Melodeon.net Forums

Please login or register.

Login with username, password and session length
Advanced search  

News:

Welcome to the new melodeon.net forum

Pages: [1] 2   Go Down

Author Topic: Treble wooden key length & leverage ratios  (Read 3603 times)

0 Members and 1 Guest are viewing this topic.

mChavez

  • Hero Member
  • *****
  • Offline Offline
  • Posts: 713
Treble wooden key length & leverage ratios
« on: August 09, 2024, 07:41:40 AM »

Hi All.

Does anyone have the information for 2-row wooden treble keyboards on the effective length of the keys (centre-of-button to axle) and key leverage ratios to get a nice & light action without self-lifting pallets? For 2- and 3- voice instruments.
Or perhaps you have 5 minutes and can quickly put a ruler to your favourite squeezebox that plays like a dream? ;)

I've got plans from an ancient soviet book with 31mm for the short row and 48mm for the long row for a 2-voice instrument, but we all know, you can't trust the soviets  :-\ The axle-to-pallet is roughly 45mm and 85mm, giving leverage ratios of 0.65 and 0.56.

For a 3-voice instrument, the measurements of keys on a Soviet stamped-metal "Shuya", 47mm & 67mm for a 3-voice box. With axle-to-pallet lengths of 47mm and  100mm, we get 1.00 and 0.67 ratios.

For another 3-voice instrument I have 33x40 56x95, giving 0.82 and 0.59 ratios.
This very quickly runs into self-playing mode on the 0.59 ratio once the springs are relaxed below 130g.

So come on. Take the ruler out, and let's see who's got the biggest (leverage ratio ;D). Any help appreciated.
« Last Edit: August 09, 2024, 10:10:06 AM by mChavez »
Logged

mChavez

  • Hero Member
  • *****
  • Offline Offline
  • Posts: 713
Re: Treble wooden key length & leverage ratios
« Reply #1 on: August 11, 2024, 08:41:00 AM »

Hohner Morino seems to try and get the ratio closer to 0.7 by the looks of it. That's a 3-row keyboard, and you can't get even ratios across all 3 rows, so I'm getting about 0.6, 0.7, 0.8 for 3 rows.

The way I understand it, the leverage affects several factors:
Effective ratio is higher, than the ratio of arm lengths, as the axle is offset a couple cm over the soundboard.
With button travel of 4mm, and arm ratio of 1, we'll be getting a very light keyboard (for a given spring strength, the required push force will be lower), but very low pad lift (< 4mm). Probably unacceptable.
With arm ratio around 0.5, we are getting plenty of pad lift (more than we need), but it is a lot harder to press the button. We might run into the case where the lightest spring that prevents self-playing pallets requires well > 120-130g of force.

I guess I'll just aim for 0.7 arm ratio, unless there are known reasons to pick something else?
Logged

Chris Ryall

  • "doc 3-row"
  • French Interpreter
  • Hero Member
  • *****
  • Offline Offline
  • Posts: 11153
  • Wirral UK
    • Chris Ryall
Re: Treble wooden key length & leverage ratios
« Reply #2 on: August 11, 2024, 11:03:59 AM »


  we've talked about this before.  There will be a measurable but not always perceptible difference in the "action"  of various keys, and that will differ between melodeons.  Many (most?) have their reeds in sets of blocks, with the layout … to some extent dependent on the preferences of the maker. Few simple machines have the reed plates mounted directly into the Fondo = soundboard.

  A more important factor is that the reeds towards the front of the machine will get more air. Some of the more sophisticated makers, e.g. Frans van der Aa, use a double-action cantilever system to open valves to the anterior chambers, to counteract this. It certainly makes for a very nice and even action on your keyboard.   But it is also … more to go wrong 😉
« Last Edit: August 11, 2024, 11:06:09 AM by Chris Ryall »
Logged
  _       _    _      _ 

Graham W

  • Respected Sage
  • ****
  • Offline Offline
  • Posts: 277
Re: Treble wooden key length & leverage ratios
« Reply #3 on: August 11, 2024, 11:34:29 AM »

Assuming you are making or fundamentally modifying an instrument . . . .
You could make a simple jig with a pivot and some kind of graduated (graph paper?) background to test out pallet lift vs button travel vs spring tension for various lever dimensions.

