable lost motion in the mechanism - you may well find that
bending the crook key will give you the results you're looking for.
As for how to do this, and what problems might arise from it (there's
always something, isn't there?) see the section on bending
the crook key.
With all that said, it's sometimes possible to experience hissing
in spite of the octave mech being in good order. Some horns are
just plain hissy, and some mouthpiece/reed combinations can highlight
It may also be be down to the choice of pad material. Cork octave
key pads are a popular option these days - they're hard-wearing
and tend not to stick, but some players find that they increase
the amount of octave key hiss. A cork pad with a domed face may
help, as might a synthetic pad - such as one made from Sugru. It
can also be the case that a leather pad has got rather hard with
age, and this too can lead to hissing.
If in doubt, swap it out.
on pin key
Bent crook key
Bent crook key
A bent crook key is by far and away the most common cause of octave
key mech problems, and in almost every case the cause of it will
be incorrect handling of the crook. When fitting the crook to the
horn, many players will casually wrap a hand around the crook and
press/turn the crook into the body. If the palm of the hand presses
down hard enough on the octave key, it'll splay it out - and this
usually means that the crook pad will no longer close onto the octave
key pip. It can also lead to pulldown,
where the tubing of the crook itself gets bent. You'll find details
of a safe grip at the end of this section.
In some instances the key gets bent when the crook is taken out
of the case - and I've actually seen people lift it out by the key.
And then there's just bad luck or a bit of knock.
But don't despair, it's something that happens to the best of us
at one point or another. Even quite experienced players can get
caught out - it just takes a moment's inattention or perhaps an
unfamiliar horn and whoops...you've bent the crook key. Fortunately
it's rarely a difficult problem to solve once you've diagnosed it.
People (like me) generally refer to bending a crook key up or down
- but it's more accurate to refer to bending the key open or closed.
This is because the key typically forms a C shape - by compressing
the ends of the key you can make the C smaller, or by pulling them
apart you make the C larger...and in each case this will affect
the resting position of both the pad and the key's ring.
While it's the case that most (intentional) bending of the key is
to do with restoring functionality to the octave key mech, it's
also the case that bending the key can improve the feel of the mech.
Too much of a gap between the pin key and the ring can lead to an
unresponsive and clunky mech - whereas too little can cause unexpected
opening of the crook key.
Let's begin with some diagnosis.
crook key that's been bent up (or open) is usually pretty easy to
spot because when the crook is fitted to the horn, the octave key
pad doesn't close over the pip. The bend has expanded the C shape
and the ring makes contact with the octave key pin way before the
pad reaches the pip.
As you can see, it's pretty obvious...once it's been pointed out
However, this is a reasonably severe example - and the functionality
of the octave mech will be just as impaired if the pad is held off
by only a tiny, barely visible amount - so if you're in any doubt
about the state of the key, you'll need to test it.
A play test is simple enough - just play and hold any note below
top A with your left hand, and gently press the crook key pad with
your right. If the crook key pad is being held off a tiny amount,
the pressure of your finger on it will be enough to close it, and
you should notice an immediate increase in the strength of your
tone and the ease of playing. Try the test playing a low G followed
by an octave G, as it's sometimes easier to hear/feel the difference
in one or other of the octaves.
that doesn't work for you, the cigarette paper test will. This simply
involves putting a cigarette paper between the octave key pip and
the pad (with the horn assembled), letting the key drop down and
then gently pulling the paper out. You should feel some resistance.
It won't be much, but it'll definitely be there. If you haven't
got a cigarette paper handy you can use a thin piece of cellophane.
At this stage it's a fairly crude test, and even a piece of writing
paper will probably do. Try doing the test with the crook off the
horn first...just to give you an idea of how much drag the pad exerts
on the paper, then fit it to the horn and test it again.
If you feel no resistance it's a dead cert that the pad's being
held off - but if you feel it's a bit of a borderline call, your
next point of diagnosis is where the pin key meets the crook key
ring. Be aware, though, that this test can be fooled by a pad that
doesn't seat all the way round - and while you have the feeler in
your hand you might as well go through the process of testing
for a leaking pad at the same time.
you're looking for here is a tiny gap between the pin and the ring
when the crook is in its playing position, as shown in the photo
on the left. It's not much, maybe a millimetre or so. If there's
no gap it means there's a good chance that the pin is preventing
the crook key pad from fully closing. On a really well set up horn
this gap might barely be visible - in which case you can use the
cigarette paper, placed between the pin and the ring, to test for
One gotcha to watch out for is a key that's been bent so that part
of it touches the crook. It's usually pretty easy to spot this,
and it'll typically be at the rear of the ring where it curves up
the crook, but a simple and effective test is to take the crook
off the horn, hold it by the mouthpiece cork, lift the octave key
pad and let it fall. I call this the drop-down test. If you hear
a ding it may mean that part of the key is hitting the crook. It
may also mean that the key just needs oiling, or that the barrel
is so worn that the key is able to wobble about.
And now for a crook key that's been bent down (or closed).
This is a bit trickier to diagnose because it'll pass the drop-down
test and it'll pass the cigarette paper under the pad test - and
it's where the ring key meets the pin where all the problems will
The big give-away here is going to be a very noticeable gap between
the pin key and the ring.
the crook key that's been bent up, this problem can have degrees
of severity. At its worst the crook key will completely fail to
open - and on some horns this will also mean that the body octave
key won't open either (the opening force being provided by the crook
key) - but with a less severe bend it might mean that the octave
key pads are able to open partially. You'll get some functionality,
but the mech will feel clunky and the tone production will be poor
(and often accompanied by a hissing
Having figured out that the crook key is bent, and in which direction,
it's time to look at how to remedy the situation.
Bending keys is never a job you should take on lightly. Not only
is there a very real risk of making things worse, there's also the
slim chance that it will go catastrophically wrong and you'll end
up with a broken key.
It has to be said, though, that you'd really have to go at it hammer
and tongs to break the crook key. We're not talking about wrenching
bits of metal here...rather it's more a case of gently easing the
key back into the required shape. I might have to bend the key back
and forth half a dozen times before I get it to where I want it,
and I've never yet broken one. That's not to say it can't happen
though - but if it does there's a good chance that there was a crack
in the key and it was always going to break sooner or later anyway.
This won't be the end of the world, but it will mean having to shell
out money to have someone fix it.
I should also say that it's not always the case that bending a
key is the solution to the problem you're having, even when it appears
to be - so take your time with the diagnosis.
With that dire warning in mind, you'll be relieved to hear that
bending a crook key is really quite simple once you know what to
do and how to do it - and that it's a technique that ought to be
as familiar to sax players as changing a reed.
First up, we're going to look at bending the crook key down - or
compressing/closing it - to correct the problem of the crook key
failing to close properly.
Another reason for compressing the key is to increase the size
of the gap between the pin and the ring when the keys are at rest.
The lack of a sufficient gap can sometimes be a problem because,
under certain circumstances, the pin key can move slightly - even
when it's not supposed to. This can be down to friction, or poor
build quality or just plain old wear and tear - and it can sometimes
be the case that although there's a gap when the mech is at rest,
there isn't one when its in use.
With the horn assembled, finger a G and press the octave key. Watch
the crook key pin - it shouldn't move at all, or at worst only very
slightly. If it rises to contact the ring it might mean that it's
just down to friction in the mech, and there won't be enough power
in it to open the crook key. Try pushing the pin gently down - you
should feel no resistance.
You can also try lifting the crook key and letting it fall back
down. You might hear a little 'ding' as the ring hits the pin -
but if you lift and drop the crook key again, you shouldn't hear
a ding (because it should have knocked the pin down).
If the pin rises enough to open the crook key, or you get that
ding on the second drop, you've probably got a bit of unwanted movement
in the octave key mech - and the proper fix for this would be to
have it repaired. In the meantime, though, you can compress the
crook key so that when the ring is at rest, it's clear of the pin's
maximum point of unwanted movement.
key designs vary, so you'll have to adapt this technique to suit
- but the essential principle is this. The key is opened, something
is placed beneath the ring section to prevent the key from closing
and pressure is applied over the key cup. In this shot I'm using
my thumb to hold the key open. This provides an ideal soft surface
against the crook tube and a reasonably firm surface (the thumb
nail/bone) against the ring. It both helps to prevent damage to
the tube, and helps to limit the amount of pressure applied to the
pad cup...mostly because it bloody hurts. If the pain is too much
to bear, wrap some cloth around your thumb or wear thick gloves.
To apply pressure to the key cup I've got my thumb under the crook
tube and my forefinger on top of the key so that I can apply a squeezing
The big question is...how hard should you press the key cup down?
There's no real answer to that one, I'm afraid - how hard you'll
need to press will depend on the stiffness and design of the key,
and it's the sort of thing you can only gauge by experience (and
even then it's largely guesswork). So, go gently. The key will probably
be quite springy - which means going too lightly will do nothing
at all. So be firm, but cautious - do a little, check a little.
If nothing seems to be moving, give it bit more welly - and if that
doesn't work, give it even more (and by now your thumb will be really
Some keys can be very tough indeed, but with this method of bending
you're very unlikely to do any real damage (bruises aside).
Horns with pseudo-underslung crook keys (where part of the key goes
under the crook but the octave key pad sits on top of the tube,
such as on Yanagisawas) can be very hard to bend...so be prepared
for a bit of swearing.
On no account should you do away with the thumb and replace it with
anything else (bits of plastic/wood etc.), because it can all go
very badly wrong very quickly indeed.
What you're aiming to achieve is for the ring key to be around
1-2mm away from the pin key when at rest. If your mech's a good
one, 'at rest' will mean whenever the crook key isn't in play -
if the mech has a few problems then 'at rest' will have to take
any unwanted movement of the pin key into account. This may mean
having to settle for a wider gap when the octave mech is truly at
rest, which will mean a slightly more clunky action when going straight
from a lower octave note to one that uses the crook key (say middle
C to top C).
If you've overdone the bending (and it's not uncommon, even for
a pro), you'll need to open the key up a tad...
out the crook key is much the same as compressing it, with the force
being applied in the opposite direction - and this time it's applied
to the ring.
As before, a thumb is used to protect one end of the key - the key
Lift the key and place your thumb over the octave key pip and let
the pad come to rest on top of your thumb. This will raise the ring
and give you more room for bending it, and it'll also prevent the
octave key pad from being squashed onto the pip - which won't do
the pad any good at all.
Pressure is applied to the ring in a downward/backward plane. Simply
pulling the ring down will work, but this lowers the ring's position
on the pin key and can sometimes lead to problems.
It looks like you could just pull the pad cup arm upwards - and
this too will work, but it's more likely to put an unattractive
bend in the cup arm and there's the potential to drive the back
of the ring into the crook tube.
How much pressure to apply is uncertain as it was when compressing
the key, but opening the key out is usually a little easier than
And as before, fit the crook to the body and check as you go along.
You'll probably find you need to cycle between compressing and opening
the key a few times before you get it just where you want it - and
you might find it more effective to 'pulse' the pressure for finer
adjustments rather than giving the key a single, solid squeeze.
