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 [3] 4 5 6   Go Down

Author Topic: Melodeon construction -- some more radical thoughts...  (Read 34555 times)

0 Members and 1 Guest are viewing this topic.

Falseknight

  • Respected Sage
  • ****
  • Offline Offline
  • Posts: 449
    • False Knight
Re: Melodeon construction -- some more radical thoughts...
« Reply #40 on: December 18, 2008, 07:57:26 PM »

This might be interesting for bases and reed blocks

http://www.trelleborg.com/en/Advanced-Engineering-Materials/Products--Solutions/Tooling-Block-TB650/

From what I remember when the material was being developed it makes MDF look like a tone-wood.
Logged

risto

  • Respected Sage
  • ****
  • Offline Offline
  • Posts: 317
Re: Melodeon construction -- some more radical thoughts...
« Reply #41 on: December 18, 2008, 08:51:42 PM »


[Still doesn't mean there might not be something better (and "modern") as far as the instrument sound is concerned :-)

Ian


OK, but how do you define better? There is always someone else preferring the other thing.

No doubt there are, for instance, violins (like Stradivarius) widely accepted to be better than other brands, but isn't this more because of the construction than the acutal material used, don't know? Anyway, different instruments from different builders using same materials probalby never sound the same. Besides, even when using a certain material you may be able to produce nearly similar response with another material just by changing constrution properties like material thickness, shapes of membranes etc.

Quote
P.S. Since wood is actually an amazingly complex composite material, it's also possible that there *isn't* anything better.

Well, if someone invented something like true synthetic wood, at least one could expect isntrument be more alike sound wise.

PS. If I had to develop a better sounding melodeon, I probably would start from trying to find a frequency spectrum which pleases the ear more than the current ones and only after that, would try to find a way to produce it.

EDIT: typos etc.
« Last Edit: December 18, 2008, 09:01:16 PM by risto »
Logged
Paolo Soprani 5-row, Roland FR1B-digital CBA button accordion. Guitar, 5-string OT banjo, mandolin, piano...
My one man band recording: https://apari.fi/Bonaparte_Crossing_The_Rhine_Risto.mp3

IanD

  • Hero Member
  • *****
  • Offline Offline
  • Posts: 1703
  • Too many melodeons...
Re: Melodeon construction -- some more radical thoughts...
« Reply #42 on: December 18, 2008, 09:33:27 PM »


[Still doesn't mean there might not be something better (and "modern") as far as the instrument sound is concerned :-)

Ian


OK, but how do you define better? There is always someone else preferring the other thing.

No doubt there are, for instance, violins (like Stradivarius) widely accepted to be better than other brands, but isn't this more because of the construction than the acutal material used, don't know? Anyway, different instruments from different builders using same materials probalby never sound the same. Besides, even when using a certain material you may be able to produce nearly similar response with another material just by changing constrution properties like material thickness, shapes of membranes etc.

Quote
P.S. Since wood is actually an amazingly complex composite material, it's also possible that there *isn't* anything better.

Well, if someone invented something like true synthetic wood, at least one could expect isntrument be more alike sound wise.

PS. If I had to develop a better sounding melodeon, I probably would start from trying to find a frequency spectrum which pleases the ear more than the current ones and only after that, would try to find a way to produce it.

EDIT: typos etc.

Of course "better" is subjective when it comes to sound -- one man's ideal melodeon sound might be hated by somebody else.

But I expect nobody would think that a melodeon made of rubber would sound better than one made out of wood no matter how it was built, so it's quite possible that something generally agreed to be "better" than wood exists. For certain, if nobody tries anything new we'll never know -- AFAIK the only melodeon builder who's really tried non-conventional materials (balsa reed blocks, foam body) is Stormy Hyde, and I don't think he's tried any of the modern composites -- either because they're difficult to work, expensive, or don't seem to offer any real advantage over wood.

Being realistic, the sound of a melodeon is much less sensitive to the materials its made from than a stringed instrument, but it is very sensitive to the way it's made (cavities, internal resonances etc). The big advantage of modern composites might not be to do with their resonant properties so much as their strength and weight -- using things like carbon fibre it would be possible to use much thinner and more open structures to build the casing which could lead to a "better" (opinion, of course) sound.