My (limited) experience though is that in reality there is a danger of overthinking this, and a lot of dimensions are already dictated by other things.

The axle to pallet distance is fixed by the Fondo you are using (which in turn is dictated by a whole host of other variables) and the position of the axle.  Unless it is a very deep box then there is not much scope for varying the axle position, it's usually pretty much as far from the pallets as you can get it.

The distance between the button rows is largely fixed, maybe scope for very small variation but in general 17mm is common.

So the biggest scope for variation is with the distance of the buttons from the axle.
Depending on the box layout some levers will give larger pallet lift than others - if you are mapping 2.5 or 3 rows of buttons onto two reedblocks (piano montage) then there might be a lot of variation here, so identify which button gives least lift for the required depression and make sure the length of that one is such as to give at least the minimum required pallet lift. Some of the others might then give significantly more but in practice that isn't a problem provided that that the minimum is sufficient.

Also, does the position of the buttons and the whole keyboard look 'right' aesthetically in relation to the rest of it?

You can then make the springs to suit each row and give the button push force you require, which will probably need a spring wire with diameters differing by about 0.05mm between rows, with individual tension adjusted to suit each button.
For me, a button travel of approx 3mm and push force of approx 115g (on a 2V instrument) makes for a nice light responsive keyboard.
Bear in mind also that even small variations in pallet weight can make a significant difference to how responsive it feels too. The lighter the better in that respect.

Good luck.

Logged

Theo

  • Administrator
  • Hero Member
  • *****
  • Offline Offline
  • Posts: 14475
  • Hohner Club Too
    • The Box Place
Re: Treble wooden key length & leverage ratios
« Reply #4 on: August 11, 2024, 12:28:34 PM »

As with many design decisions there care conflicting factors that you have to balance to reach a good compromise. 

Thinking about button to axle length and with other variables fixed the two main ones are:
Longer button lever = shorter pallet lift (bad up to a point) and better mechanical advantage (MA) in overcoming inertia of the lever/button/pallet
Good playability will also depend on typical repertoire and style of the player. 
Fast playing with lots of notes would be favoured by minimising inertia, short button travel and possibly maximising MA.
If maximum volume is required then food pallet lift is the driving factor.
For legato, soulful playing then short button travel could be more important and MA and inertia less so.
In practice one box might be used for all these styles, so thoughtful compromise is the key to a good design.

And as Grantham has said, there are many other variables that affect the above to a greater or lesser extent.
« Last Edit: August 11, 2024, 12:30:12 PM by Theo »
Logged
Theo Gibb - Gateshead UK

Proprietor of The Box Place for melodeon and concertina sales and service.
See my YouTube channel  for videos on repair and tuning

mChavez

  • Hero Member
  • *****
  • Offline Offline
  • Posts: 713
Re: Treble wooden key length & leverage ratios
« Reply #5 on: August 11, 2024, 02:32:15 PM »

Designing my fist scratch build, so there's some room for imagination. And cutting my own everything, so not tied to a particular fondo design.
Still to decide whether to go for a 2- voice or a 3-voice. Keeping it simple, just a two row keyboard. 31x17.5cm bellows cross section, so quite a bit of space in the box for tweaking the key levers.

No need for a jig - this can be modelled in inkscape in a few minutes, but there's still a question of how much minimum pallet lift is actually required before lowest reeds become unstable.

Three limiting factors in this design are the button travel (4mm feels good for me), minimum acceptable pallet lift before the reeds become unstable (unknown) and the minimum strength of springs that will prevent the pallets from self-lifting on a strong push. The key leverage ratio determines how light the keys can be. A ratio of, say, 1, will not provide enough pallet lift. A ratio of 0.4 is highly unlikely to allow 120g of button resistance. You can make springs that will give you 120g of resistance, but these springs will be too weak to hold down the pallets, at least on a 3-voice box.

I could not find any plans or information on melodeon keyboard, apart from an ancient 1937 soviet book that suggests what looks like a very short keyboard, hence the question.
My experience with squeezeboxes is, unfortunately, underthinking details like this to be honest, and it's too easy to make a pig's ear of the job.

