Reed voicing / profiling / tuning etc.

[4to5to6]

Searching through the forum post, I have found a lot of discussion on reed voicing, tongue profiles, etc. but nothing with all the factors combined..,

 

I need help...

 

After two days of fiddling around with an instrument, it is much, much better but still not acceptable.  I have concluded that it was tuned to concert pitch probably by using a Dremel.  Ouch!  It has been butchered but I am determined to make it live again.  The high reeds are especially being a total pain.  They are weak and unresponsive and take a lot of pressure to sound. The upper octave is in reality still unplayable.  I've been trying a lot of different things.  What was done to these reeds?!?!

 

I have been able to voice many of the reeds by bending the tongues (many were curved) to shape and adjusting the gaps but had to reprofile many of the reeds by filing and also had to completely replace at least six tongues.  A couple of reeds had brass tongues so I replaced these.

 

The lower octave are sounding great now but the high notes are still being a real pain!

 

There seems to be a delicate balance between reed thickness, reed profile, reed gap, belly (center of reed) thickness, etc. as well as the weight of the valve, etc.  Many of the high notes are very weak and thin sounding and lack in responsiveness.  I've made many adjustments but just can't seem to get them to be responsive with a decent volume.  I've been experimenting with different reed parameters:

 

- profile of tongue

- thickness of belly (stiffness of reed / 

- curve of reed

- gap set between reed and frame

- ratio between root / tip tongue thickness high/low reeds

- weight or thickness of valve.

 

I remember Steve Dickinson going into great deal with me about setting the air flows by adjusting the pad hole sizes to match the requirements of the reeds, the superior responsiveness of long scale reeds, over blowing reeds, acoustic dampening, resonant chambers, slot relief, mass / tensile balance, original/best/long scale, Bessemer process, key travel vs. key height, etc.  What did I miss? but I've never had the opportunity to learn these skills hands on.

 

Has anyone ever summed all this up in one place?  It would be very helpful.  How about a video of someone voicing a butchered reed... a high reed and a low reed?  How about going through the process of lowering a reed from old high pitch to A440 and then reprofiling the reed.  The metal seems to bend down at the place where it is filed.

 

What am I missing?   I need a bit of advice especially in getting these high reeds right...  Many are weak and unresponsive and take a lot of pressure to sound.  I've tried changing the curve, adjusting the gap, adjusting the profiling.. while better, still not there.  Comparing to other working reeds is not immediately obvious.  What am I missing?  What was done to this instrument?!?! 

 

[Matthew Heumann]

With regard to someone butchering it with a Dremel: " A good craftsman never blames his tools".

I've been tuning concertinas for 30 years with a Dremel using a rubber bit infused with micro abrasive and have NEVER had a bad result. Quite the contrary, the rubberized bit, using gentle brief quick strokes, does not over-heat the reed, and actually buffs away the metal without leaving the abrasion grooves of files or automotive sandpaper.

Be it a file, sandpaper, or Dremel, an unskilled or careless user will damage a reed.

 

That being said, I have also encountered individual reeds that defy voicing, am actually working on a lower reed that refuses to sound "nice". Profile is good, position in frame is good, fit in reed slot is good, yet the damn thing is harsh. So I understand the frustration.

 

When you come to the Northeast Squeeze-In, introduce yourself  to me please. I'll be doing a Friday Night workshop on maintenance & repair.

[4to5to6]

1 hour ago, Matthew Heumann said:

With regard to someone butchering it with a Dremel: " A good craftsman never blames his tools".

I've been tuning concertinas for 30 years with a Dremel using a rubber bit infused with micro abrasive and have NEVER had a bad result. Quite the contrary, the rubberized bit, using gentle brief quick strokes, does not over-heat the reed, and actually buffs away the metal without leaving the abrasion grooves of files or automotive sandpaper.

Be it a file, sandpaper, or Dremel, an unskilled or careless user will damage a reed.

 

That being said, I have also encountered individual reeds that defy voicing, am actually working on a lower reed that refuses to sound "nice". Profile is good, position in frame is good, fit in reed slot is good, yet the damn thing is harsh. So I understand the frustration.

 

When you come to the Northeast Squeeze-In, introduce yourself  to me please. I'll be doing a Friday Night workshop on maintenance & repair.