You can also take a look at the article on making
and fitting synthetic octave key pads, which contains a section
on bending the octave key - and shows a couple of variations on
the two grips mentioned above.
to bear in mind when adjusting the key is your playing position
- specifically the position you usually put the crook in prior to
The crook key ring is supposed to allow for a little variation in
the angle of the crook (you might, for example, turn the crook round
a little more when playing in a seated position). Ideally the gap
between the pin and the ring should remain unchanged no matter what
angle you put the crook at (within reason), but some rings are better
made than others (or haven't been squashed out of round) and you
might find that the size of the gap varies with the crook position.
If so you'll have to set it at the most extreme position you're
likely to use and put up with a slightly larger gap in any other
position. See the section on distorted
rings for more details.
mentioned at the start of this section, it's important to get into
the habit of forming a good grip when fitting, removing or adjusting
There are many variation on the theme, but the vital points of consideration
are that little or no pressure is applied to the crook key and the
downward/upward force should be as close to the tenon sleeve as
The grip shown here probably looks much like the grip you're already
using, but the detail is in the point of pressure. The forefinger
and thumb have a good grip on the tube - as does the crooked little
finger (against the fleshy part of the base of the finger on the
other side of the tube). Anything inbetween is relaxed and loose.
It's a little like picking up a cactus, or a bramble or rose stem.
You find a couple of clear spots where you can get a grip, and the
middle of your hand instinctively arches upwards to avoid getting
stabbed by a spine or a thorn.
All of the pushing/pulling force is applied via the forefinger and
thumb. The little finger provides stability and sideways force so
that you're able to slightly twist the crook back and forth while
fitting or removing it.
An easy way to test whether you've got the right grip is to operate
the crook key with your other hand. It should be free to move. It
won't move much as it'll hit your hand almost immediately, but it
at least tells you that there's little or no downward pressure on
the key. And because your little finger is crooked underneath the
crook tube there'll be little or no downward force on that end of
Body/crook key not rising
Bent pin key
Leaking octave key pad
Pulldown means you have a bent crook - and as the term suggest,
it's bent downwards.
is quite a common problem, and aside from damage due to an obvious
impact it usually comes about due to excessive force when fitting
the mouthpiece or when fitting the crook to the horn.
More often than not the only visible sign is that the crook key
won't close - which leads to most players suspecting the problem
is just a bent octave key. However, a trained eye can usually spot
a slightly bent crook - it often just 'looks wrong'.
This TJ RAW crook has copped a whack and is showing very typical
signs of pulldown. Take a look at the section of tube just below
the crook key's pillar- notice the way the light catches it. It's
quite a subtle thing, but if you hold the crook level in your hands
and tilt it to and fro in the light you'll see that the reflection
behaves differently when it hits the peak of what's now an oval
You should be able to feel the pulldown too, by gently pinching
the crook tube between your finger and thumb and rotating it while
feeling for a slight ovalness. For comparison, pinch the tube just
after the mouthpiece cork and see how it feels - then go back to
the suspected oval section and note how it feels like your fingers
are being pushed slightly apart as you rotate the tube.
may also be small signs of damage...such as a slight indentation
in the tube between the tenon sleeve and the end of the brace that
runs underneath the crook. Sometimes it can be rather more obvious,
such as on this Yanagisawa alto crook, where you can clearly see
a hefty crease beneath the tip of the brace.
There's no home fix for this, it'll require some extremely careful
manipulation of the crook - and if you get it wrong the potential
for disaster is enormous. You can, however, compensate for the pulldown
by bending the octave
key - but while this will work, it leaves the tip of the crook
at a lower angle than before and this can sometimes lead to aches
in the neck and back as well as embouchure problems. It may also
have some slight impact on tuning, though this is rare for small
amounts of pulldown.
It's worth bearing in mind that once a crook has been pulled down,
it tends thereafter to be more prone to it even when fixed - and
this is because the structural integrity of the tube has suffered.
Not a lot you can do about this, other than to be careful when putting
the mouthpiece on or assembling the horn.
Prevention is better than cure, and what really helps is having
a correctly-sized and greased mouthpiece cork. A very general rule
of thumb is that the mouthpiece should slide freely to around halfway
onto the cork before it starts to snug up, and thereafter requires
merely a moderate twist and push to secure it in the playing position.
If it doesn't go on smoothly then you probably need to grease the
cork, and if it binds it may mean the cork is too thick. You might
also find it helpful to fit the mouthpiece to the crook before fitting
the crook to the horn.
Pullup is the reverse, and is much rarer - and usually only happens
due to an impact. However, some players are in the habit of lifting
their horns off their stands by the crook...which is just asking
for trouble. Aside from bending the octave key up to compensate,
the fix for this is the same as pulldown.
For details of a safe grip to use when fitting/removing the crook,
see the end of the article on bent
Bending the crook key
Bent pin key
Bent pin keys account for a sizeable proportion of octave mech
problems, simply because the pin itself is extremely vulnerable
to damage - almost all of which happens when the horn is being placed
in or removed from the case. This is why all horns come supplied
with an end cap - which is, typically, a small plastic plug that
fits into the top of the body tube. This prevents the pin, which
would otherwise be sticking out a centimetre or so, from getting
caught up in the case liner or from impacting against the side wall
of the case.
As a general rule of thumb the octave key pin should lie parallel
to and in line with the centre of the body tube. If that sound confusing,
place a mug on a table and stand a pencil upright next to it (or
just take a peek at some of the photos in this article). This is
the 'technically correct' position of the pin, but it's not uncommon
to find that it's sometimes sitting a little off line - and by little
I mean a couple of degrees. It may have been built this way, or
it may have been deliberately bent for some reason (as a quick fix
or to overcome another problem within the mech). This isn't usually
anything to worry about, though further testing will reveal if it's
If the pin is bent more than a couple of degrees it's highly likely
to have an adverse effect on the mech, with this effect being more
severe the greater the degree of bend.
The plane in which the pin is bent will usually determine how the
mech fails. If it's bent to either side it will make the operation
of the mech highly dependent on the angle of the crook (and thus
the position of the crook key ring). This might mean that your octave
mech won't work with the crook in its usual position...but if you
move it to one side you might find it starts to work again.
the pin has been bent to the left, and you can see the original
position of the pin in translucent red. Because the height the pin
stands off the body hasn't changed, there's half a chance that the
mech will still work. There's usually a gap of a millimetre or so
between the pin and the ring, which is sometimes just enough to
accommodate the change in geometry - but this really depends on
how far the pin has been bent and how circular the ring is (and
most of them aren't all that circular). A temporary fix for this
would be to bend the
crook key down a little, or simply to adjust the angle of the
Note where the pin itself rises out of the key arm (bottom right
of shot), and note too the position of the strengthening brace fitted
to the underside of the crook. If you were looking at this setup
straight on you'd see that the end of the brace would line up with
the tip of the pin, which would line up with the point where the
pin meets its key arm.
Not all horns are built like this - but where they are it's a handy
rough-and-ready reference to how straight the pin is.
If the pin is bent up or down it's rather more likely to stop the
mech dead in its tracks. A pin that's bent up can hold the crook
key open all the time, and this'll mean it'll be all but impossible
to play any note below octave A without sounding like a duck with
a sore throat. The quick fix for this is to bend the crook key down
(or closed). This will restore the necessary clearance between the
pin and the ring, and all should be well.
However, bear in mind that some mechs rely on the pin key resting
a certain height above the body, and if this dimension has changed
you could well find that the body octave key opens unexpectedly.
Much will depend on where the pin key's buffer cork is placed. In
the shot you can see that the cork is fitted beneath the pin key
arm - so unless the arm itself is bent, there's a good chance that
the rest of the mech will remain unaffected by the bent pin.
A pin that's been bent down will result in a dramatic increase
in the distance between the pin and the ring when the mech is at
rest. It could just mean that the crook key rises only very slightly,
or it could mean that this distance is large enough to take up all
of the throw of the octave mech, which would mean that the crook
key would never rise. It can also result in the mech's touchpiece
being moved closer to the body (thus further reducing the throw)
and the body octave key opening when it shouldn't. A quick fix,
assuming the only problem appears to be the failure of the crook
key to open, would be to bend the crook key up a little.
Again, keep in mind that the resting height of the pin key could
affect other parts of the mech - and with the pin itself being bent
down it may well contact the body (or the crook) before the buffer
cork on the pin key does...thus raising the resting height.
shot on the right shows a Buescher alto mech, and being a vintage
horn its mech is 'active'. In this instance the pin key is sprung
so that it always wants to rise - and is prevented from doing so
by being held down by either the G key foot or the thumb key. As
far as the thumb key is concerned its 'holding force' is applied
via a stub attached to the pin key (in red). While the key is at
rest it's in contact with the pin key's stub and the body octave
key (as ringed in the centre). The G key acts on another, smaller,
stub (as ringed on the right) - when G is pressed, its foot rises
and butts up against this stub, thus preventing the pin key from
rising. All of these relationships are critical, and have to be
precisely balanced at all times.
If the pin key was to be bent down it would act like the thickness
of the buffer cork beneath the pin had been increased, and this
in turn would raise the height of that red stub. Because the arm
from the thumb key rests on it, the arm would also be raised - and
it would no longer rest on the body key...which would open.
The bend would also lower the smaller stub - and as you pressed
the G key its foot would rise and meet the stub rather sooner than
expected. It wouldn't be able to move the pin key (because it's
already resting on the body) and this would prevent the G key pad
from closing properly. In other words, the mech's now completely
out of balance - and neither it or the horn will work again until
the pin key has been put back into its proper position.
And all of the above assumes that the mech still works - but it's
entirely possible that the pin key is now binding
on its pivot...in which case it's game over, and nothing's going
to work until you get the mech straightened and freed up. Such collateral
damage notwithstanding, the solution to a bent pin key is obvious:
bend it back to its proper position.
What's not so obvious is how much damage you can do by attempting
to realign the pin key. At best you might chew up the metal because
you don't have the right tools for the job, and at worst you could
snap the pin key off the mech - especially if it's a screw-in one
(see the section on Broken
it's not always the case that it's the pin that bent, rather it's
the arm to which its attached. No
problem then - just bend the key arm, right? Again, you run the
risk of chewing up the metal, but there's also a risk that in bending
the key arm you might distort the barrel on which its mounted...which
may already have been distorted by the impact that bent the pin
key in the first place. And, as many a DIYer has discovered, bending
a bent key back into place often results in simply putting another
bend in the key. By far the most common example of this annoying
phenomenon is that of an enthusiastic parent trying to straighten
out a flute that their kid has just sat on. What was a nasty-looking
but nonetheless gentle bend - and one that's easily dealt with when
you know what you're doing - turns into a double bend...complete
with a nasty crease in the body (usually around a tonehole). Much
wailing and wallet-emptying will ensue. If you feel you're up to
bending the pin I'd recommend you read through the section on dealing
with distorted rings first, as it contains more details about
the geometry of this particular part of the mech - but my advice
would be to leave well alone and take comfort in the fact that you've
successfully diagnosed the problem...and give your repairer a call.
will also pay you to try working out exactly how the pin got bent.
If you dropped the horn on the floor, then stood on it, that's likely
to be a dead give-away...but maybe it was OK yesterday and now suddenly
it's not. This usually means there's been a problem while putting
the horn away or taking it out of the case, and it's nearly always
because the end plug/cap was missing.
The end plug fulfils two functions; it keeps the dust and fluff
out of the crook receiver and it protects the pin key from getting
caught in the lining of the case and being bent. It also protects
the pin if the horn rattles around in the case. You'd be surprised
at just how much force goes through the horn if you, say, knock
your case against a door frame. There'll be a few kilogrammes of
brass moving at around 3 mph that suddenly comes to a complete stop...and
if the pin key isn't protected and prevented from colliding with
the wall of the case, it'll take the full force of the impact.