Chris's point to the angle of pallet lift is valid for 4+ voice instruments, but I think it's negligible for a 2-voice, and even a 3-voice can be set up well enough not to require reverse-mounted pallets.
Logged

mChavez

  • Hero Member
  • *****
  • Offline Offline
  • Posts: 713
Re: Treble wooden key length & leverage ratios
« Reply #6 on: August 11, 2024, 03:00:07 PM »

Weight of pallets, buttons, etc... Interesting thought, but I am sceptical to accept it as one of the defining factors of a good keyboard in all honesty.

If it really mattered, we'd see hollow carbon fibre pallets on all high-end concert accordions and cf buttons instead of the incredibly heavy MoP. The lightest/fastest keyboards I've tried so far are Hohner's old brass mechs (Morino, 235) with quite large beech pallets. Brass is heavy. Beech is not particularly light either. Italian accordions often use quite a bit of reed wax to hold the pallets in place, and that stuff can add a few grams too. I reckon the feel is mainly defined by the leverage & spring tension that we need to overcome.
I'm sure that pallet weight, as well as lever stiffness do come into the equation, but I wonder how important these factors are compared to lowering the spring tension by 5-10 grams.
Logic suggests that increasing the soundhole area will also allow the air to flow to the reed faster, resulting in faster response for a given reed, but we're probably splitting hairs with all of this, and the impact is dwarfed by good leverage+spring selection.

On the other hand, there's nothing stopping me from using a cf-balsa laminated sandwich for pallets. And I intend to use titanium rods to link pallets with keys. (Mainly because I've got a stack of them lying around, doing nothing  ;D). Heavier than aluminium - probably, but also a lot stiffer, which can't be a bad thing for a loaded lever.
Logged

Graham W

  • Respected Sage
  • ****
  • Offline Offline
  • Posts: 277
Re: Treble wooden key length & leverage ratios
« Reply #7 on: August 11, 2024, 06:12:33 PM »


I'm sure that pallet weight, as well as lever stiffness do come into the equation, but I wonder how important these factors are compared to lowering the spring tension by 5-10 grams.
Logic suggests that increasing the soundhole area will also allow the air to flow to the reed faster, resulting in faster response for a given reed, but we're probably splitting hairs with all of this, and the impact is dwarfed by good leverage+spring selection.

There are a lot of hairs to split though!    The things you mention;  but also the reed quality, valve selection, tuning and reed chamber volume then come into that equation and have their own influences on the 'feel' that your finger ultimately perceives at the button.  I'm not sure that any one thing dwarfs any other.

Making a prototype or test-bed box and then being prepared to re-make or replace parts of it to try the effect of changes is an interesting and revealing exercise . . . .
As for pallet lift, I once had a Castagnari 2V MM D/G which measured 5mm lift at the tip of the pallets, so somewhat less at the heel. For me, min 5mm at the heel works if there are low voice reeds.
Logged

mChavez

  • Hero Member
  • *****
  • Offline Offline
  • Posts: 713
Re: Treble wooden key length & leverage ratios
« Reply #8 on: August 15, 2024, 07:41:00 AM »

Just in case anyone's after some plans.
I've never managed to find any melodeon plans online at all, so here's the only one I've managed to dig up, which is from a 1937 soviet book on harmonikas.

Maybe somebody will find it useful as a starting point if they are building their own box.

I disagree with how the key leverage is done here, and in my own build I've extended the short key by 5mm, and the long key by 5+3mm (giving ~21-22mm between rows, which was my target).
For a 2-voice box you might get away with the leverage in the plan, but I expect the extension will give a lighter keyboard anyway.
For a 3-voice box I don't expect these sizes to work without some self-raising pallets at 120-130g at the button.
« Last Edit: August 15, 2024, 08:06:16 AM by mChavez »
Logged

mChavez

  • Hero Member
  • *****
  • Offline Offline
  • Posts: 713
Re: Treble wooden key length & leverage ratios
« Reply #9 on: August 20, 2024, 07:32:56 PM »

Well, she might not be perfect, but not bad for the first attempt, even if I say so myself.

Since it's a 2-voice, I think that I'll have some wiggle room as to how far back I can position the keyboard, so the ratios can still be adjusted.
Next step is the keys, but fitting them so that they are not loose in their slots, yet, they don't bind when the humidity rises, might be a challenge for a first time builder.