I’m using a tuning bellows I received from Bill Crossman maybe 10 years ago. It works fine.  It is just a nicely finished platform clamped to the table with an old bellows closed off and weighted at the bottom that drops down pulling air through a couple of different size reed holders on the top and an air valve.  I can compare two reeds or seal one reed holder off.  I use a couple of different files, filing roughly 45 degrees with the courser one so there are no straight across stress points up to almost straight across with a micro fine file as a final touch up. it’s like 2000 grit sandpaper. Sometimes I’ll use a convex file on the belly of a smaller reed to get a bit of a curve in the central belly area when thinning a new small tongue.  I make sure that I’m not in a rushed head space or will just leave it all for the next day.  Patience is key and also humming or whistling while you work 😄.

 

i was once told by a restorer near Spokane Washington that most concertinas have had their reeds butchered and he scared me away from tuning for years.  Steve Dickinson was a big help in couching me along and helped me greatly in understanding how the acoustics of concertinas work.  David Hornett also greatly helped me when he sent me a package of different thickness sheared off blue steel strips and some old reeds to practice on.  What a great guy!

 

The bad reed tongues were either gouged out or thicker on one side of the reed than the other.  I’ll attach a photo.  Many reeds were curled down where the metal was removed and were left this way so many needed to be reshaped by gently bending them up close to the root and then bending the entire tongue down to adjust the gap.  It’s been quite the learning process.  I reversed engineer about a dozen of my high end Aeola doing reed scale studies before ever touching a reed ranging from tiny tin foil thin micro piccolo reeds to large weighted reeds in a double acting GBass and a large single acting C bass so got to know intimately what a very good reed looks like.  Obviously still learning. Fixing this butchered instrument is my apprenticeship!  What a challenge.  I’ll attach a photo of a bad reed I had to replace the tongue in.  I usually just copy the opposite reed or the next note up or down as a starting point.  I’ve found loose clamp screws and also broke a clamp screw the other day from over tightening one. Luckily there was enough of the broken screw threads to grip it and remove it.  I’m learning the hard way.

 

I’m open to all the device I can get and accept that there is no exact one “right way”.  I often wish I could go back a hundred years and hang out at the Wheatstone factory where they had multiple generations of accumulated experience.  The old mid 19th century 12 Guinea and better concert level instruments with the early steel reeds are amazing!  Regondi, Case and Blagrove used them as well as Emily Bulteel and Miss Binfield.  I wish I could play half this well but that is a different topic.

 

All the best,

John

[Matthew Heumann]

Yep! That looks like one messed up reed. I recently replaced a reed that had been filed so thin at some point, that the tip literally crumbled off. Decades ago, Frank Edgley, once a concertina restorer,  told me that one of the reasons he started making new concertinas was that it was more time-efficient & less frustrating than trying to fix wonky old ones!

Best of luck!!!

[Chris Ghent]

In your quest to get faster response times for your higher reeds, try taking off the valve on the other side of the reedpan to assess the potential of the unvalved reed.  This at least eliminates the valve.  Create the smallest gap you can between reed and frame; if you don’t need a microscope to fit it then it probably could be tighter.  Make sure the reed makes a nice curve when you raise the tip.  No flat spots and no places where it curves suddenly.  Get as much of the reed curving as you can, no (oh OK, very little) long unbending parts close to the clamp. Flatten the underside of the reed frame after you have tightened the clamp, they often bend up and this is a source of leaks. Copy the profile of a reed you know works well. Pay attention to the size of the chamber.  Make it smaller by temporarily gluing wood slips to take up space. 
 

In general a fast loud reed is tight in the frame and has a small chamber. 
 

Its true there are a lot of factors! 

 

Edited to add, if you look sideways at the profile of a reed and you can see under the reed before the point at which it bends up to its resting position, so about the first third of the reed, that will cost you efficiency.

 

 

Edited April 19 by Chris Ghent

[4to5to6]

Thanks Chris. Can you recommend a scope to view the reed gaps?  What magnification is best?  I’ve been just holding them to the light but should build a light table with a scope for setting the tongue.  
 

It seems to me that the gap should be as small as possible taking into account expansion of the wood that could squeezes the frame in therfore binding the reed.  When I was doing my reed scale studies, I was very impressed when the gaps were under 0.0005” (0.013 mm)… practically unmeasurable!  
 

