Wednesday, May 7, 2014

Finding the Leak When a Leak Light Doesn't Find It

The leak light has become the standard when installing new pads on a saxophone or when trying to chase down an issue that is usually caused by a leaking pad (stuffiness or warbling of lower notes).   Intonation isn't generally thought of as a leak issue, but it very often is.  I don't want to get side tracked (like that won't happen), but if you play low D and then use different palm keys to vent instead of depressing the octave lever and opening the one of the octave pips, you can get a variety of notes, usually off pitch.  So small leaks can also do odd things to intonation, and sometimes odd things that are only apparent with one particular note.

That's how I got into checking for the few things that I'm going to discuss.  After rebuilding a horn, sometimes a horn with a great reputation for intonation, I found odd things that I couldn't correct by running a leak light through the horn.
There are areas where the leak light just can't do it, yet it is very important to get the horn leak proof for intonation and power.  Here are a few things I've learned. 


Starting at the top, take the neck off and test the upper octave pip.  I place the tenon of the neck against the heel of my thumb like this.  Sometimes, I even lick the area to make sure that I'm getting a good seal.  

Then, I suck on the cork end to create a vacuum.  Once a vacuum is formed, I put my tongue over the end of the neck tube.  (Don't create too much of a vacuum or you can actually hurt your tongue if you have a sharp edge.)  This checks the seal on the upper pip.  The vacuum should hold indefinately, even with a microtuner.  If the vacuum fails after several seconds, you've got a problem.  I have become a fan of using cork for the pip pads.  Cheap, fast, moisture proof, long lasting, and seals just as well if not better than leather pads.


Next thing to check is the neck tenon seal.  This can be difficult to test, however, if you suspect that there may be an issue, you can "fix" it and see if it makes a difference in how the horn plays.  The first fix is to wrap a single layer of PTFE tape (also known as Teflon plumbing tape) around the neck tenon and insert it into the horn using a rotating motion in the opposite direction from your wrapping.  If the fit is tight, the tape will "smear" and it will look something like this.


That picture is of my Martin, and I just leave the tape there.  It stays there for several months and then I do it again.  On a Conn with a double neck socket, I rolled the PTFE tape into a string and inserted it like a gasket to be compressed when the neck was put on.  Worked like a charm.  If the improvement is really apparent when you play, you might want to actually do some tenon work or take it to a technician.  If it gives the horn a little boost and more even intonation, then I'm good with that.  It's not like having PTFE tape showing on your mouthpiece cork.  That always reminds me of the geek who fixes his glasses frames with BandAids.




Another neck area that can be an issue is the expansion slot on horns where the neck is attached with a screw clamp.  The slot has to extend below the tension screw used to clamp the neck, and that can be the source of a tiny leak.  PTFE tape to the rescue again.  The neck receiver slot on this Beaugnier has been "packed" with white PTFE tape (shown here as a white stripe).  When the neck is clamped down, the PTFE works as a packing gland and assures an airtight seal at the slot.




But aren't we are fixing things on the neck that might not even need fixing?  Okay.  But since the fixes are so easy, I'm not too worried.  

There are ways to find out if the neck connection is actually leaking.  Here's one.  


It looks like a creepy medical device, but it is actually a plumbing tool called an inflatable test bladder.  So it even sounds like a creepy medical device.

Air is pumped into the tire valve and the other end inflates inside the pipe, or in our case, the horn, to make a tight seal.  
In order to put the device into place on altos and tenors, it is usually necessary to remove one of the upper stack pads, thread the device into position through the bell, and then pull the tire valve out of the upper stack key hole.  The black rubber part (the inflatable bladder) is placed right below the neck tenon and then inflated.  
Now the neck is put on, and since we have already done a vacuum check on the neck octave pip and microtuner, we know that when we suck on the neck any leak is because of the neck tenon assembly.  You can see now why "fixing" a possible tenon leak before a full diagnosis is a viable option.  The diagnosis is time consuming and tedious, while the cure is quick and easy.  

Which leads me at last to my strange testing method (and the purpose of this whole blog).  Wouldn't it be nice if we could just vacuum test the entire horn to find out if it is leaking?  Not just pip and neck tenon leaks (which don't show with a regular leak light), but also pad leaks in those few places where we can't get a clear view with the leak light.  There might even be other places where leaks can occur that we haven't even thought of.  

Think of the bandwidth this would save!  Go to any saxophone website and read the posts.  "My horn plays sharp.  What could it be?"  "My horn has a stuffy Eb.  What should I do?"  "My horn jumps registers.  Who knows how to fix it?"  What could it be, what should I do, who can diagnose my horn sight unseen over the internet?  12% of all posts start out similarly (I just made up that number).  Well, the first thing that you should do is to make sure that your horn is sealing properly everywhere.  Once you have done that, then you can post a "what should I do" question.

Here's what I've come up with.  Place a piece of Saran Wrap, aka plastic wrap, over the bell and fasten down with a rubber band to keep it in place.  



Put the neck on the horn, finger low Bb, and suck in.  Cheap plastic wrap simply pops from the vacuum, assuming the horn is fairly tight.  Good plastic wrap is a bit more difficult, and that's why I would recommend using the best you can find or a piece of plastic grocery bag. 

Here's a picture of the wrap sucked into the bell a little bit and indicating a good seal around the bell.  I place my tongue over the end of the neck, as in the prior test.  I consider holding a vacuum for 5 seconds or more to be great.  That doesn't usually happen on the first test.