The extreme case of this would be the "reeds-out" construction I suggested, in which case the materials will have very little effect on the sound anyway (because there aren't any between the reeds and the listener) but could make it much easier to build something strong enough without needing big chunks of wood to take all the stresses. We might still not like the sound, but I'm certain it would be quite different.

Hmm, how difficult would it be to turn the end of a melodeon "inside-out" and still have it playable somehow? Might need something like Rob's "glove-box" idea, but even if a lot of bodging is needed it would be a lot quicker and easier than building an inside-out box from scratch only to find that it sounded awful...

Ian

P.S. The idea of "design-a-spectrum" would work -- if only we knew how to get from where we are to where we wanted to be :-)
(the resulting "ideal sound" may not be acoustically achievable)
Logged
Oakwood Model 4 D/G, Castagnari Dony D/G/#, Castagnari Tommy G/C, Baffetti Binci D/G, Hohner Preciosa D/G, Melos Bb/Eb, Lightwave SL5 and Kala California fretless basses

Falseknight

  • Respected Sage
  • ****
  • Offline Offline
  • Posts: 449
    • False Knight
Re: Melodeon construction -- some more radical thoughts...
« Reply #43 on: December 18, 2008, 10:09:29 PM »

Specify what the components are meant to do, and then you can specify the materials best suited to achieve this.

I suspect that the materials used are more to do with ease of production than the accoustic suitability.
Logged

IanD

  • Hero Member
  • *****
  • Offline Offline
  • Posts: 1703
  • Too many melodeons...
Re: Melodeon construction -- some more radical thoughts...
« Reply #44 on: December 18, 2008, 10:31:07 PM »

Specify what the components are meant to do, and then you can specify the materials best suited to achieve this.

I suspect that the materials used are more to do with ease of production than the accoustic suitability.

True of many instruments affordable by mere mortals :-)

Ian
Logged
Oakwood Model 4 D/G, Castagnari Dony D/G/#, Castagnari Tommy G/C, Baffetti Binci D/G, Hohner Preciosa D/G, Melos Bb/Eb, Lightwave SL5 and Kala California fretless basses

Waltham

  • Regular debater
  • ***
  • Offline Offline
  • Posts: 198
Re: Melodeon construction -- some more radical thoughts...
« Reply #45 on: December 18, 2008, 10:59:10 PM »

Aside from the problem of specifying the goal I would argue that it's not just (or even predominantly) the spectrum that you want to design, in almost every instrument the starting transient seems to be more important to the sound; I can dig out references if anyone wants but not tonight, Josephine.  Charles Taylor used to give a lecture in which he'd play a clip from a classical piece where a whole orchestra was playing and then all stopped apart from one instrument, and he'd ask who thought it was a french horn and who thought it was an oboe and so on, there'd be no clear consensus.

I'm a little skeptical about the desirability of improving the acoustic transmission through the walls as I suspect that most of the sound comes out of the hole under the pallet so another path is as likely to cancel as reinforce.  But this could be tested.

Since radical ideas have been called for here are mine:

1. Replace the valves with sliding shutters.

2. Put conical 'plugs' on the pallet faces that protrude into the cavity in the reed block so that their effective volume changes as the button is pressed.

3. Just build a full sized instrument configured like a Preciosa.

Each of these has to do with the Helmholtz resonator formed by the cavity under the reed (which, incidentally is responsible for tuning being slightly different when the reed is on the reedblock to when it is off).  It's my feeling that this resonator, is a very important factor for the sound (though I'm not sure I could put a percentage on it).  If it were a purely acoustical effect then the shape of the cavity wouldn't matter as the wavelengths are much too long to 'see' the shape; since it does there must be some interaction with the airflow but I don't think it's well understood yet.
Logged

stillicidia

  • Member
  • *
  • Offline Offline
  • Posts: 26
Re: Melodeon construction -- some more radical thoughts...
« Reply #46 on: December 19, 2008, 08:10:00 AM »


The extreme case of this would be the "reeds-out" construction I suggested,

Isn't that called a harmonium? - except they cover the reeds in that to keep the dust off
Richard
Logged
Dyslexic fingers and deaf as a post!
member of accordions anonymous - I admit it I am an accordion addict!
it all started with just one! just a little one! then another, and another...