If you haven't got an end plug, go buy one - they're cheap enough.
A basic plastic one will do, as long as it's a snug fit in the receiver.
A loose one is next to useless, as you'll get tired of the thing
always falling out...and either it gets lost or ends up rattling
around in the accessory compartment because it's too much bother.
If you use a shove-it type swab it'll probably have a built-in end
plug - and if you want something a little more upmarket you can
always ask your repairer to knock you up a custom plug on the lathe.
This won't cost a great deal and it'll be a perfect fit for your
horn and the case...and as daft as it sounds, you might find you
grow rather attached to it. If you're a bit of a cheapskate you
can make your own end plug out of an old champagne cork...though
you might have to go through a few bottles before you find a cork
that's large enough (there's always a downside, eh?).
One last tip - never tighten the crook receiver screw against the
end plug. You'll end up stretching the receiver, which leads to
a loose crook and thus a leak.
Body key not closing
Over-thick buffer beneath
Broken pin key
itself, an octave mech problem due to a broken pin key is pretty
rare. It's far more likely that any impact sufficient to break the
pin will bend the key. However, it does happen from time to time
and can usually be traced to a previously bent pin that was hurriedly
straightened...which resulted in a crack in the pin. Such a repair
will last for a while, but the next knock might be enough to shear
the pin off.
You might also come across pins that are screwed into the key arm
(this one's fitted to a Selmer Balanced Action alto), and if the
thread loosens up it's entirely possible for the pin to work its
way free over time.
not a lot you can do about a broken pin, other than to have a new
one fitted - but a get-you-out-of-trouble fix is to fit a length
of stiff tubing over the remaining stub. It needs to be reasonably
stiff else it'll simply bend when it tries to open the crook key.
A small nail pushed into the tube may help to stiffen it up. Car
windscreen washer tubing is ideal, though you might need a drop
of glue to help prevent it from slipping off.
If there's no stub, or only a very small one, you're a bit stuffed.
As for screw-in pins, things are little easier. It's generally not
too much trouble to remove any stub that might remain, and once
that's done it's a simple enough job to make a new bolt to replace
the broken one. Which is just as well, because pins like these tend
to be a bit fragile.
Players who've bought new crooks may sometimes find that the pin
key is too short to contact the ring of the crook key.
In this instance it might be possible to bend the crook key to bring
the ring into play - but so doing usually throws out the geometry
of the key, and you have to resort to bending the pad cup arm to
correct it. This is fairly advanced key bending, and is best left
to a professional.
Another solution is to have an extension fitted to the pin key -
and if it's done properly it should mean that you'll still be able
to use the original crook (which is particularly important for players
who've bought alternative crooks rather than replacement ones).
The windscreen washer tube trick may also work, and if you like
switching between various crooks it's possible to have a set of
removable sleeves made up to suit each crook.
Bent pin key
A distorted crook key ring can have unpredictable effects on the
functionality of the octave mech depending on precisely how and
where it's been bent out of shape.
This is distinct from a ring that's too high or low due to a bent
crook key - and refers more to the shape of the ring rather
than its position. The unpredictability comes about because there
may be some positions of the crook where the relationship between
the pin and the ring is just fine, and others where it's not. You
might find the horn works fine when you're playing in a standing
position, but when you sit down to play and move the crook slightly
to the side, the octave mech no longer works.
To understand the nature of the problem we must first understand
how the ring key works when it's in good shape.
In an ideal world the ring should form part of a perfect circle.
The pin key sits close to the inner diameter, and no matter what
position you put the crook in, the relationship between the pin
and ring remains the same.
In the real world octave key rings are seldom completely circular
- and there's really no need for them to be because the average
player is only going to want to move the crook a limited distance
away from its central point...which, for the sake of argument, can
be thought of as having the mouthpiece cork roughly in line with
the thumb hook (note, not the octave key thumb rest...which is usually
offset to the left).
From this position a player might move the crook a little to the
left or the right for the sake of comfort, and perhaps a bit further
to accommodate different playing positions (such as sitting down).
Going much beyond this would suggest a very unusual playing position,
and there's a chance you'd need to have the ring specifically tailored
to your needs. The amount you can move the crook and still maintain
the functionality of the octave mech is called the swing.
Before we get started on the diagnosis, there's something we have
to be sure of - and that's the state of the pin key. Is it straight
and in line with the body of the horn? If the pin
key is bent it'll duplicate many of the symptoms common to a
distorted ring, and while you may well be able to fix those symptoms
by moving the ring around you'll be better off putting the pin right.
the crook in its normal position. The red mark on the ring lies
directly over the centre of the pin. Note the small gap between
the pin and the ring - this is very important.
Octave mech are seldom very precise - and even if they start out
that way, they soon begin to wear. So this little gap is a margin
of error - it's small enough not to cause any noticeable clunkiness
in operation, and large enough
to ensure that when the pin key is at rest it can't hold the crook
key open. It also allows for a little stiffness or temporary binding
in the mech which might otherwise briefly lift the crook key a fraction
- which usually results in a very unexpected squeak.
With this setup everything's right with the world, and any bum notes
you play on the gig are going to be down to your lack of practice
(or the amount of beer you've had).
Now the crook's been tilted to the left (as you hold the horn in
front of you). You might favour this position if you're in the habit
of slinging the bell at an angle across your stomach, which reduces
the risk of a nasty accident when you're doing a spot of marching
Note the new position of the red mark, and note too how that gap
between the pin and the ring has all but disappeared. It's just
about there...but only just, and this position represents the extreme
at which the octave mech will remain functional. Move the crook
any further to the left and you'll effectively run out of the ring's
circularity...and the pin will force the crook key open.
time the crook's been moved the same distance to the right. This
is a very common position for seated players, where the bell rests
on the right thigh.
Again, you can see how far the red mid point has moved, and you
can also see that there's still a gap between the pin and the ring.
It's a little larger than when the crook was tilted to the left,
and that's because this ring is ever so slightly out of round. Not
by much, and not by enough to be of concern - but enough to serve
as an example of how the circularity of the ring can affect the
You should be able to see that's it possible, and quite easy, to
check the geometry of the crook key ring simply by fitting the crook
to the horn, centring it and then then turning it from side to side
while keeping an eye on its position relative to the pin key. You're
almost certain to see some change in the relationship, but provided
the gap between the pin and the ring doesn't disappear or increase
dramatically you're not likely to have any problems with this part
of the mech...assuming the rest of it's in good order.
When the ring gets bent it throws the geometry out, and one of
three things is likely to happen; the gap between the pin and the
ring disappears, the gap increases in size...or the whole ring gets
shunted to one side, which could close the gap or increase it depending
on which direction the ring goes.
a simplified diagram showing the effects of the most common ring
The good ring is just fine - you can swing the crook from side to
side at your leisure and the mech will still work.
A squashed ring usually results in an increased gap between the
pin and the ring. The mech might still work but it'll feel clunky,
and the opening heights of the octave pads will be reduced. You
can correct this by bending
the crook key - which will leave you with the situation in the
diagram. This ring will work within a very limited swing. A dead
centre position is fine, but it you swing the crook even a small
amount you'll lose the gap and the crook key pad will be forced
A flattened ring usually leaves you with no gap between the pin
and the ring - and in most cases it means the crook key is already
being held open.
Again, bending the crook key will sort this out - and in the position
shown you're still likely to have enough swing to get by.
An offset ring is a bit of a wild card because its effects will
vary depending on which direction the ring has been pushed. In this
case it's copped a shove to the right and it's closed the gap between
the pin and the ring. As before, you can remedy this by bending
the crook key, but this will leave you with a mech that can only
be swung one way (in this case, to the player's right). If you swing
it left you'll run the pin straight into the ring and force the
crook key open.
But this assumes a simple left or right offset - the ring could
quite easily go another way. Forwards, backwards, a bit forward
and a bit to the side etc...
In such cases I'd recommend focussing on fixing the gap between
the pin and the ring before attempting to move the ring in any other
direction. And keep in mind that if the ring looks more or less
central on the pin, but with no or a large gap, then you're simply
dealing with a bent crook key rather than an offset ring.
DIY fixes? You need to be very careful before tackling a distorted
ring, because it's often the case that the damage will be a combination
of issues. The ring might be squashed, but also offset - or it might
be flattened at an angle...as well as being offset. And on top of
all that, the crook key will probably be bent too.
If you have a vintage horn it's quite likely to have a very simple
crook key - just a single key arm with a relatively thin ring fitted
to it. You'll probably get away with bending this type of key without
too much fuss. But if your crook key has two arms that rise up to
attach to the pad cup arm, you'll find it's a great deal more difficult
to move. Similarly, semi-underslung crook keys (such as those found
on Yanagisawas) will be a real pig to shift without the aid of some
tools - and once you take a tool to a key, you're messing in the
big league - and it can all go very badly wrong very quickly.
You might also have to contend with a key guide - a U-shaped channel
fitted to the crook, in which the crook key sits. Any bending you
do will have to take into account the need for the key arm to clear
the sidewalls of this channel, otherwise it'll bind.
My advice is to do what you can to get the horn going for the gig,
if needs be, and let your repairer sort it out once the emergency's
At this point I should imagine that some of you are looking a your
crook keys and thinking "But mine looks nothing like that!".
The ring, despite its name, doesn't have to be a ring. Some horns
have just a section of a ring...a piece of curved rod about a couple
of centimetres long. Some may have slightly curved flat piece of
In every case the same principles apply to these keys as to the
standard ring design - but by being that much smaller they're often
more difficult to move around...though it's fair to say that they're
also less likely to get bent in the first place.
When it comes to baritone saxes, all bets are off. There may be
no ring at all - or what passes for a ring might be fitted to the
pin key...and the crook key might not even be on the crook itself.
The same principles still apply, though - but the best way to deal
with a bent bari mech is to A: swear a bit and B: Call your repairer.
You have been warned.
Bent pin key
on pin key
Crook key opening unexpectedly
There are a number of reasons why this can happen; sometimes it's
down to wear or free play in the mech, allowing certain parts of
the mech to move when they shouldn't. Sometimes it's down to damage,
with a misaligned key moving too far or not enough and sometimes
it's just down to the build quality of the mech and its geometry.
In the first two cases you just need to correct the anomalies and
all should be well, but in the last case the fix could be rather
most likely cause tends to be a binding mech - and this can be down
to damage or the lubricant in the mech having dried out and gone
all sludgy. Less common is that one or other of the springs has
failed or is incorrectly set. Many mechs rely on a bit of closing
force coming from the crook key - it provides resistance against
the pin key and diverts power along the path of least resistance...which
is typically the body octave key.
Missing corks/buffers are also a common cause, in particular the
buffer that regulates the throw of the octave key touchpiece. Some
mechs are very tolerant of the amount of throw - but others will
only work properly within a defined range of movement. Lose the
cork under the thumb key and it might mean the mech gets pushed
beyond this limit...at which point just about anything could happen.
A quick test for this is to find something to stuff beneath the
thumb key to limit how far it can be pushed down (a bit of folded
card, for example). If the mech now works when the throw is restricted,
you've found the problem. If it doesn't, you've got other issues.
or compressed corks can be a problem, and can often be difficult
to detect unless you know what to look for. The older the horn,
the less tolerant a mech is likely to be of corks and buffers that
sit outside a rather narrow range of thickness - but it can also
be a problem on relatively modern horns.