Sometimes a reed worked fine with a poor fit strangely enough… I found this true in the 31xxx instruments so concluded it was the type of steel used that made them sound good. It was rare to find a nice scaling curve and nice workmanship, etc. together… only my amboyna 31xxx TT had it all.  And Steve Dickinson did the work on it.  Balance, responsiveness, quality gaps… and then Steve’s setup and voicing plus it is in absolutely pristine, well taken care of instrument.
 

I think your advice about long sweeping curves with no sudden changes or flat spots may be the key.  I’ll remove the valves as well and flatten the bases and make sure there is no gap in the first third of the tongue looking sideways. It is especially important to me to have good balance between the high and low notes so quieter bass reeds and louder treble reeds with ideally the same response high to low. I’ll fine tune this with the valve weights as well.  Another topic.  I’m not going to give up.  Thank you.

.

Edited April 19 by 4to5to6

[Chris Ghent]

4,

I have never seen a gap under 20 microns on a vintage reed. It is very hard to measure; I had mine measured with a very fancy piece of kit by a bloke who worked in an armaments factory.  I find 10x magnification is too much, mine is 7 and I wouldn’t mind if it was 6. You can see a lot with a loupe but a steady platform for viewing and controlled lighting while working is an asset.  
 

Expansion of wood is no trouble with the majority of concertinas as the gap is not tight enough to make a difference.  Once the gap is tight there is nothing you can do to stop the wood expanding with humidity and trapping the reed. Consequently you must either relieve the frame or the wood. Some Wheatstones I have seen have the wood relieved alongside the reed window.  You can feel the frame click into the tighter part of the wood as you push it in, just before it stops moving. 
 

 

[4to5to6]

Thanks Chris, I will search for a 6x digital scope and mount it on a small light table.  I have handled maybe 1000s of reeds and did get really good at eyeing them, holding them up to the light but having them held perfectly flat against a surface and looking at them straight on just makes sense to increase consistency.  My measuring tools would measure zero gap but obviously there was some so would write down 0.0005 (0.013mm). 0.0008 may be the actual amount.  Strangely enough, I have heard some very good sounding reeds with inconsistent gaps so am not convinced this is the most critical factor.  Still learning so don’t take what I say as absolute.

 

Where I live, humidity can vary between 20% indoors in the winter to 80% outdoors when raining.  I have had to do exactly what you said at times… take a couple of swipes with a fine file to remove a very slight amount of wood right next to the reed slot if a metallic sound develops (reed tongue rubbing on frame).  I have very few problems over all.

 

I’m home in a few days and will continue on until my reeds are perfect.  All advice is welcome.  What is the ideal belly (center) thickness/stiffness?  Too thin, they choke… too thick. Low responsiveness.  There’s definitely a learning curve in all this… especially when rescuing butchered reeds.  
 

Reeds have to “played in” so results aren’t instant which makes it difficult as well.  I presume this “playing in” is building up the work hardening of the reed.  I’ve had instruments that haven’t been played in 20 plus years come to life in the most amazing way after a few days to a few weeks playing.

.

Edited April 20 by 4to5to6

[wunks]

6 hours ago, 4to5to6 said:

  All advice is welcome.  

.

Something to try that is easy:  Use a small thin piece of refrigerator magnet as a weight/thickener at various locations along the reed.  moving it towards the tip will increase pitch so try mid-reed first.  A piece a bit less than a third of the reed length for starters.  This may not solve the problem but might be a useful indicator of how to proceed.

 

Sorry.  toward the tip lowers the pitch

Edited April 20 by wunks
correction

[LesJessop]

Would a set of feeler gauges be a useful tool for measuring and setting the gaps? I don't speak from experience - I've only ever re-set one reed, and that was a low-B on a tutor Lachenal that just sounded 'Blerrrr'. The gap between the end of the reed and the block seemed enormous, and moving the reed to narrow the gap solved the problem.

 

So, 100% success. But if my aeola had the same issue I would leave it to an expert ! 

[4to5to6]

I have always found that the better the instrument, the easier it is to work on.

 

I like the tiny piece of magnet idea.

 

New reeds are easier to set as you have a piece of reed hanging out the end to use as a handle before snapping it off.