Martin J

  • Hero Member
  • *****
  • Offline Offline
  • Posts: 902
  • Poole, Dorset
Re: Melodeon construction -- some more radical thoughts...
« Reply #47 on: December 19, 2008, 12:16:24 PM »

Each of these has to do with the Helmholtz resonator formed by the cavity under the reed (which, incidentally is responsible for tuning being slightly different when the reed is on the reedblock to when it is off).  It's my feeling that this resonator, is a very important factor for the sound (though I'm not sure I could put a percentage on it).  If it were a purely acoustical effect then the shape of the cavity wouldn't matter as the wavelengths are much too long to 'see' the shape; since it does there must be some interaction with the airflow but I don't think it's well understood yet.

Helmholtz resonator - wow.  Now if this is as scientific as it sounds it must have formulae to predict it's optimum performance for each wave length and even assist tuners in deciding how much difference it is going to make to a reed tuned outside the box.  Armed with such knowledge one should be able to design the 'perfect' sound chamber for each note and by extention a whole instrument.  I appreciate the wavelength comment but exact divisions usually work ie. ¼, ½, ¾ the wavelength etc.

PS. There is already a way of predicting the difference between in box and out of box tuning
Logged
Castagnari, Weltmeister, Giustozzi, Streb, too many Hohners.  No Strings Attached ceilidh band

TomB-R

  • Hero Member
  • *****
  • Offline Offline
  • Posts: 598
Re: Melodeon construction -- some more radical thoughts...
« Reply #48 on: December 19, 2008, 01:34:44 PM »

Each of these has to do with the Helmholtz resonator formed by the cavity under the reed (which, incidentally is responsible for tuning being slightly different when the reed is on the reedblock to when it is off).  It's my feeling that this resonator, is a very important factor for the sound (though I'm not sure I could put a percentage on it).  If it were a purely acoustical effect then the shape of the cavity wouldn't matter as the wavelengths are much too long to 'see' the shape; since it does there must be some interaction with the airflow but I don't think it's well understood yet.

Helmholtz resonator - wow.  Now if this is as scientific as it sounds it must have formulae to predict it's optimum performance for each wave length and even assist tuners in deciding how much difference it is going to make to a reed tuned outside the box.  Armed with such knowledge one should be able to design the 'perfect' sound chamber for each note and by extention a whole instrument.  I appreciate the wavelength comment but exact divisions usually work ie. ¼, ½, ¾ the wavelength etc.

PS. There is already a way of predicting the difference between in box and out of box tuning

I'm very open to being corrected on this, but taking it somewhat to the extreme,
1) wouldn't that change the whole character of the instrument in that you'd be moving to reeds "coupled" to an air resonance rather than free reeds.  (Is the cheng a bit of a hybrid in this respect?)
2) the volumes involved if you're going to reinforce the fundamental of each note are considerable, whether you use air columns ("whistle" or oboe like) or helmholz resonators (ocarina like.) Your melodeon might end up looking like a double ended portative organ!

On the other hand is the "cassotto" a related and more realistic principle?
Tom

Logged

Waltham

  • Regular debater
  • ***
  • Offline Offline
  • Posts: 198
Re: Melodeon construction -- some more radical thoughts...
« Reply #49 on: December 19, 2008, 01:55:53 PM »

As I say, it's not a wavelength issue, the wavelength is the sound speed divided by the frequency, so at A above middle C that's 343/440, about 79 cms, far bigger than a reedblock cavity.  As for optimizing them, as far as the acoustical effects go I think that's already been done by trial and error.  I'm not suggesting they reinforce the fundamental, just that they have an effect, by the way.  What could possibly be improved is how the airflow through them interacts with the reed, particularly at the start of a note.  Clearly it makes a difference (Preciosa vs Lilliput, highest reeds being the mounted the 'wrong' way round etc.) but to know how best to shape them we'd have to either a) understand the flow patterns and particularly the vorticity around free reeds better than we do or b) do a lot of trial and error and get lucky from time to time.

Quote
PS. There is already a way of predicting the difference between in box and out of box tuning
I am indebted to you for bringing this to my attention.  Might I ask what it is?
Logged

IanD

  • Hero Member
  • *****
  • Offline Offline
  • Posts: 1703
  • Too many melodeons...
Re: Melodeon construction -- some more radical thoughts...
« Reply #50 on: December 19, 2008, 02:20:16 PM »

Each of these has to do with the Helmholtz resonator formed by the cavity under the reed (which, incidentally is responsible for tuning being slightly different when the reed is on the reedblock to when it is off).  It's my feeling that this resonator, is a very important factor for the sound (though I'm not sure I could put a percentage on it).  If it were a purely acoustical effect then the shape of the cavity wouldn't matter as the wavelengths are much too long to 'see' the shape; since it does there must be some interaction with the airflow but I don't think it's well understood yet.