On the right I'm testing the mech
on a Yamaha 61 alto. Everything looks fine and moves freely but
the notes from mid D to G are rather stiff and unresponsive, and
I've traced it to the crook octave key opening ever so slightly
when it shouldn't. It's doing so because the buffer beneath the
thumb key has worn, and the throw of the key now exceeds what the
mech can handle (it's quite a picky mech in this respect).
There are three points of interest. The first is the thumb key,
the second is the opening height of the body octave key and the
third is the clearance between the pin key and the crook key ring.
The G key is pressed and held down, followed by the octave thumb
key. As I press this key slowly down I'm watching to see when the
body octave key contacts the buffer on the G key's foot. As soon
as contact is made my focus shifts back to the thumb key. Am I able
to push it down any further? When it comes to a stop, can I push
it down still further if I press a little harder?
It seems that I can - at which point the pin key begins to lift
the crook key. In this case the movement was noticeable (I could
see the crook pad rising), but in some cases it may be difficult
to spot any movement. This is where the feeler (a cigarette paper)
comes in handy. Placed between the pin and the ring it should remain
free to move throughout the entire operation. As soon as you detect
the pin gripping the feeler it means the crook key is about to rise.
The fix, in this instance, was to replace the buffer beneath the
thumb key with a slightly thicker one. When I repeat the test I
now find that the thumb key comes to rest at the moment the body
key touches the G key foot, and cannot be forced any further down
(with reasonable force). At this point the feeler is still just
about free to move - which means that the crook key will not open.
You can hopefully see that there's a relationship between the height
the body key opens and the point at which the crook key starts to.
Once the body key has opened as far as it's able to, any further
movement of the thumb key is unnecessary and likely to cause problems.
What this means is that the throw of the G key effectively sets
the throw of the octave key mech - and that should you wish to increase
the amount that the octave keys open, you'll have to increase the
throw of the G key by thinning out the buffer on the G key's foot.
This may or may not be a problem depending how how high or low the
rest of the stack keys are set.
Modern swivel mechs are rather more forgiving and will usually
tolerate a little extra throw on the thumb key - but the feeler
test can still be used. In this case you may notice a little bit
of drag on the feeler as the pin comes up against the ring, but
it should still be far short of actually gripping it.
Keep in mind that the above also applies to a thumb key that's
been bent upwards so that the touchpiece sits too high.
a general rule of thumb (quite literally), the top of the touchpiece
should sit just slightly higher than the top of the thumb rest (the
ideal position is shown in the photo on the left).
Players will vary in their preference; some like it a little higher,
some like it a little lower - and it often depends on the design
and shape of the touchpiece. However, there comes a point where
it becomes impractical (and this can be too low as well as too high)
- and in the shot on the right you
can see that it's clearly way too high.
If your horn arrived like this from new then it's a good chance
that it's a cheap horn that's been poorly assembled - but although
the operation of the thumb key is likely to be cumbersome, it's
also likely that the octave mech has been set up to work with this
setup. If, on the other hand, the thumb key never used to be in
this position it's pretty much a dead cert that it's been bent -
and the most likely cause is careless handling when removing the
horn from its case.
fix for this, unsurprisingly, is to bend the key back into place
- but it's not always as simple as that because if you bend a key
at one end...it'll also bend at the other unless you take steps
to prevent it so doing. It may also bend any other keys that are
linked to it. When you remove the bend, you can't guarantee that
you've removed the bend from the other end of the key, nor that
you've restored the relationship between any of the parts attached
to the key barrel.
In this case we've got the touchpiece at one end and a slotted arm
at the other - which connects to a pin on the swivelling mech. As
the touchpiece got pulled up, the pin will have resisted the force
(if you're lucky) and more or less held the slotted arm in place.
When you bend the touchpiece down, you'll be leaving that slotted
end pretty much where it ended up...and it may now need adjusting
to restore its proper relationship with the touchpiece. And if the
swivel pin bent too, you could be in big trouble.
Unbending keys is a really a job for a professional - but if you're
in a fix and you can't possibly make things worse, you can try this.
Slide a support under the slotted arm. I'm using a lolly stick,
but the thick end of an old reed will do just as well. This'll hold
the arm in place and prevent it being pushed down into the body.
Now press down carefully but firmly on the touchpiece. Don't go
mad, just give it a little push, then remove the support and see
where the touchpiece sits now.
it's still too high, give it another push - but keep in mind that
you'll be better off leaving it a bit too high than pushing so far
down that it's too low...as it's a lot trickier to raise the touchpiece.
It should get you out of trouble...assuming the rest of the mech
didn't get pulled out of line.
Incidentally, the cause of this raised touchpiece isn't a bent
key - it's a dirty great dent under swivel mech's lower pillar.
It might seem odd pointing it out with a bright red arrow, but when
this horn came in for repair the player's complaint was that the
octave key was bent up. The dent had to be pointed out before they
could see it. And the moral of this little illustration is to look
around the general area of the problem, because it may not be what
you think it is.
Finally, unexpected movement can also be a result of a springing
error or a lost cork when reassembling a mech after cleaning. It's
sometimes hard to see how a spring is supposed to be set, or exactly
where a piece of cork was (or how thick it was) - which is why I
recommended taking photos of the mech at the start of this troubleshooting
Weak, dislodged or broken
crook key spring
Worn or squashed pad
mech errors due to worn or squashed pads are rare, but there are
some cases in which a change in the thickness of the pad can cause
leaks. Most mechs will tolerate a little change in pad thickness,
but if it becomes too extreme it may lead to a situation where the
setup of the mech may result in an overly thin pad being held off
A visual diagnosis will often be sufficient, but a test with a feeler
will provide more certainty. Bear in mind that octave key pads tend
to get squashed into a cup shape, and that viewed from side on they
might look fine...but if you lift the key and examine the face of
the pad you might find it doesn't look quite so good.
Here are four octave key pads in various states of functionality.
At the top is a pad in good order. The leather is clean and supple,
the face isn't compressed and there's an even dimple where the pad
sits on the octave pip.
Next down is a slightly scuffed pad. How it got this way is a mystery,
but it may have been down to over-zealous cleaning, or perhaps it's
due to wear in the key which allows the pad to move around over
the octave pip. There may also be some roughness on the pip. The
pad will still work, but it's definitely on its way out and should
be changed reasonably soon.
Third down is a split pad. In this instance the central portion
of the pad looks to have lost its flexibility (due to the constant
cycle of wetting and drying) and has cracked. Repeated operations
have hammered away at the brittle leather and parts of it have disappeared
completely - leaving the felt core exposed. This pad will still
work, but it will be leaking slightly...so there'll be some loss
of tone and response. It should be changed as soon as possible.
The last pad looks intact, but the leather is dry and scuffed -
and it's an even bet that the felt core is just as dry and tough,
When a pad gets this old and hard it simply can't maintain a decent
seal over the octave pip. It probably still just about functions,
but needs changing.
Any pads that look ever worse will need changing immediately.
A handy get-you-through-the-gig tip is to wrap an iffy octave key
pad in clingfilm. Tear off a small square, place it over the face
of the pad then bunch it up over the top of the key cup to hold
it in place. This will also work as a quick fix for a pad that's
See the section on leaking
octave key pads for more details.
Leaking octave key pad
Leaks from octave key pads (assuming they're in good condition)
are reasonably rare. More often than not the cause of a leak is
down to poor regulation rather than a problem with the seat of the
If you suspect one of the pads is leaking the first thing to do
is to visually check the condition of the pad. Does it look clean
and fresh? If it's a leather pad, is the skin in good condition,
with no splits or frayed areas? If it's a cork or a synthetic pad,
does the surface of the pad look even? How does it compare to the
examples shown in the section on worn
or squashed pads?
a general rule a well-seated pad relies on a flat key cup that comes
directly down onto a level tone hole, but many an octave key pad
functions perfectly well even if this ideal can't be achieved. However,
there's a limit to how much 'angle' such pads can accommodate, and
this can leak to leaks from the front or the rear of the pad. This
is especially true of pads that are quite firm...such as those made
Here are three octave pads. The top one is nice and level, and
you can see that it sits evenly on top of the octave key pip.
The middle one is sitting in a cup that's at a slight angle, but
it's not so acute as to cause any problems. Provided the pad is
set in the cup at a suitable angle it will still present a level
face to the octave key pip. Pads set like this are often seen on
poorly made mechs - and a better bet would be to either bend the
key cup arm so that the cup sits level, or fit a thinner pad.
The lower pad doesn't stand a chance. The angle of the key cup
is too acute - and while the rear of the pad is resting on the pip,
the front is being held off and is leaking.
the theory, at least. In practice it can be almost impossible to
spot a leak because the octave key pads and pips are very much smaller
than those in the diagrams.
The simple test of the seal of the crook key pad is to place the
tenon sleeve (the larger end) of the crook against your lips, place
your palm over the cork end to seal it up and blow down the crook
(one some crooks you might need to reverse the lip/palm positions
due to the keywork). You should feel a very immediate and firm resistance
that doesn't fade. You can also try suction - but I tend to feel
that suction is of more use for detecting leaks in the soldered
joints of the octave key pip and the tenon sleeve.
Keep in mind that if the crook
key spring is too weak, the blow test will cause the key to
open and the pad to leak, and this issue will need to be corrected
(and the suction test will not detect this problem).
It's just about possible to test for a leak using a feeler (typically
a thin cigarette paper - silver Rizlas, for example - or some thin
cellophane, such as might be found wrapped around a box of chocolates),
but it requires a steady hand. You'll need to cut a tapered slice
out of the paper, bringing the tip of the feeler to a fine point.
The octave key pad is lifted, the feeler placed over the rim of
the pip and the pad allowed to fall. The feeler is then drawn out
- and the resistance noted. If the pad is doing its job a paper
feeler will probably tear (don't forget to remove any bits from
underneath the pad) - but if you can pull it out with ease then
it probably means you've found a leak.
You have to be careful that the tip of the feeler doesn't bridge
the hole in the pip, as this will give a false reading. Looking
at the lower of the three examples above, you can see that a feeler
so placed would be gripped by the rear of the pad and would show
a seal...even though the front of the pad was leaking. It's a delicate
job because the top of the octave key pip might only be three or
four millimetres across...if that.
This test differs from that used to check that the pad actually
comes down at all (perhaps because of crook
pulldown). In that test the assumption is made that the pad
is seating, and all you're checking for is that there's nothing
preventing the pad from doing its job.
loose or wobbly crook can sometimes be a source of a leak because
it may allow the octave key pad to come down onto the pip slightly
It's pretty rare, given that the key tends to self-centre itself
- but excessive free play could cause problems.
To check how much wobble you have on the key, grasp the key cup,
lift it up above the key guide (if possible - don't force it) and
simply wiggle the key gently from side-to-side. On this key you
can see that the wobble allows the key to go as far as the side
walls of the guide. The horn works just fine with this amount of
wobble, but I'd consider this to be the maximum amount of free play
that's allowable. If you can move the key much more than this then
it's going to be worth having the action tightened up. I might not
necessarily improve the way the horn plays but it'll certainly make
the octave mech feel more snappy in use.