[wunks]

See the above edit.  I've had great success lowering the pitch on very low pitched reeds. My low cello Eb ( to make a bisonoric pair with the low F ) never budged since I installed it but a recent try with a much higher pitched reed and a thicker strip of card stock didn't work, so a shorter, thinner strip might be better.

[Chris Ghent]

2 hours ago, LesJessop said:

Would a set of feeler gauges be a useful tool for measuring and setting the gaps? I don't speak from experience - I've only ever re-set one reed, and that was a low-B on a tutor Lachenal that just sounded 'Blerrrr'. The gap between the end of the reed and the block seemed enormous, and moving the reed to narrow the gap solved the problem.

 

So, 100% success. But if my aeola had the same issue I would leave it to an expert ! 

The smallest gauge in a feeler set is usually around 1.5thou, or .037mm.  I ground one down to .025 (1thou) about 20 years ago but it was impossibly unwieldy especially since one of the parts between which you are measuring is not fixed.  You don’t need a gauge to adjust the reed into the centre of the frame slot as you can very successfully eyeball this.  The only time you need to know the “exact “ gap is if you are trying to create a certain size through predictive engineering and want to know how well you are doing.  
 

When I got the gap measured (optically) and was told in that example it averaged 18 microns I kept that reed assembly, and I think of the gap in it as “9” . Comparing it under the microscope to other assemblies I can easily see this one is a 7, that one is a 10.  

[Jake Middleton-Metcalfe]

11 hours ago, Chris Ghent said:

The smallest gauge in a feeler set is usually around 1.5thou, or .037mm.  I ground one down to .025 (1thou) about 20 years ago but it was impossibly unwieldy especially since one of the parts between which you are measuring is not fixed.  You don’t need a gauge to adjust the reed into the centre of the frame slot as you can very successfully eyeball this.  The only time you need to know the “exact “ gap is if you are trying to create a certain size through predictive engineering and want to know how well you are doing.  
 

When I got the gap measured (optically) and was told in that example it averaged 18 microns I kept that reed assembly, and I think of the gap in it as “9” . Comparing it under the microscope to other assemblies I can easily see this one is a 7, that one is a 10.  

That is quite interesting, like you I  don't actually measure the gap, but just judge it by eye under magnification, one just gets on and does it. I always wondered what it actually was though. How did they actually measure the gap? Was it some sort of microscope with a very small ruler?

[4to5to6]

I measured the shoe opening width and subtract the measured width of the reed at the same location.  Small high quality reeds were commonly under 0.0005” (12.7 microns)… A piece of paper is 0.004 (twenty times this) so the gap is really unmeasurable  by usual methods.

Edited April 22 by 4to5to6

[Chris Ghent]

I struggle with numbers with too many noughts in them, but I think we have some crossed wires here.  I make .0005 (half a thou) to be 12.7microns rather than 127.  Your piece of paper which you quote at .004 (4thou) I think is 8 times as thick, rather than 20. A micron uses the symbol “μm“

 

I was brought up in NZ under the old imperial measurement system and while we intrinsically understood base 10 metric had to be better it wasn’t until the 70s we started to change.  Many people my age still talk in yards, feet and inches. Mostly I have abandoned it in practice except for one measurement, “thou”.  This is because when talking of the thickness of a reed there is an ability to talk in whole numbers with thou.  “This reed is 7.5thou at this point, I’ll ease it down to 6.5thou” is easy to assimilate compared with “it is .187, I’ll ease it down to .162.”  There are 40 (roughly) thou to a millimetre and 25 microns to a thou.  
 

I have never seen (not the same as it doesn’t exist) a vintage instrument with reeds at 12.7 microns.  I have seen a couple of Dippers at about that level, had one with John Dippers name on it here a week ago with about 15, and I don’t doubt Wally Carroll could do it but I haven’t seen it.   127microns as you say, then? Say 62 each side, almost 2.5thou a side. Seen plenty like that.  There are good reasons older instruments mostly did not have tightest clearances, amongst them the cost and the higher harmonics generated by tight clearances when the EC was being aimed at a parlour audience and edginess in the sound was not valued.  The best clearances I have seen in vintage concertinas are in Wheatstone Model 59 and associated models (the 59 has 30 keys, the other similar models have for example 36 keys with the same quality) with Class C reeds. These instruments are played by several name players in Ireland, Noel Hill, Hugh Healy, Jack Talty come to mind. From memory they would be around the 18 micron mark.  
 