Helmholtz resonator - wow.  Now if this is as scientific as it sounds it must have formulae to predict it's optimum performance for each wave length and even assist tuners in deciding how much difference it is going to make to a reed tuned outside the box.  Armed with such knowledge one should be able to design the 'perfect' sound chamber for each note and by extention a whole instrument.  I appreciate the wavelength comment but exact divisions usually work ie. ¼, ½, ¾ the wavelength etc.

PS. There is already a way of predicting the difference between in box and out of box tuning

There are indeed formulae, they're exactly the same as those for reflex speaker boxes. It's a classic mass-spring resonator, the mass is the air in the holes and the spring is the volume of air inside the box.

The big question is,  do you actually want a resonance,and if so at what frequency do you want the resonance to be?
-- output is boosted at the resonant frequency (clue -- honk honk) and rolls above above it (muffled treble)

If you want to hear the effect of this in a melodeon or accordion it's quite simple -- compare the sound of the bass end with and without the end plate (with sound holes in) attached.

The air in the sound holes does form a Helmholtz resonator with the volume of air inside the bass end, and the resonant frequency is usually in the 500Hz-1KHz region (you can calculate this or measure it, I've done both). The frequency of the resonance decreases as the volume of air inside the bass end goes up, and also as the total area of the holes becomes smaller; if the holes are too big in a small melodeon the resonance moves further up into the midrange and you get a honky/nasal bass end sound.

With the end plate on you usually have a noticeable "honk" (nasal tone) to the bass end sound, with the high buzzy harmonics of the reed cut out.

With the end plate off the "honk" disappears and you get much more treble from the reeds -- possibly too much if it then covers up the treble end sound.

I very much doubt that this is what is really wanted on the treble end, at least for most people. Anyway, the volume of air inside the reed chambers is so small compared to the size of the hole that the Helmholtz resonance will be at a pretty high frequency. The resonances along the length of the reed chamber are also at very high frequencies (a 3cm chamber is one wavelength long at about 10kHz). With a cassotto both the volume and length of the chambers are greatly increased so the resonant frequency comes down to a few kHz, hence the more mellow less bright sound.

The Helmholtz resonances aren't what causes the change in pitch when reeds are bit inside the box (or on the block), it's the change in acoustic impedance seen by the reed -- basically, the reed resonant frequency is slightly affected by the mass of air that it drives, which is affected by how much sound energy is radiated into the air, which is different when the reed is inside or outside the box/block. If the reed is louder (at the fundamental frequency) then the acoustic impedance will rise (more air mass loading on the reed) and the reed frequency will drop, and vice versa if it's quieter. This is exactly what happens with a speaker compression driver on a horn.

The amount of the shift is variable because it depends on the stiffness and mass and area of the reed and the amount of air loading added when placed on the block or in the box -- I would say unpredictable because it seems that way, though it would be perfectly predictable if you knew what all the numbers were for each reed to plug into the equations.

For a typically constructed box (reeds/frame/blocks/case) I expect there's a rule-of-thumb which predicts the frequency shift; however if you changed the construction significantly (reed/frame/block sizes/shapes/thicknesses) from "typical" this would no longer work.

Ian
« Last Edit: December 19, 2008, 02:22:26 PM by IanD »
Logged
Oakwood Model 4 D/G, Castagnari Dony D/G/#, Castagnari Tommy G/C, Baffetti Binci D/G, Hohner Preciosa D/G, Melos Bb/Eb, Lightwave SL5 and Kala California fretless basses

Quebecois

  • Hero Member
  • *****
  • Offline Offline
  • Posts: 1206
  • Accordez-vous donc, c'est si beau, l'accordéon!
Re: Melodeon construction -- some more radical thoughts...
« Reply #51 on: December 19, 2008, 02:24:28 PM »

Wow, do you really need to know this stuff in order to determine if you like the sound of an old Hohner ???
Logged
Hohner Morgane D/G, pre-Erica Hohner in C/F and G/C,
Hohner Erica A/D, Roland FR-18