It should be noted, though, that most crook keys are pretty crudely
designed and built - and because of the amount of punishment they
take (particularly when fitting or removing the crook), it seldom
takes long before even the slickest setup shows signs of wear. Correcting
it is often a fairly major job, particularly on keys where the barrel
is fixed to the crook tube and the key fits over it (which you'll
find on just about every modern horn).
certainly never hurts to check that the rod screw isn't loose though,
and this is a simple matter of finding a screwdriver small enough
to fit the head of the screw and giving it a clockwise turn. Snug
it up firmly...and no more. The threads on the screw are very small,
and there's often not a lot of 'meat' on the other side of the key
for them to screw into. A drop of oil when you're done will help.
And while you're poking around in the area, try to have a look
across the rims of the octave key pips. They should be level - but
they aren't always. It might not be so much of an issue with a leather
pad (which will be soft enough to take up the unevenness) but a
cork pad will really struggle to seat on a warped pip.
If you're still not sure whether the pad is leaking or not, try
playing the instrument and have someone press the octave key cups
down firmly. If you notice an improvement in the quality of the
notes you can produce, this probably means one or other of the pads
Assuming you find a leak the solution is to have the pad or pads
replaced. This won't cost a great deal - but if the leaks are due
to misaligned key cups or warped pips, you'll have to dig a little
deeper into your budget.
If the leak is due to a damaged pad you can try wrapping a small
piece of clingfilm over the pad (and key cup). This will provide
a temporary skin, and should keep you going for a while. It's a
simple emergency fix for the crook key pad, but wrapping clingfilm
around the body key pad will be a little more difficult, because
you may have to make an allowance for the G key foot.
Worn or squashed pad
Weak, dislodged or broken
crook key spring
Octave key pip blocked
This is another octave mech problem that can be quite hard to diagnose,
simply because there are no easily visible symptoms. It's also the
case that the diagnostic tests for a blocked pip aren't always conclusive.
For example, if you suspect the crook pip is blocked you ought to
be able to test it with a simple blow/suction test. Place the palm
of one hand over the tube at the corked end, place the other end
against your lips, raise the octave key and blow down the tube (or
suck). If you meet with a firm resistance it'll be because the pip
However, this hardly ever happens - because the pip is only partially
blocked. There's enough air getting through to show a 'pass' on
the test, but not enough air to allow the pip to vent properly -
which results in poor note production and/or air noise (hissing).
And testing the body pip in this fashion isn't really possible.
However, while cleaning the pips is a bit of a faff, it's not a
difficult job - so if you've even the faintest suspicion that they
might be blocked, you might just as well clean them.
for what might be blocking the pips, it's usually a sludgy build-up
of the stuff that gets blown into a horn (saliva, fats, sugars etc.)
or it might be some trapped fluff from a pull-through.
Here's an excellent example of the latter. As you can see, there's
a build-up of gunk around the rim of the octave key tube. It doesn't
extend very far into the tube - in fact it's paper-thin - and it's
made up of collected fibres from the cloth that gets pulled though
the horn to dry the bore out. Each time the cloth passes by the
tube it gets compressed, and fibres catch on the rim's sharp edges.
Over time more and more fibres get caught, and eventually you end
up with a sort of spider's web over the mouth of the tube. This
reduces the working diameter of the tube and results in a loss of
tone and reduced stability on the body octave notes (mid D to G).
Cleaning this gunk out will yield instant and noticeable improvements.
This shot was taken through the F key tonehole after the key was
If you look very closely you can see that the 'web' sits just a
little way in to the tube. This is because pull-throughs have a
very limited ability to clean out the toneholes in the bore and
tend to push gunk into them - where it collects at a point that's
just out of reach of the passing cloth. This is why I tend to recommend
'shove-it' swabs (such as the Padsaver), which do a much better
job of scouring the toneholes. Even an occasional use will help
- as would poking an old toothbrush down the bore every once in
a while and giving the octave pip tube a bit of a brush.
To clean the pips you'll need a pipe cleaner. You can buy these
at most tobacconists (if you can find one these days), and they're
often sold at craft stores. For preference, get tapered ones - the
narrow end will slip inside a pip a little easier than the wider
end. I'd also recommend using some sort of fluid. Cigarette lighter
fluid is ideal, but failing that you can use a little soapy water.
you start, though, it's worth bearing in mind that the part of the
pip that's visible on the outside of the horn is only half the story.
Most pips also have an internally tapered tube that extends into
the bore - which you can see if you peer down the top of the body
tube. The length of this tube varies - on some horns it might be
almost a centimetre in length, on others it might be no longer than
a few millimetres. The body pip tube is nearly always longer than
that on the crook pip.
Cleaning the crook pip is pretty easy on most horns - dampen the
tip of the pipe cleaner with lighter fluid, lift the key up and
carefully insert the pipe cleaner into the pip (you might find access
easier if you put a little bend in the pipe cleaner). Give it a
turn or two and a good wiggle, then pull it out. If the tip looks
dirty, cut it off or use fresh pipe cleaner and repeat the process
until it comes out clean.
gets a bit trickier on single-piece sopranos and modern baritones
(where it's essentially another body pip) because the key cup won't
rise all that far, but it should still be doable. When you're done,
seal up one end of the crook and blow through the other while holding
the key open, just to blow out any fibres that may have been left
Cleaning out the body pip is a bit trickier as you'll have limited
access - unless you fancy removing a few keys. Fortunately the pipe
cleaner is quite flexible, and if you bend the end down (about 5mm
back) you should be able to slide it beneath the pad and into the
pip hole. Hold down the G key and the octave thumb key to raise
the pad and very carefully ease the tip of the pipecleaner into
the pip. Once it's in the hole you'll be able to ease it in still
further. Go carefully though, because you don't want to snag the
pad and tear the skin.
As before, give it a bit of a wiggle and then gently pull it out.
you can go in via the bore - but first you'll need to stiffen up
your pipe cleaner.
Bend it almost it half, so that one leg is slightly longer than
the other by a centimetre or so. Now grip the pipecleaner at the
bend and hold the tip of the short leg against the longer and twist
the cleaner up tightly. Finally, bend the remainder of the longer
leg at right angles. You should end up with something that looks
Dampen the tip and, as before, hold down the G and octave keys
to open the body key pad. Guide the pipecleaner down the bore of
the horn and ease the tip into the pip tube.
The extra stiffness of the twisted pipecleaner should help, but
you still won't be able to apply that much pressure. Just take it
easy, and keep an eye on the top of the pip for the tip of the pipecleaner
poking through. It's unlikely it'll get very far, but you want to
avoid it touching the pad.
before, give it a bit of a wiggle then pull it out and see whether
there's any gunk on it. Repeat as necessary. Blow down the bore
as best you can when done to clear out any loose fibres (a tin of
'air duster', as used for cleaning computers, is ideal for this
job - and can be used to blow dust off the horn's action).
This technique is also good for cleaning crook pips on modern baritones,
where the pip is situated just a little way down from the crook
If you're at all comfortable with removing keys then the whole
job can be made a great deal easier by removing the top F key on
most horns - and removing the octave mech makes it easier still.
There are other, though rarer, causes of blocked pips. It's not
unheard of to find that the pips are blocked up with polishing (buffing)
soap. This is most likely to happen on cheap new horns, or those
that have been recently buffed (prior to relacquering, for example).
It shouldn't happen, but it sometimes does. Buffing soap residue
is quite hard and greasy, so it can be quite stubborn when it comes
to removing it. Fortunately, cigarette lighter fluid will cut into
it like a hot knife through butter - and a little drop applied to
a pipe cleaner will see it off in no time at all.
Hard deposits are rather more difficult to deal with. These will
be down to the formation of scale (calcium carbonate) - just like
the stuff that forms in the bottom of a kettle. It takes a fair
while to build up such an encrustation, so it's more likely to be
an issue with older horns - and while 'soft' cleaning the pip may
help, it won't be long before you run into the same problem again.
If you suspect this is the problem you're suffering with, your best
bet will be to have the pip cleaned out by your repairer because
a certain amount of dismantling of the keywork may be required -
though it wouldn't hurt to give your crook
a proper clean.
Body key not closing fully
Failure of the body octave key to close is a relatively rare problem
on modern octave mechs because the design defaults to the body key
being held down when the mech is at rest. For the most part this
problem is likely to be due to damage to the mech after an impact,
or perhaps due to corrosion seizing up the component parts. Your
first test should be to see whether the mech is moving freely -
details of which you'll find in the section on worn
or badly made mechs.
can also be quite hard to spot that there's a problem because the
downward force of the G key foot tends to compress the pad and squash
it over the octave key pip. This forms a dish in the centre of the
pad, so that the actual sealing part of the pad is somewhat recessed.
When this happens it's possible to find that although the pad looks
as though it's sitting snugly on top of the tip, it actually needs
to drop down another millimetre or so before a seal is made.
The graphic shows what's happening. On the left the pad is seating
firmly down on the pip. On the right the octave key cup is raised
slightly, but because the pad is dished it still looks like it's
seating over the pip - but the seat is now raised, and the blue
line is an invisible gap that would allow air to escape. You can
prod and poke the key cup to see whether it moves, but by far the
best method of checking the integrity of the seal is to check
it with a feeler.
are, however, a couple of other likely suspects, of which a missing
buffer cork under the pin key is the most common. In theory the
loss or thinning of this buffer shouldn't cause any problems on
a modern swivelling mech, other than allowing the octave touchpiece
to sit higher than normal and producing some noise when the pin
key hits the receiver - but if the mech is worn, or perhaps not
as well built as it ought to be, it's just possible that the thickness
of the pin key buffer will become critical.
Clearly, a missing buffer will need to be replaced - but if you're
due on stage in 10 and your friendly neighbourhood horn repairer
is nowhere to be seen, a quick fix is to pop a blob of adhesive
tack underneath the pin key...with a strip of paper over the top
of it to prevent it sticking to the key.
you have the means to fit a new cork, the adhesive tack trick is
helpful in experimenting with the size of buffer required - of which
the three important factors are that the pin should not be able
to hit the receiver, there should
be a gap between the pin and the crook key ring when the mech's
at rest and the body key should be able to close fully.
The other suspect is a crook key that's set too open, so that
the ring key applies a constant pressure to the pin key when at
rest. This would normally result in the crook key being held open,
but if the spring on the crook key is set excessively high it might
have enough power to overcome the force from the pin...which would
mean the crook key closes and forces the body key open. It would
also require the pin key buffer to be overly thin and the mech to
be badly built. As such this is very rare, but I've seen it happen
once or twice.
You're far more likely to come across the body key problem on 'active'
vintage mechs, where there's often a pair of levers that control
the opening and closing of the two octave key pads - and these levers
may be interconnected by yet more levers. The hard part is working
out which lever does what, but as a general rule of thumb you're
likely to find that there'll be a pair of arms coming off the touchpiece
key which control the closing of the two key pads (take a look at
the vintage soprano mech photos in the first part of this article).
Where each of these arms connect to other keys there should be a
small cork or felt buffer installed to prevent metal-on-metal clacking.
If these buffers aren't 'balanced' it'll mean that one or other
of the octave key pads won't close when it ought to - and in the
case of a body key that won't close, it usually means that a buffer
has come off or worn down. If this is the problem then you'll need
to adjust or replace these buffers - but for a quick fix you can
try wrapping some tape around the levers.