There are issues in the simple measurement, subtract the reed overall from the slot overall.  With my reeds it is impossible because the reeds are tapered.  A vernier calliper is a good daily tool but when looking for microns you really need a micrometer and one with the right jaws would be a rare thing.  Lastly, you would need to make these measurements under a microscope because the local fluctuations (think file marks) in the surface of the side of the reed and to a lesser amount the frame could be up to .5thou (12.5microns).  In fact the engineer who measured my reed clearance said “it averaged 18thou”, something I was happy with in what is essentially a scraped fit.  
 

A tight fit is a factor in response and maybe volume, but it is not the only way to skin the cat.  The shape of everything else, the profile, the slot, plus the steel, these things are factors.  No doubt at all Wheatstone had some great reed makers.   A tight gap actually has some inherent issues apart from higher harmonics, plenty of dust is thicker than 18 microns and some of it is very hard.  Human hair ranges between about 18 and 180 microns.  

If any of my figuring seems out I’m happy to be corrected, never too late to learn though it won’t affect my working practice as that is now a visual recognition method based on knowing the size of a particular gap and I could call it 3 freds as long as I get close to it.  
 

Cheers

 

Chris

[Chris Ghent]

Jake,

the bloke who measured my reed rattled off a number of instruments they had which could measure things to various levels and after each he would say things like, that one cost $1,000,000.  We were at the dog park at the time and I was not on record so much as focused on the goal of getting a good reading from any of the machines.  And whether my dog had had a crap yet.  My recollection is the machine had an optical element but also some sort of computer interface such that they could ask it the width of the light line, or perhaps it was the area of the light line which it could turn into an average width.  
 

Optical comparators were common once  (I think David Hornet has one) but they are relatively useless in industry now and would be cheap.  They essentially have a lens and a graticule and would probably be adequate for a simple inquiry. I would try and find someone with one and ask them to measure because then I would not have to store the machine.  
 

I would be very surprised if Dave Eliot was not very across this technology.  
 

Cheers

 

 Chris

No dogs were harmed etc. 

Edited April 21 by Chris Ghent

[4to5to6]

127 was a typo.  12.7 microns equals roughly 0.0005 thousands of an inch.  I’m in Canada and even though we switched to the metric system in the 70s, most machinists tools, lathes, etc. are still in inches.  Sorry for the confusion… I was just trying to be more internationally friendly… obviously being very badly at it.  “There are 40 (roughly) thou to a millimetre and 25 microns to a thou.” I’ll try to remember this Chris.

 

I think I mentioned that I reversed engineered the reeds on maybe 20 or so of the best Aeola Wheatstones that I could get access too.  I still own many of these concertinas and they are absolutely wonderful.  These are mostly golden era amboyna, tort, gold ends, etc… the cream of the crop but also concert level from the mid 1800s when steel reeds were just staring to be used.

 

I think that I also mentioned that the reed gap does not seem to be the critical factor in the performance of a reed as I have seen and heard reeds that I would probably reject visually yet sounded and played wonderful.

 

My question to those with much more experience than I is what are the design factors that make a good reed?  There is the tip, the belly and the root closes to the clamp… the pitch is determined by the length, width, thickness of the reed, etc…. the ratio between the tip weight and thickness or stiffness of the root.  What is the ideal thickness of the belly?  When is it too thin?  A nice flowing, filed curved profile seems to be ideal.  Should the bottom surface of the tongue be flat?  Bent down, than up to have the first third below the surface then slowly rising up?  Metal seems to bend away from the surface that is filed??? Do you simply bend it back into the proper curve after filing?  I can look at my charts and see what has been done as far as physical thickness and reed scaling and look at the spread sheet curves but I’m looking for advice on the theory for setting the reed by bending it and the ideal gap along the length is still a bit of a mystery to me.  Also belly thickness… when is a reed tongue too thin?

 

Some theory on the purpose and ideal thickness of valves may be helpful.

 

I’m back home now so will be back at it tomorrow.  Any advice on reed design, profiles, voicing, etc. will be most welcome.