IanD

  • Hero Member
  • *****
  • Offline Offline
  • Posts: 1703
  • Too many melodeons...
Re: Melodeon construction -- some more radical thoughts...
« Reply #52 on: December 19, 2008, 02:31:10 PM »

As I say, it's not a wavelength issue, the wavelength is the sound speed divided by the frequency, so at A above middle C that's 343/440, about 79 cms, far bigger than a reedblock cavity.  As for optimizing them, as far as the acoustical effects go I think that's already been done by trial and error.  I'm not suggesting they reinforce the fundamental, just that they have an effect, by the way.  What could possibly be improved is how the airflow through them interacts with the reed, particularly at the start of a note.  Clearly it makes a difference (Preciosa vs Lilliput, highest reeds being the mounted the 'wrong' way round etc.) but to know how best to shape them we'd have to either a) understand the flow patterns and particularly the vorticity around free reeds better than we do or b) do a lot of trial and error and get lucky from time to time.

Quote
PS. There is already a way of predicting the difference between in box and out of box tuning
I am indebted to you for bringing this to my attention.  Might I ask what it is?

Unfortunately analysing nonlinear flow patterns round a vibrating reed is a massively difficult computing problem, and I'm not sure that the theory actually exists to do it properly -- it's a similar problem to things like aileron flutter in planes and vortex shedding, there have been recent advances in theory but they're still not there yet, and simulation takes a very long time right now even on the fastest supercomputers. If you want to simulate the start-up the problem is many times worse because this takes many cycles of oscillation to stabilise, so you can't even use pseudo-steady-state analysis methods.

Maybe in the future (10 years?) if Moore's Law holds we'll be able to do this, for now I guess trial and error and experience is the way forward.

Ian
Logged
Oakwood Model 4 D/G, Castagnari Dony D/G/#, Castagnari Tommy G/C, Baffetti Binci D/G, Hohner Preciosa D/G, Melos Bb/Eb, Lightwave SL5 and Kala California fretless basses

IanD

  • Hero Member
  • *****
  • Offline Offline
  • Posts: 1703
  • Too many melodeons...
Re: Melodeon construction -- some more radical thoughts...
« Reply #53 on: December 19, 2008, 02:32:45 PM »

Wow, do you really need to know this stuff in order to determine if you like the sound of an old Hohner ???


Not at all. But you do if you want to know how to change a design to improve it without relying on guesswork or serendipity :-)

Ian
Logged
Oakwood Model 4 D/G, Castagnari Dony D/G/#, Castagnari Tommy G/C, Baffetti Binci D/G, Hohner Preciosa D/G, Melos Bb/Eb, Lightwave SL5 and Kala California fretless basses

Quebecois

  • Hero Member
  • *****
  • Offline Offline
  • Posts: 1206
  • Accordez-vous donc, c'est si beau, l'accordéon!
Re: Melodeon construction -- some more radical thoughts...
« Reply #54 on: December 19, 2008, 03:03:51 PM »

Wow, do you really need to know this stuff in order to determine if you like the sound of an old Hohner ???
Not at all. But you do if you want to know how to change a design to improve it without relying on guesswork or serendipity :-)
Ian
Sometimes guesswork and inspiration give the best results! :)
Logged
Hohner Morgane D/G, pre-Erica Hohner in C/F and G/C,
Hohner Erica A/D, Roland FR-18

Waltham

  • Regular debater
  • ***
  • Offline Offline
  • Posts: 198
Re: Melodeon construction -- some more radical thoughts...
« Reply #55 on: December 19, 2008, 03:26:49 PM »

The Helmholtz resonances aren't what causes the change in pitch when reeds are bit inside the box (or on the block), it's the change in acoustic impedance seen by the reed -- basically, the reed resonant frequency is slightly affected by the mass of air that it drives, which is affected by how much sound energy is radiated into the air, which is different when the reed is inside or outside the box/block.
Indeed so, but I rather carefully avoided saying it was the Helmholtz resonance that caused the shift, but the Helmholtz resonator, that is to say the cavity coupled to the reed, as I say it doesn't operate at its resonance frequency.  So one can think of it in terms of effective mass if one prefers, it's just a matter of terminology.  I'm not as convinced by the energy argument, wouldn't that imply a nonlinear effect since energy is quadratic in amplitude? and yet the tuning change can be observed at the lowest amplitudes, and in the absence of flow by twanging the reed. 