[img width= height= alt=PXL-20240820-180255014" border="0]https://i.ibb.co/mCPDt31/PXL-20240820-180255014.jpg[/img]
[img width= height= alt=PXL-20240820-180319933" border="0]https://i.ibb.co/FJ3pQSH/PXL-20240820-180319933.jpg[/img]
[img width= height= alt=PXL-20240820-180305361" border="0]https://i.ibb.co/kxDYTQ3/PXL-20240820-180305361.jpg[/img]
Logged

Aaron Wildman

  • Member
  • *
  • Offline Offline
  • Posts: 36
Re: Treble wooden key length & leverage ratios
« Reply #10 on: August 27, 2024, 11:35:34 PM »

Wooden levers with the hinge pin through the keybed?  I like the idea in general, it is a good simple approach, as long as the wood is well seasoned so it doesn't swell/warp/shrink. 

I have two ancient boxes with that basic design, although they are cheaply made.  My biggest gripe is that the hinge wire diameter is so small that there was no way to safely remove it.  Snap!  Now they just sit in pieces, relegated to the bottom of my project list.
Logged

mChavez

  • Hero Member
  • *****
  • Offline Offline
  • Posts: 713
Re: Treble wooden key length & leverage ratios
« Reply #11 on: August 28, 2024, 08:16:58 AM »

Well seasoned or not, you can't stop wood the from moving. If it's not well seasoned, it will be a disaster, but even if it's seasoned, it will move with humidity changes at these thicknesses.

I'm using 3 mitigation techniques in my first build:
- Keys laminated from two pieces of wood instead of one. 3 would have been better, but I'm a lazy bugger.
- Light touch of shellac on comb and keys - the shellac does not stop the moisture exchange with the air, but it slows it down significantly and prevents shocks. The jury is out on whether this is a good thing or not, but I'm a big believer in shellac.
- I'm installing PTFE bushings into the keys, so instead of going straight through the wood, the axle goes into ptfe.
Logged

mChavez

  • Hero Member
  • *****
  • Offline Offline
  • Posts: 713
Re: Treble wooden key length & leverage ratios
« Reply #12 on: February 02, 2025, 05:50:08 PM »

Slow & steady. My wrist has healed and I've made some moderate progress on the build.

Not only should we think about treble key ratios, but there's quite a bit of thought that needs to go into the bass end, especially if it's a roller mech.
The force of the spring needs to be sufficient to keep the pallets down, yet, the leverage starts getting a bit more complex than the treble end. There's some room to play with the heights of roller posts, but I just had to eyeball it for the first build.

Will keep you updated on the progress.

[img width= height= alt=PXL-20241215-092148163" border="0]https://i.ibb.co/xSQXXgTs/PXL-20241215-092148163.jpg[/img]
[img width= height= alt=PXL-20250202-154456751" border="0]https://i.ibb.co/TDRbr4ck/PXL-20250202-154456751.jpg[/img]
Logged

mChavez

  • Hero Member
  • *****
  • Offline Offline
  • Posts: 713
Re: Treble wooden key length & leverage ratios
« Reply #13 on: March 25, 2025, 12:23:56 PM »

Just to show that I'm not all talk and no action (pun intended >:E), here's some melodeon keyboard porno for you.

PXL-20250324-194349478" border="0

I think there is something undeniably sexy about a wooden action. It will never be as smooth or easy to install as the aluminium one, but if you ask me, aluminium belongs in beer cans. Keep your keyboards wooden.

I've managed to put about 110-120g on the keys without any excessive self-playing. Pallet lift is 4mm for bottom row and more for top row. I reckon there's some scope for lightening the springs on future builds. Problem is - I've lost my ratios notes, so will have to re-measure.
Logged

mselic

  • Hero Member
  • *****
  • Offline Offline
  • Posts: 1683
Re: Treble wooden key length & leverage ratios
« Reply #14 on: March 25, 2025, 08:06:12 PM »

Very nicely done! Impressive : )

It's funny, I was just today working in the shop on a vintage accordion with a wooden keyboard assembly and thinking how much I loathe them (or at least the problems one encounters with them) ;D  I can't think of any advantages wooden keyboards have over aluminum ones other than the wooden ones can sometimes look quite lovely like yours...:)