 
What we are after isn't simply the ability to pop the wrapping.  We are looking over the entire horn for leaks.  Actually, we are "hearing" over the entire horn.  

As you inhale slowly, you will get an indication as to how leak proof your horn is.  If you inhale fast, you can suck the pads down that wouldn't normally seal when inhaling slowly.  You may hear air hissing into the horn from some leak.  On a fairly tight horn, you should be able to raise one finger at a time off of the keys and the vacuum will hold the key down (unless you have your pads poorly floated or your keys over sprung, IMHO).  A key that lifts easy under vacuum (i.e., while you are sucking) is a suspect.  

If a hissing sound continues despite using a gorilla grip on the keys, you know that the leak is either from one of the "un-fingered" normally closed pads that's not sealing properly or some place other than the pads.

So, let's say that you've done the test and, although you can pop the plastic if you wish, when you inhale slowly there is a slight hissing sound and the horn doesn't hold a vacuum.  How do you locate that hiss or those hisses?  The first way is using an assistant.  Nah, that's no fun.  What I use is a desk top microphone, headphones, and my computer.  Here's the microphone "sniffing" for a leak on the low Bb pad as I slowly inhale.
 

Actually, I thought that my leak sounded like it was further up on the horn, so I started from top to bottom.  After checking the neck tenon, I quickly found the culprit.  I could hear the little C# pad "whistle" from the vacuum when the rest of the upper stack keys were pressed down, even when using a gorilla grip.  With the microphone/headphone setup, the hissing was much louder than what I was originally hearing.  It sounded like a tornado.  A real bonus.
It wasn't a problem with a defective pad or an uneven tone hole lip, it simply wasn't regulated correctly with C, B, and A, which should depress and seal the C# pad when the other notes are fingered.  But I couldn't see that with the leak light, maybe because the palm keys are in the way on this horn.  Or it could be that the pad was pressed tight enough so that light wouldn't show, because of the rim edge "footprint" on the leather, but air could get through.  And ultimately it is air leakage, not light, which is our concern.  

I placed paper on to the cork that regulated the C# key, and that fixed the leak.  For this type of regulating, I initially use yellow Post-It notes (a lightly adhesive paper).  Post-Its stay in place, can be layered for any thickness, and can then be easily removed when the cork is redone.  The thickness of the paper is about .003 inch.  Two layers (.006" or .15mm) was enough to close the key and stop the hissing leak that I was hearing.  Here's a picture with the slips of paper in place.


Two good things came from this.  The horn speaks better.  It just seems to have a little more power.  And even better, this was a horn that suffered from the infamous "stuffy second D," as mentioned at the start of this blog.  2nd D was both stuffy and sharp, although it was possible to play 2nd D in tune with enough aforethought (memo to self: remember to lip down second D every bloody time).  The "stuffiness" could probably be more accurately described as a "weak and unpredictable" 2nd D.  What a pain.  

The cause of the problem was that I basically had two vent points open when playing 2nd D.  The lower octave pip would be open, and right below it, the C# key was cracked open .006", creating a "ghost vent."  The C# key is close to, but not quite, a natural venting point for 2nd D.  The natural venting point is actually the palm D key.  Play low D and hit palm D instead of the octave register key.  You'll generally get a nice 2nd D even with the pip closed (which is a traditional remedy for a stuffy D).  But pressing the octave register and having C# leaking caused my horn to have a sharp and stuffy 2nd D (and probably other lesser intonation problems throughout both registers).  

Even though the C# leak was open and affecting every note below C# in both registers, it made itself most noticeable as a faulty ghost vent point for 2nd D.  A tiny undetected leak on a horn might be barely noticeable, or might make one certain note a pain in the butt.

But back to the "sucking" method of leak detection.  There's nothing besides the pads on the lower horn that can leak, right?  Wrong.  Old Martin horns can leak around the tone holes if the solder corrodes.  I also did a "fix without diagnosis" when I rebuilt my Martin.  I wiped tung oil on the inside lip of the tone holes.  If there was any pitting, crevices, corrosion, etc., the tung oil would plug the pitting and seal off the solder from further corrosive liquids.  Keep in mind that tung oil takes several weeks to polmerize.  Not really a problem unless you leave a fuzzy pad guard in during this time.

Although I haven't actually come across leaking soldered tone holes, I have found a leak around an extruded tone hole.  I couldn't believe it when I found one.  As tone holes are extruded, the thin brass of the body tube is forced to make a right angle bend.  This can cause a stress crack in the metal that is basically invisible to the naked eye and, like the solder corrosion, may not be in a "straight line" through the horn such that light can emit from the leak.  

Most of these flaws are probably sealed with lacquer during manufacturing and have no real effect, but this one wasn't.  After finding it, I was able to put a leak light to this stress fracture and get a picture of it.  This is the bottom of the Eb tone hole chimney, looking up from the bow.  Click on the picture and it's right in the center.


Instead of sealing the leak with tung oil, I chose to "tin" the area with solder from the inside.  Not an easy task, but the low notes on this tenor now speak with authority.  It also helped with a slightly sharp low Bb.  Again, a double bonus.  I feel sorry for whoever played this horn from 1957 until I bought and rebuilt it.  They never experienced the full extent of this Dolnet's low end power.

There are other non-leak light methods of finding leaks, but I think this method is the best.  You can instantly check the whole horn.  You can check for suspected leak problems when on a gig.  No electricity required.    By applying varying pressure on keys, it's often even possible to quickly identify the problem pad.

Here's another leak that I found that would have never been detected using a leak light.