It's interesting to me that even the Hohner pokerwork 1-rows with their flat-mounted reeds still have a cavity beneath them (thanks to Theo for pointing this out a while ago, at a time when I hadn't taken mine apart).  It might be interesting to make an instrument with as little cavity coupling as possible: reeds mounted flat on the [not-a-]soundboard, holes almost as big as the reedplate.  I wonder what that would sound like.

I'm not as skeptical as IanD  about the possibility of computing the coupled flow-vibration problem.  The things that make this challenging in aeroelasticity are high Reynolds number, high Mach number and the need to resolve fine length-scale vorticity in turbulent regions as well as a few other things, none of which should be too bad in a free reed.  I suspect that a small PC cluster would be up to the job, it would make a nice self-funded PhD for someone.  In any case, there are other approaches besides 'brute force' computation, for example that taken in

Ricot D, Causse R, Misdariis N, Aerodynamic excitation and sound production of blown-closed free reeds without acoustic coupling: The example of the accordion reed, J. Acoust Soc. Am. 117(4) 2279-2290 (2005). 

Sorry I can't find a public domain version but here's the abstract:
Quote
The accordion reed is an example of a blown-closed free reed. Unlike most oscillating valves in wind musical instruments, self-sustained oscillations occur without acoustic coupling. Flow visualizations and measurements in water show that the flow can be supposed incompressible and potential. A model is developed and the solution is calculated in the time domain. The excitation force is found to be associated with the inertial load of the unsteady flow through the reed gaps. Inertial effect leads to velocity fluctuations in the reed opening and then to an unsteady Bernoulli force. A pressure component generated by the local reciprocal air movement around the reed is added to the modeled aerodynamic excitation pressure. Since the model is two-dimensional, only qualitative comparisons with air flow measurements are possible. The agreement between the simulated pressure waveforms and measured pressure in the very near-field of the reed is reasonable. In addition, an aeroacoustic model using the permeable Ffowcs Williams-Hawkings integral method is presented. The integral expressions of the far-field acoustic pressure are also computed in the time domain. In agreement with experimental data, the sound is found to be dominated by the dipolar source associated by the strong momentum fluctuations of the flow through the reed gaps.
I'm not saying I agree with their approach, but it's not raw computation.
Logged

IanD

  • Hero Member
  • *****
  • Offline Offline
  • Posts: 1703
  • Too many melodeons...
Re: Melodeon construction -- some more radical thoughts...
« Reply #56 on: December 19, 2008, 03:27:38 PM »

Wow, do you really need to know this stuff in order to determine if you like the sound of an old Hohner ???
Not at all. But you do if you want to know how to change a design to improve it without relying on guesswork or serendipity :-)
Ian
Sometimes guesswork and inspiration give the best results! :)


And sometimes they waste an awful lot of time and go nowhere :-)

It's always useful having a map to tell you where to go instead of blundering around in the dark...

Ian
Logged
Oakwood Model 4 D/G, Castagnari Dony D/G/#, Castagnari Tommy G/C, Baffetti Binci D/G, Hohner Preciosa D/G, Melos Bb/Eb, Lightwave SL5 and Kala California fretless basses

risto

  • Respected Sage
  • ****
  • Offline Offline
  • Posts: 317
Re: Melodeon construction -- some more radical thoughts...
« Reply #57 on: December 19, 2008, 03:38:28 PM »


...Charles Taylor used to give a lecture in which he'd play a clip from a classical piece where a whole orchestra was playing and then all stopped apart from one instrument, and he'd ask who thought it was a french horn and who thought it was an oboe and so on, there'd be no clear consensus.

Haven't heard of that before, but quickly thinking, sounds more like a question of how the human ear adapts itself to sudden changes and how it recovers back to normal. Seems as a previous "false" information can distract the ear for some time. In any case trying to find the Mother of all Melodeon Sounds is something I would leave to the Sirens in the Greek mythology.

Quote
... Each of these has to do with the Helmholtz resonator formed by the cavity under the reed (which, incidentally is responsible for tuning being slightly different when the reed is on the reedblock to when it is off).  It's my feeling that this resonator, is a very important factor for the sound ....