It can also be the case that too much buffering is the cause of
the problem, especially on vintage/active mechs that rely on the
pin key being free to 'float' in mid air. See the sections on bent
pin keys and over-thick
buffers which goes into more detail about how the body key on
active mechs functions.
Leaking octave key
Bent crook key
Worn or badly made mech
Over-thick buffer beneath
Bent pin keys
Octave key pip unsoldered or leaking
This is a very pernicious problem because it's often extremely
hard diagnose, mostly because there are rarely any visible pointers.
It's also unexpected, so it tends to be the very last thing players
look for - if they even look for it at all.
There are three main reason why a pip fails in this fashion, the
most common being that the solder has simply degraded over a long
period of time due to Galvanic
Corrosion. The second most common reason is that the pip was
never properly soldered in the first place - and although this is
more likely to be the case on cheap horns, it still happens on quite
expensive ones. And lastly there's 'collateral damage'. This usually
happens when someone's put a dent removal bar down the horn, and
hasn't been too careful to avoid it colliding with the octave key
tubes. In such cases it's very likely that the pip's tube will have
been damaged as well.
If you're very lucky (or unlucky, I suppose) you might be able
to see that the pip is loose. Depending on the design of the pip
it might be possible to see that the soldered joint has broken.
A failed joint has a very distinct look about it...the surface is
lighter than solid solder, and looks rather grainy. You might also
see jagged edges. With the aid of a magnifying glass you might also
be able to see a clear gap where the tube goes into the body. In
extreme cases it's possible for the solder to simply break down
over a long period of time (due to a process called
selective galvanic corrosion) - and it slowly goes from a state
of leaking slightly to being completely loose. Rather amusingly
(for me, at least), when combined with a sticking pad it's sometime
possible to see the pip lift out of the body when the octave key
If nothing's particularly visible or you're still uncertain, there
are a few things you can do to test the integrity of the octave
key pips - of which one of the most effective is the time-honoured
'poke it with a stick' method. Yep, this is a bona fide test.
Find a small wooden stick - a pencil is ideal - and with the crook
off the body finger an octave G. This will lift the G key foot off
the body octave key pad, and although the pad might not rise (because
the crook isn't fitted), it won't have any downward force acting
upon it. Carefully insert the stick down the bore. You'll see a
small tube or protrusion into to the bore around a couple of inches
down - this is the body octave pip tube. Position the tip of the
stick over the end of the tube and give it a gentle push outwards.
If the pip has come unsoldered you'll be able to push the tube out
of the body.
same procedure can be repeated for the crook octave key pip - though
on some horns this might be quite difficult (on tenors, for example),
and you'll be better off trying to pinch the pip from the outside
between your thumb and forefinger and giving it a bit of a tug.
The job is a great deal easier if you have reasonable fingernails.
Assuming all is well it just means that neither of the pips have
come completely unsoldered - but it still means that they might
be partially unsoldered, or that there's a 'micro leak' (a very
tiny hole in the solder, usually caused by soldering flux 'blowing
out' just at the moment when the solder is setting).
Checking for these problems is rather more difficult.
You can use an air test to check the crook. Place the palm of your
hand over the mouthpiece end of the crook so that the tube is sealed.
Stretch out a finger and slide it between the octave key pad and
the pip so that your finger is sealing the pip (and the key cup
is resting on top of your nail). Don't press down too hard, you
only want to seal the hole in the pip - if you press much harder
you finger might seal the outer circumference of the pip, or cause
a partially unsolder pip to seal temporarily.
Place the other end of the crook against your lips and blow hard
down the tube. You should feel a firm and consistent resistance.
Try sucking too - again, you should feel a firm and consistent vacuum.
If you can't maintain a pressure/vacuum then you've probably got
a leak - but unless you can positively nail the leak down to coming
from the pip you should be aware that it might not be the octave
pip that's leaking...it might be the crook tenon sleeve. Squirt
some washing up liquid into a cup of water, mix it up to a froth
and splash a little around the pip. Repeat the air test - if there's
a leak you should be able to see tiny bubbles forming.
Similar testing for the body octave key is all but impossible,
but there's still something you can do - and for this you'll need
a tin of cigarette lighter fluid (the stuff you fill petrol lighters
with, such as Zippos etc.). Before you get started, ensure the bore
of the horn is completely dry. You're going to be looking for sign
of moisture leaking into the bore, and the last thing you want is
any residual moisture from your last blowing session.
Place the horn down on a flat surface and position it so that the
body octave key pip points directly upwards (use cushions/cloths
etc. to prop up the horn as necessary). You now need to pop a drop
of lighter fluid around the base of the octave key pip - and the
best way to do this is to squirt a little of the fluid into a bottle
cap, then pick up a drop on a cocktail stick or cotton bud...or
even the tip of a pencil.
Once you've got everything ready, clean and completely dry your
Carefully apply the fluid to the base of the pip. Be very careful
not to get any fluid onto the octave key pad...where it may find
its way down into the tube. If you use too much fluid there's a
chance it might run around the body and find its way into the bore
via one of the palm key pads (it won't do any harm, it just knackers
the test). Now stick a finger down the bore of the horn and try
to feel around the base of the octave key tube - then quickly withdraw
your finger and see if there's any fluid on it. Give it a sniff
- if a tiny drop of lighter fluid made its way through it may have
already evaporated, but it will have left a distinctive aroma (this
is why you cleaned your hands). You can also make any lighter fluid
on your finger more visible by rubbing chalk on it - see the section
on other leaks
in crook for more details.
If you found nothing, repeat the test a couple of time just to be
on the safe side. If, however, you found some lighter fluid on your
finger, you've got a leaky pip.
You could, at a pinch, use other solvents - but cigarette lighter
fluid won't harm the lacquer/plating and won't damage any pads or
corks. It's also remarkably good at finding its way through the
tiniest of holes. It's also brilliant at cleaning sticky pads, so
it's well worth having a tin of it in the cupboard.
Assuming you've found a leak, what can you do about it?
Well, if a pip has come completely unsoldered you'll have to have
it soldered back in place.
If, for some reason, you need to be able to play the horn and you're
miles away from a repairer, your best bet will be to carefully prise
the pip up and wrap a small 'sausage' of adhesive tack around the
tube, before gently pushing the pip back down (really work the tack
between your fingers first, to get it nice and tacky). This'll get
you out of a tight spot for a while.
I wouldn't advise using anything more permanent - but (and this
is strictly for competent DIYers) I can't in all honesty see any
reason why an epoxy adhesive will not do a perfectly good job, provided
you're able to do a neat and tidy job. JB Weld in particular works
For micro leaks and partially unsoldered pips I'd obviously still
recommend having them resoldered - but some adhesive tack squished
around the base will offer a temporary fix. You could also push
a little wax (candle or beeswax) around the base of the pip, and
a hot needle will help the wax flow and adhere around the base.
Similarly, a drop of nail varnish (very carefully applied) will
get you out of trouble.
Other leaks in
Other leaks on body
Other leaks in crook
If you're reading this section it probably means you've read the
above section on unsoldered
octave key pips, and have found a leak on the crook - and now
you're not sure whether its the pip that's leaking...or the air
is escaping from somewhere else.
Assuming you've carried out the washing up liquid test and it's
shown no leaks around the pip, you can turn your attention to the
crook joint, which comprises the receiver or socket and the crook
Both of these parts are typically soft soldered on to the body and
crook respectively - and either because of a poor soldering job
during manufacture or subsequent deterioration of the solder due
to selective galvanic
corrosion, leaks may develop. In extreme cases the parts may
even fall off.
If your crook has developed a leak around the tenon sleeve then
it's a fair bet that the receiver joint may be in equally poor condition,
so it makes sense to test both parts given that the technique and
the tools are the same.
Even a very small leak in this joint can lead to instability in
note production - and it's nearly always the transition between
top G and A where the instability shows up most. This is precisely
why this kind of leak is often misdiagnosed as an octave mech problem.
The proper fix for this is to have the relevant part removed, cleaned
and resoldered - but if it's just a cheap horn or an old banger,
a drop of judiciously-applied superglue may do just as well. Keep
in mind, though, that a small gap in the soldered joint might be
only the tip of the iceberg, and the underlying problem may be much
worse...and you won't really know until the receiver falls off halfway
through a gig. On very cheap horns it's not uncommon to find (once
one of the parts has fallen off) that it was only held on by the
merest ring of solder around the top of the joint instead of being
start by looking for a leak in the receiver joint. The big problem
is finding the leak. You're unlikely to have much luck with shining
a torch at the joint and looking for light getting through - and
a far better test is to drizzle a little cigarette lighter fluid
(naphtha) over the joint and see where it leaks through. A piece
of coloured chalk will come in handy too.
There are two stages to this test, the first being the 'sniff' test.
Wipe out the receiver with some tissue to clean out any grease,
gunk and grime - and to make sure it's completely dry. Prep your
tin of lighter fluid by opening the nozzle. Now wash your hands,
taking care to rinse off as much of the scent of your soap as possible
and any contamination from opening the lighter fluid nozzle.
Sit yourself down and place the horn horizontally over your knee
in such a position that the slot beneath the crook clamp screw is
facing upwards. This prevents any excess fluid from spilling over
the surface of the receiver and seeping through the slot into the
bore...which would skew the test. Now tilt the receiver end down
Dribble a few drops of lighter fluid onto the joint, being careful
not to splash it onto the receiver itself - and being even more
careful not to get any on your hands. It will run around the joint
Now stick a finger inside the receiver, feel for the joint and quickly
run your finger around it...withdraw it and give it a good sniff.
If there's a leak in the joint, lighter fluid will have made its
way through and you'll smell it on your finger. If the leak is large
enough you may even see some fluid on your finger...but a small
leak will allow so little fluid through that it will evaporate on
your finger almost immediately. But it will leave an aroma behind
for a bit longer...hence the sniff test.
what's the chalk for? Well, chalk will readily soak up lighter fluid
- so if you coat your fingertip with chalk it'll show any fluid
contamination far more clearly than on bare skin...and the chalk
will slow the evaporation process. I find blue chalk seems to work
best, and you can clearly see the difference between the part of
my finger that's dry and the darker part that's contaminated with
the lighter fluid.
If the test proves positive you'll now have to locate the leak.
Dry the socket out again, and wipe over the exterior. It'll need
to be completely dry - so either leave it a while or play a hair
dryer over it for a minute or so.
repeat the drizzling...but this time you only want a drop or so
- just enough to seep through the leak and not to flood the inside
of the receiver. Shine a light into the receiver and you should
be able to spot a dark patch where the fluid has seeped through.
You can make use of the chalk again though - but this time you need
to break a bit of and pummel it into dust. Dip your fingertip into
it then rub it against the joint at the bottom of the receiver.
Repeat until you've got a line of chalk all the way around the joint.
Apply the fluid to the exterior, and any that seeps through will
show up clearly in the area where the chalk has soaked it up.
If you now inspect the exterior of the joint there's a very good
chance that you'll see either a break in the solder joint around
the receiver or a suspiciously dark area that looks like it might
be a gap in the joint. Don't forget to clean the chalk up when you're
finished - it should brush/wipe away easily.
Temporary fixes? Well, as just mentioned, a drop of superglue will
seal the hole - but it can be a messy fix. If it's a nice horn,
a piece of tape will serve as a short-term fix until you have have
it repaired properly, as will rubbing some candle wax or beeswax
into the hole.