Incidentally, Melodie (one-row accordion) builder Sylvain Vezina uses some kind of delrin-like material for his 'wooden' keyboard levers which he then tops with a wooden veneer.  You can't tell that it's not all wood to look at it, but he manages to avoid all kinds of humidity/age-related issues that one encounters with wooden levers.  Not that I'm suggesting you do anything other than what you've already done!
Logged
C#/D player

mChavez

  • Hero Member
  • *****
  • Offline Offline
  • Posts: 713
Re: Treble wooden key length & leverage ratios
« Reply #15 on: March 25, 2025, 09:57:58 PM »

Well, a lot of plastics are not really dimensionally stable and can bend all over the place.
I reckon my approach (2-ply laminate, shellac finish, PTFE bushings) will be quite stable.
The metal levers are not hammered into the keys like they used to do in the old days, but I've got a threaded insert press-fitted, then glued into the key (less stress and lower chance of cracking with age).
The buttons are also screwed into threaded inserts rather than directly into the wood.

The only part that can really cause some issues is the comb. Slithers of wood with the grain exposed on both sides.
On my second keyboard I was daft enough to only put shellac on one side of the slots and then head into the house for dinner... When I came back to the keyboard, all the comb walls have bent really badly - probably close to 45 degrees!
Took a few days in dry room for them to stabilize. Then soaked them in shellac and they were stable ever since.

There's highest quality hand-made reeds on the other side of that fondo... I'm in love. Unfortunately, today's society is not going to let me get officially married to this squeezebox. Even in Europe.
Logged

boxcall

  • You got to love it!!!
  • Hero Member
  • *****
  • Offline Offline
  • Posts: 1801
Re: Treble wooden key length & leverage ratios
« Reply #16 on: March 26, 2025, 12:00:37 AM »

Looking good!
On the one row I built, I used cherry for the levers which is a very stable wood when properly dried. The levers are the part that have the biggest risk of movement or swelling mainly because of the way the grain is running. The dividers between the levers should not move much at all if it is the end grain that touches the levers ( well not really touching) wood doesn’t swell or move much in that direction, just in the other two directions.
I have not had a single problem with the keyboard on my one row and it gets pretty humid around here in the summer and very dry in the winter. I also must have done something right with the stop slides as they don’t stick and they stay up :)
Fingers crossed that I didn’t just jinks myself!!
Anyway looking forward to seeing your finished box!
Logged
Hohner 1040 C, Beltuna one row four stop D, O'Byrne Dewitt/ Baldoni bros. D/C#, Paolo soprani "pepperpot" one row D

mChavez

  • Hero Member
  • *****
  • Offline Offline
  • Posts: 713
Re: Treble wooden key length & leverage ratios
« Reply #17 on: March 26, 2025, 02:42:47 PM »

Thanks. Still a bit of work to do.

My problem is that over the years I've accumulated a bag of offcuts that are too small to sell or make anything...apart from a melodeon. When I say offcuts, I mean woods like cuban mahogany, bocote, Indian rosewood, figured ozzie blackwood, walnut. Enough for at least for 6 squeezeboxes.
I guess that's my weekends planned for a few years ahead  >:E
Logged

boxcall

  • You got to love it!!!
  • Hero Member
  • *****
  • Offline Offline
  • Posts: 1801
Re: Treble wooden key length & leverage ratios
« Reply #18 on: March 26, 2025, 05:21:02 PM »

Nice! Yes that should keep you busy for a while 😀
I built my box out of Spanish cedar which is like a little mahogany but much lighter.
Also I didn’t use anything on the wood to finish other than the exposed areas.
I’m working on a couple more boxes ( slowly) made out of Birdseye maple with a bit of figuring.
Logged
Hohner 1040 C, Beltuna one row four stop D, O'Byrne Dewitt/ Baldoni bros. D/C#, Paolo soprani "pepperpot" one row D

mChavez

  • Hero Member
  • *****
  • Offline Offline
  • Posts: 713
Re: Treble wooden key length & leverage ratios
« Reply #19 on: March 27, 2025, 09:01:23 AM »

I've made a few guitar necks out of Spanish Cedar, but I prefer mahogany for this job.
Smells lovely! Easily dented though.

Birsdseye I've worked with too - can be a bit tricky.

Lovely one-row!
Logged
Pages: [1] 2   Go Up
 


Melodeon.net - (c) Theo Gibb; Clive Williams 2010. The access and use of this website and forum featuring these terms and conditions constitutes your acceptance of these terms and conditions.
SimplePortal 2.3.5 © 2008-2012, SimplePortal