Beause of the very small volume of the reed block cavities and large hole sizes my guess is that Helmholz resonance is a minor factor in accordion type isntruments, at least compared to its effect in instruments with considerably larger air reservoirs. I leave this to anyone to prove otherwise, in any case the frequencies produced by the Helmholz effect are only a part of the overall frequency spectrum.
Logged
Paolo Soprani 5-row, Roland FR1B-digital CBA button accordion. Guitar, 5-string OT banjo, mandolin, piano...
My one man band recording: https://apari.fi/Bonaparte_Crossing_The_Rhine_Risto.mp3

IanD

  • Hero Member
  • *****
  • Offline Offline
  • Posts: 1703
  • Too many melodeons...
Re: Melodeon construction -- some more radical thoughts...
« Reply #58 on: December 19, 2008, 03:53:14 PM »

The Helmholtz resonances aren't what causes the change in pitch when reeds are bit inside the box (or on the block), it's the change in acoustic impedance seen by the reed -- basically, the reed resonant frequency is slightly affected by the mass of air that it drives, which is affected by how much sound energy is radiated into the air, which is different when the reed is inside or outside the box/block.
Indeed so, but I rather carefully avoided saying it was the Helmholtz resonance that caused the shift, but the Helmholtz resonator, that is to say the cavity coupled to the reed, as I say it doesn't operate at its resonance frequency.  So one can think of it in terms of effective mass if one prefers, it's just a matter of terminology.  I'm not as convinced by the energy argument, wouldn't that imply a nonlinear effect since energy is quadratic in amplitude? and yet the tuning change can be observed at the lowest amplitudes, and in the absence of flow by twanging the reed. 

It's interesting to me that even the Hohner pokerwork 1-rows with their flat-mounted reeds still have a cavity beneath them (thanks to Theo for pointing this out a while ago, at a time when I hadn't taken mine apart).  It might be interesting to make an instrument with as little cavity coupling as possible: reeds mounted flat on the [not-a-]soundboard, holes almost as big as the reedplate.  I wonder what that would sound like.

I'm not as skeptical as IanD  about the possibility of computing the coupled flow-vibration problem.  The things that make this challenging in aeroelasticity are high Reynolds number, high Mach number and the need to resolve fine length-scale vorticity in turbulent regions as well as a few other things, none of which should be too bad in a free reed.  I suspect that a small PC cluster would be up to the job, it would make a nice self-funded PhD for someone.  In any case, there are other approaches besides 'brute force' computation, for example that taken in

Ricot D, Causse R, Misdariis N, Aerodynamic excitation and sound production of blown-closed free reeds without acoustic coupling: The example of the accordion reed, J. Acoust Soc. Am. 117(4) 2279-2290 (2005). 

Sorry I can't find a public domain version but here's the abstract:
Quote
The accordion reed is an example of a blown-closed free reed. Unlike most oscillating valves in wind musical instruments, self-sustained oscillations occur without acoustic coupling. Flow visualizations and measurements in water show that the flow can be supposed incompressible and potential. A model is developed and the solution is calculated in the time domain. The excitation force is found to be associated with the inertial load of the unsteady flow through the reed gaps. Inertial effect leads to velocity fluctuations in the reed opening and then to an unsteady Bernoulli force. A pressure component generated by the local reciprocal air movement around the reed is added to the modeled aerodynamic excitation pressure. Since the model is two-dimensional, only qualitative comparisons with air flow measurements are possible. The agreement between the simulated pressure waveforms and measured pressure in the very near-field of the reed is reasonable. In addition, an aeroacoustic model using the permeable Ffowcs Williams-Hawkings integral method is presented. The integral expressions of the far-field acoustic pressure are also computed in the time domain. In agreement with experimental data, the sound is found to be dominated by the dipolar source associated by the strong momentum fluctuations of the flow through the reed gaps.
I'm not saying I agree with their approach, but it's not raw computation.

Even without any resonator at all you still get a frequency shift due to the air mass loading -- compare a reed just in a frame to this mounted in the middle of a flat board (no cavities either side). This is the same effect as a speaker in free air (little sound) compared to mounted in an infinite baffle (big flat plate), the resonant frequency drops as the sound output increases due to air loading (change in acoustic impedance). A resonator -- even driven away from resonance -- also changes the acoustic impedance, as do other things nearby due to reflection of sound energy (actually, anything that changes the acoustical or mechanical impedance seen by the reed could cause a small frequency shift).