Such leaks are, fortunately, quite rare - but there's a far more
common source of leaks from the receiver...namely a loose crook.
A loose crook or neck can cause all manner of problems. The most
common problem is a loss of tone - the horn seems less vibrant,
less clear and stuffier to blow. Generally speaking, when the crook
locking screw is tightened up hand tight (you should never have
to force this screw) the crook should be held quite snugly and barely
move (if at all).
I tend to advocate a little bit of slippage - if you happen to have
the horn knocked while you're playing it, a bit of give on the crook
might well save your teeth - but it should be noted that this movement
must only be confined to the side-to-side plane. If you're able
to rock the crook up and down then you really do have a problem.
cause of loose crooks is usually down to the tenon socket stretching
with age or use - but excessive tightening of the crook screw in
order to overcome a bit of looseness will eventually make things
worse. It leads to the receiver becoming tapered (tighter at the
top than it is at the bottom), and air can rise up through the gap
at the bottom of the joint and exit at the screw slot. In more severe
cases the metal around the bottom of the slot may crack.
This is very bad news indeed - but as with most problems on a horn
it's fixable...if you throw enough money at it. The cheapest fix
is to have a patch soldered over the crack. The next cheapest fix
is to have the receiver removed and the crack filled with silver
(hard) solder - and the most costly option is to have the receiver
replaced...which may required building a new one.
If you spot a crack early enough it's possible to stop it in its
tracks by eliminating the source of the stress that's causing it,
and by either elongating the slot into the crack or drilling a small
round hole at its base to help prevent the crack from propagating
Another common cause loose crooks is a stiff crook screw. This
mechanism often gets wet - and when water mixes with brass it often
lead to the formation of a blueish/white deposit. This stuff is
quite gritty, and it'll hamper the operation of the screw. The fix
is easy - simply remove the screw, clean it (and the sockets) with
a pipe cleaner soaked in a little lighter fluid, then pop a drop
of grease or oil on the screw before refitting it. Tighten the screw
up without the crook in place (not something I generally recommend),
and when it starts to lock up give it just a half turn more (but
don't force it)...then back it off. Fit the crook and tighten up
An easy way to test if your receiver is causing a tonal dropoff
is to smear the crook tenon sleeve with a generous dollop of cork
grease (Vaseline will do in this instance). Fit the crook and work
the grease in by rotating the crook back and forth. Tighten up the
crook screw (don't overtighten though - with the tenon greased it's
unlikely that the clamp will be able to lock the crook tight) and
give the horn a blow. The grease acts as a temporary seal, and if
you notice a significant improvement in tone or stability then you'll
need to have the crook joint properly tightened. Remember to wipe
the grease off afterwards.
The proper fix is to have the crook tenon sleeve expanded and
refitted to the socket, which is a time-consuming and messy job
(the tenon sleeve may have to be lapped to fit the receiver), so
it tends not to be cheap. It is, however, the sort of job you might
only need once or twice in the life of a horn...and the resultant
improvement in tone and stability more than makes up for the cost.
A cheaper fix (AKA a bodge) is to grind a little of the screw socket
cheeks away so that the screw can pinch up more metal. It often
works, but it places yet more strain on the screw slot and increases
the chances of the receiver cracking.
test the tenon sleeve joint you can use the technique described
above - hold the crook so the tenon sleeve is only very slightly
pointing down from horizontal, apply a few drops of lighter fluid
to the joint then use your finger inside the bore to catch any fluid
that might have leaked through.
Alternatively, because you're able to seal one end of the tube off
and blow into the other you can also use the suds test as described
in the section on unsoldered
octave key pips.
As to fixes - well, as already mentioned you can seal up a small
leak with a drop of superglue...though I must reiterate that unless
you know for sure that the problem is a small leak and not a major
breakdown of the solder (or indeed a lack of it), you could find
that the joint fails while you're playing the horn and the crook
hits the deck. This would be extremely bad news for your wallet.
Another temporary (and potentially less messy) fix is to wrap some
plumber's tape (PTFE tape) around the outside of the joint. You
can even melt a little candle wax around the joint with the aid
of a hairdryer (or a small gas gun, if you're careful).
Leaks from elsewhere on the crook are very uncommon, but not unknown.
Major culprits are leaks due to cracked tubing (typically on older
horns that have had numerous repairs made to dented/crushed crooks,
leaks from corroded tube seams (typically on the underside of the
tube, and nearly always just after the crook cork - and leaks via
the cork itself, due to air gaps beneath the cork...so that air
blown in through the mouthpiece can slip under the cork and escape
out the other end. Each of these leaks will require fixing properly,
but plumber's tape or adhesive tack will provide temporary solutions.
If you own a horn with a double-sleeved tenon or a tuning attachment
(Microtuner), it adds
a whole 'nother layer of complexity to the affair. Both of these
require specialist knowledge to deal with them appropriately - particularly
when it comes to diagnosing problems and deciding on solutions.
You can certainly use some of the techniques described here (with
some adaptation) to test them, but I really wouldn't advise attempting
any home fixes.
mouthpiece cork can be an occasional source of problems.
Its job is to hold the mouthpiece firmly in the desired position
and to seal against any air that wants to escape (rather than go
down the bore of the horn). The cork should be in good condition,
and supple enough to still hold the mouthpiece securely when moved
a moderate distance from the ideal position (typically three-quarters
of the way onto the cork - see embouchure/mouthpiece
issues). If the cork is chipped, cracked or not very tolerant
of different mouthpiece positions, it's probably time to have it
replaced. It's a beer money job - and it'll save you the indignity
of having to use things like plumber's tape (as seen on the cork
above right) to hide the shame of a knackered and worn-out cork.
The quality of the cork may have bearing on the efficacy
of the seal; if it's particularly grainy or pitted, there's a chance
(admittedly small) that air may 'track' its way through these imperfections
and leak out of the rear of the cork.
A lot of cheap cork sheet can be shot full of holes too - and while
these probably don't present much of a risk of leakage in themselves,
they nonetheless allow moisture to get beneath the cork...and over
time that may lead to failure of the adhesive
way air can track through a mouthpiece cork is by going underneath
it. It's exceptionally rare, but it does happen. It's due the cork
being poorly glued to the crook. Air enters at the tip of the crook,
channels beneath the cork and exits at the other end. It's quite
hard to test for, but it you pop your mouthpiece on, brush some
suds over the end of the piece and the cork and then blow the horn
while someone else watches the end of the cork, it may be possible
to see bubbles forming.
And, finally, there's the wonderfully-named butt joint leak.
This is down to how the mouthpiece cork was fitted. Standard practice
is to use a lap joint (left, figure A), which increases the surface
area where the two end of the cork sheet meet and are glued together
- but some repairers (and manufacturers) use a butt joint (left,
figure B). There's nothing wrong with a butt joint, but great care
must be taken to ensure that the joint is fully closed all the way
along its length.
Figure C shows what happens when a butt joint is incorrectly formed.
It's rather exaggerated but you can see that because the ends of
the cork sheet aren't square, gaps have formed above and below the
have to be a particularly half-arsed job to leave the cork like
this, and it's far more common to find either a leak above or below
the joint rather than both - and you can't rely on the glue to fill
such a large gap.
How can you tell which joint has been used?
Have a look around the surface of the cork. You're looking for a
line that extends from the front of the cork (at the tip of the
crook) to the rear. If a lap joint has been used you'll typically
see a wavy line - and a butt joint will show a more or less straight
An even better indication is to look at the rear face of the cork
(you might need the assistance of a magnifying glass).
lap joint, as shown here, presents a clearly sloped line. This is
quite a generous lap, and you may well see a far shorter line at
a much steeper angle. If the angle gets too steep the joint becomes
less of a lap joint and more of an angled butt joint - and a proper
butt joint will show a vertical line.
If you spot any gaps it could well mean the cork is leaking, or
is likely to leak at some point, so it would make sense to have
For a quick fix, however, you could melt a little wax into the gaps.
I'm not going to recommend attempting to glue up any gaps at the
rear of the cork on the basis that it could get very messy and damage
the finish on the crook - but gaps at the front end are a very different
Not only can they let air under the cork, they'll let moisture in
too - and that can lead to lots of problems over a period of time.
You could just fill the gap with some wax, but as this part of the
cork takes a lot of punishment it's unlikely to last very long.
I also doubt you'll have much luck with conventional glues because
there's likely to be a lot of grime tucked inside the gap (cork
is really the only way to go, and I very strongly recommend
one of the gel types. You can give yourself a fighting chance by
cleaning the end of the cork with a drop of lighter fluid on a soft
brush. Be gentle or you might simply rip the cork off, and give
it ten minutes or so to dry out. Now very carefully wipe a little
cork grease over the surface of the front end of the cork. Wipe
it on gently, don't force it into the cork. This will help prevent
any glue that oozes out from sticking to your fingers, and lessen
the chances of your ripping the cork away.
need something to poke the glue in under the cork. A pin will do,
but a better bet is a cocktail stick with the pointed end cut off
and the end of the stick shaved to a flat to form a small spatula.
Pop a blob of glue on the end of it and tease it into the gap. Work
it around so that you cover as much of the surface area beneath
the cork as possible. Gel superglue tends to set less quickly than
the fluid kind, so you'll have a little working time. Once you've
got some glue in the gap, press the cork firmly down to close the
gap up. Hold it down for a minute or so (or longer, if you have
the patience). Repeat as necessary for as many gaps as you have.
It's unlikely to matter that much if you make a bit of a mess at
this end of the cork...it usually looks a bit tatty anyway.
If you're tempted to use this fix at the other end of the cork,
do yourself (and your horn) a favour and wrap a bit of masking tape
around where the end of the cork sits on the crook tube. As an extra
belt-and-braces measure you can pick up a tube of superglue remover
when you're buying the superglue. It's handy stuff, works a treat,
and a tube of it will last you for years.
Octave key pip unsoldered
Well, I'm guessing this is not the place you wanted to be - but
take heart, half the battle of sorting out a problem is knowing
what the problem is...and as it looks like you're the problem, there
are things you can do to improve your playing skills and develop
your embouchure (pronounced 'ombowshure').
There is one 'getout' though, and that's your mouthpiece. If you're
playing on a duff piece you're likely to run into all sorts of unexpected
problems - many, if not all, of which can be solved by buying
a better mouthpiece - though it's always worth bearing in mind
that just because you paid a lot of money for a mouthpiece, it doesn't
mean it really suits you.
Reeds can be a factor too. A bad reed can cause a lot of problems.
Changing it is a good bet, but you can't assume that the next reed
out of the box won't also be bad. Try a different strength; too
soft or hard a reed can lead to unpredictable results.
matching may be a problem. This is usually more of an issue on vintage
horns. Putting a bright, high-baffled modern piece on a horn that
was built in the 1930s is asking for trouble. It's not that it can't
be done - it's just that it takes a lot of practice to make it work.
It's also the case that 'extreme' pieces (very bright or very dark)
can be quite challenging even on a modern horn, and I'm often surprised
at how many new players have been suckered in to buying entirely
Mouthpiece position is critical. As a very general rule of thumb,
a mouthpiece should be pushed about three quarters of the way onto
the mouthpiece cork. This is the 'sweet spot' - it's where the good
people who designed your horn expect your mouthpiece to be so that
(in very simplistic terms) all the acoustics line up. If
you stray out of this area by more than about 5mm either way, the
horn becomes more and more unstable in terms of internal tuning
(the tuning between individual notes) note production. The red line
in the photo shows where the end of your mouthpiece should sit -
at least as rough starting point.
The smaller the horn, the more critical the placement of the mouthpiece
And then there's you - or rather your embouchure.
You can give someone the most perfectly-built horn and the most
easy-blowing mouthpiece fitted with the best reed in the world...but
it still won't guarantee that they'll be able to play a clean note.
The sax is fairly forgiving, but there are still some hurdles to
overcome. Middle D is a notoriously weak note, and the break-point
between octave G (the highest note that uses the body octave key
pip) and octave A (the lowest note that uses the crook key pip)
can be difficult to make cleanly.
Lots of long note practice will help - it's the sax player's equivalent
of weight training. It strengthens and tones the muscles and gives
you more control over the notes. There are plenty of websites out
there with all sorts of hints, tips and advice about technique -
and one I particularly recommend is Pete
Problem with additional links to octave
On most horns the octave mech concerns itself only with opening
the body octave key (of which there may sometimes be a pair) and
the crook octave key.
However, some horns have additional keys which link to the mech,
of which the C# link is the most common. This link closes the auxiliary
B key when an octave C# is played, and is used to tame a tendency
for this note to play quite sharp. It's very often found on sopranos
and one or two altos but is rarer on the larger horns. It's often
thought to be a modern invention, but in fact it can be found on
a number of vintage horns (there's seldom anything new - just variations
on a theme or 'reinventions' of long-forgotten patents). As you
might imagine, such additional linkages add another layer of complexity
to the octave mech and can throw your troubleshooting efforts into
disarray if you don't have a clear understanding of what these links
do and how they do it.
a C# link on a soprano - I've removed a few keys so that we can
get a better look at it.
There's an additional key arm attached to the lower part of the
swivel mech. When the octave key is pressed down, this arm will
rise - and when the octave key is released it will fall. As there's
no official name for this arm we'll simply call it the primary C#
arm. It sits beneath another arm coming off the rear of the Auxiliary
B key - which we'll call the secondary C# arm. As the primary arm
rises it lifts the secondary arm - which closes the Aux.B pad (seen
on the lower right).
Note the hole in the centre of the pad. The Aux.B key is sometimes
'split' into two keys - a lower key cup with a doughnut-shaped pad
in it and an upper key with a very small pad which closes down over
the small centre hole in the lower pad. It's only this lower part
of the Aux.B key that is moved by the C# link.
There are three ways in which this link can cause problems, the
first of which is that the primary arm doesn't rise far enough to
close the Aux.B pad.
This isn't a major problem insomuch as it will only affect the clarity
and tuning of the octave C# - and not the functionality of the octave
The second mode of failure is that the primary arm is unable to
rise far enough because it has already closed the Aux.B pad before
the octave key has been fully depressed. In effect this limits the
throw of the mech, which may in turn mean that neither the body
or crook key cups are able to rise fully. This can lead to hissing
and instability. The technical terms for this is 'holding off',
which describes how one key prevents another from fully opening
And the third mode of failure is that the primary C# arm doesn't
fall far enough to allow the Aux.B pad to open fully. On its own
it doesn't affect the operation of the octave mech - but it will
throw out the regulation of the horn's upper stack (the A, B and
C keys), which leads to a clunky feel and may also affect the tuning
of the lower C#.
However, it's often the case that this mode of failure is responsible
for the second mode (and vice versa).
There are a few ways you can check the mechanism. The first is
to check the operation of the thumb key, and that it can be pressed
all the way down.
If you look on the underside of the key you'll see one or two buffers.
There's often one directly below the touchpiece - and another one
may be located beneath the forked arm at the other end of the key
(you can see it in the shot above, immediately below the red shaded
primary arm). You should have at least one of these buffers, and
possibly both. Press the touchpiece down and check whether any of
the buffers touch the body tube. If they don't it means that something
is preventing them from doing so - and it may well be because the
primary arm is being held off...though it may well also be because
of other problems within the octave mech (told you it was complicated).
Take a peek at the Aux.B key pad - it should be fully closed.
all is well, release the octave key - and now (very) gently lift
the touchpiece up. Not by much...it only needs to move a millimetre
or so. Look at the point where the primary C# arm meets the secondary.
As you lift the octave key, the primary arm should fall...but the
secondary arm should stay put. This will show up as a gap appearing
between the two arms. It won't be much (hopefully), but even the
smallest gap tells you that the upper stack action is correctly
balanced with the octave mech.
If no gap appears it'll mean the Aux.B key is being held off by
the primary arm and you'll have a spot of double-action in the upper
finger a G. The secondary C# arm will rise. Place a cigarette paper
feeler under it, then press the octave key down. As the octave key
closes, the primary arm should pinch the feeler paper. Look again
at those buffers beneath the octave key - they should be touching
the body. When this linkage is set right you should be able to press
the octave key down very, very lightly and just about be able to
pull the feeler out. As you press the octave key down a little more
firmly you should feel the grip on the feeler increasing. If the
feeler is fully gripped before the octave key touches the body it
means it's being held off...and you may not be getting full opening
on the body or crook pads.
If you find any problems I'm going to recommend that you avoid
trying to fix them, simply because it's likely to get very complicated
very quickly. This is especially true of cheap sopranos (where this
mech is commonly found) because there are other factors in play
- such as the mechanism being poorly built in the first place, and
the possibility that it was never set up to work properly from new.
More often than not such mechs have to be set up using the 'That'll
have to do' principle, simply because it's the only way to get the
damned things to work.
the only thing that's relatively easy to fix is that of the primary
arm not rising high enough to close the Aux.B pad - and in most
cases this'll be down to a need for a thicker buffer between the
primary and secondary arms. You can see on the examples above that
the buffer is a clear plastic tube...and it's not uncommon for these
to fall off. On some horns there's an adjustable pin on either the
primary or secondary arms. By moving this pin up or down the arm
you can adjust the relationship between the two arms using the checks
as described above to find the 'sweet spot'. However - on cheap
horns this adjuster is usually very poorly made...and even more
poorly fitted - and attempting to move it often results in something
breaking (usually the screw that holds the adjuster on) or the fork
arms splaying out when you try to tighten the screw up.
The best that you can do is check that the screw isn't loose (just
give it a bit of a pinch up with a screwdriver...and I mean just
a bit) and otherwise leave well alone.
On horns with even more complex additional links (such as altissimo
octave key cups etc.) the recommendation to leave well alone and
let a professional deal with it is even stronger.
Body/crook key not rising
Other leaks on body
Well, the good news is that your octave key mech seems to be working.
The bad news is that your problems are more likely to be down to
leaks elsewhere on your sax, which are really beyond the focus of
this article. Such leaks often present themselves as octave key
problems because they're more noticeable in certain group of notes.
You might, say, have no problems getting a low G - but when you
try for an octave G, the note fails (or vice versa)...and it's therefore
easy to deduce that the octave mech is at fault.
And it might be - but it might not be the only problem. Leaks
tend to be either catastrophic (such as a pad falling out, stopping
the horn dead in its tracks) or cumulative (the horn gets less responsive
the lower you play).
Your dodgy octave G might be down to a small leak on, say, the A
key pad - and while on its own this might not have much effect on
your octave G, another small leak on your octave key mech may well
be all that's needed to push the note into instability. Fix either
one of the problems and the G works again. Fixing both problems,
of course, would be the best bet.
Assuming you've carried out the tests for checking for structural
leaks on the crook, the next step is to check the pads from
top to bottom - and details of how do this can be found here.
A noisy mech is usually just a nuisance rather than a show-stopper,
but it's often a sign that there's a problem. Most of the noise
will be down to metal hitting metal - and this'll either be because
a buffer has worn or fallen off or there's simply no lubrication
on the mech.
Checking for a missing buffer is reasonably easy - just work the
mech slowly, noting where each part rises and falls, and look at
the various point where either one key contacts another, or where
part of a key contacts the body. In almost every case there should
be some means of preventing one piece of metal colliding with another.
Inspect thin buffers carefully as it's quite possible that they
may have worn through and there's now a 'bald patch' that's allowing
metal-to-metal contact. If in doubt, stick a piece of paper or masking
tape under the key to see whether it stops the noise.
Some care is required when dealing with vintage mechs because certain
keys may look as though they ought to have some buffering fitted
to them when in fact they rely on a certain amount of clearance.
As a very general rule-of-thumb, the places where keys connect to
each other will need to be buffered - but where keys look as though
they may contact the body tube there may be a need for a certain
amount of clearance beneath them, and any buffering present may
only be there to prevent clanking on the odd occasion that the key
bounces. If in any doubt, use the paper/masking tape trick to see
how it affects the functionality of the mech.
If all looks to be in order then it's a fair bet that all your
mech needs is some lubrication, and for the most part this is going
to be key oil. For an in-depth overview of oiling the keys, check
out the article on Oiling the
said, you might spot some areas where there's obviously metal-to-metal
contact - but no practical means of fitting a buffer between the
parts. A very good example of this is the swivel pin used on Selmer
style mechs. Unless the swivel has been factory built with buffers
(usually small tubes) fitted or subsequently modded to accommodate
them, there will be metal-to-metal contact at each end of the swivel
bar and in the centre, where the bar pivots. You can certainly pop
a drop of oil on these contact points, and it'll go quite some way
to quietening things down - but the swivel ends will do better with
a drop of grease.
Just about any grease will do, but the best bet is thick silicone
grease. This will tend to stay put rather longer than standard grease.
Pivot and Roller Lubricant' is excellent for the job, and I
find that bog-standard High Tack (or HT) silicone grease will do
nicely too and is readily available (plumbing stores, hobby shops,
eBay etc.). High tack simply means it stays in place rather better
than standard silicone grease, which tends to slowly dribble away
from the joint.
The hardest part is getting the stuff in without dismantling the
mech (and trust me on this, you really don't want to do that if
you can help it) - but if you pop a blob near the offending joint
and use a finger or an old reed to force it into the gaps, it should
work just fine. Just make sure you wipe off any excess afterwards.
Another source of noise could be hardened buffers. Cork can go
quite hard over the years, and felt can absorb moisture...which
dries out and hardens it over time. Plastic tubes are notoriously
noisy from the off, but even these can harden eventually. In such
cases the only remedy is to have them replaced.
for 'other' - as you can see by the sheer size of this article there's
an awful lot that can go wrong with an octave mech, and no matter
how many issues I've covered there's always room for more. For example
- is there a dent near the mech that might have pulled one of the
pillars out of alignment? Has a bit of superglue (commonly used
on cheap horns to disguise excess play in the keywork) got into
the mechanism? Have you recently lost an earring? Maybe it fell
into the mech and jammed it up (seen that one).
With a bit of luck, this article will cover around 90% of the problems
you're likely to encounter - and as I said at the start, even if
you can't work out what's wrong, you can at least narrow the field
down a little...and if all else fails you can give your repairer
All of this assumes, of course, that your diagnosis of the octave
key mech was up to scratch. If you've read through all the relevant
sections and done all the checks then it's fair to say you've given
it your best shot - but that doesn't mean you didn't miss something.
It happens, and because octave key mechs can be notoriously fickle,
there's no shame in it. If in doubt, get a repairer to check it
out - or get a more experienced player to blow your horn to see
if they have the same problems.