As you say, simulating it would make a very nice PhD for someone -- but I do suspect that a couple of years solid work and a *lot* of computing time (see below) are needed, which is not something than any of us (or an instrument maker) is likely to be able to spare.

The simulations you referenced are (I think -- correct me if I'm wrong) a pseudo-steady-state type of analysis, similar to harmonic balance simulations in RF circuits, which shows the settled state of the system -- we do the same kind of analysis for IC design, and this is typically 100x faster than a transient analysis which is what you'd need to simulate the start-up behaviour. Add on the need to analyse say 10 different reeds across the pitch range and you're looking at 1000x more computation; now assume 10 design iterations to do your tweaking/improvement and you have 10000x more computation -- which is why I guessed we might be 10 years away from this being feasible in a reasonable time.

Your idea of flat-mounted cavityless reeds with very big pallet holes will come close in sound to the "reeds-out" idea, the concept is similar and it might be easier to try out quickly to see what it sounds like (in case we hate the result!). The difficulty in a real box is getting enough area to fit all the reeds in; even with the Preciosa (2 reeds) Hohner had to resort to "stacked" reeds with cavities and not-so-large pallets, with 3 reeds (or 3 rows) this would be well-nigh impossible.

Ian
Logged
Oakwood Model 4 D/G, Castagnari Dony D/G/#, Castagnari Tommy G/C, Baffetti Binci D/G, Hohner Preciosa D/G, Melos Bb/Eb, Lightwave SL5 and Kala California fretless basses

Waltham

  • Regular debater
  • ***
  • Offline Offline
  • Posts: 198
Re: Melodeon construction -- some more radical thoughts...
« Reply #59 on: December 19, 2008, 04:30:13 PM »

Even without any resonator at all you still get a frequency shift due to the air mass loading -- compare a reed just in a frame to this mounted in the middle of a flat board (no cavities either side). This is the same effect as a speaker in free air (little sound) compared to mounted in an infinite baffle (big flat plate), the resonant frequency drops as the sound output increases due to air loading (change in acoustic impedance). A resonator -- even driven away from resonance -- also changes the acoustic impedance, as do other things nearby due to reflection of sound energy (actually, anything that changes the acoustical or mechanical impedance seen by the reed could cause a small frequency shift).
I really think we're talking at cross-purposes here.  Perhaps it would help if I re-stated my position as follows:
  • The frequency shift is caused by a change in the impedance seen by the reed.
  • A major component of this impedance is due to the reed being coupled to a cavity, inside the reedblock in most instruments.
  • This cavity is a Helmholtz resonance, although it is not operating at its resonance frequency in most cases.

Quote
As you say, simulating it would make a very nice PhD for someone -- but I do suspect that a couple of years solid work and a *lot* of computing time (see below) are needed, which is not something than any of us (or an instrument maker) is likely to be able to spare.
Well, the joy of long computing jobs is that one can do other things while they're running,  and it's far from uncommon for Musical Acoustics PhDs to be done part-time.  But if anyone were to try this they'd have to be doing it because they really wanted to, rather than any dramatic increase in commercial gain, fame, sex appeal, etc...

Quote
The simulations you referenced are (I think -- correct me if I'm wrong) a pseudo-steady-state type of analysis, similar to harmonic balance simulations in RF circuits, which shows the settled state of the system -- we do the same kind of analysis for IC design, and this is typically 100x faster than a transient analysis which is what you'd need to simulate the start-up behaviour. Add on the need to analyse say 10 different reeds across the pitch range and you're looking at 1000x more computation; now assume 10 design iterations to do your tweaking/improvement and you have 10000x more computation -- which is why I guessed we might be 10 years away from this being feasible in a reasonable time.
Yes they were steady state (though not directly computational) and yes the transient solution will take longer, but in flow acoustics time-domain methods are creeping up on frequency domain ones and I'd conjecture that the factor wouldn't be as big as the one you experience in your line.  And my thinking was that seeing the results from just one simulation of a typical reed might tell us enough about what the air and the reed do to each other to give us some ideas about changing cavity shapes that could be tested directly on the analogue computer we call a melodeon.

Quote
Your idea of flat-mounted cavityless reeds with very big pallet holes will come close in sound to the "reeds-out" idea, the concept is similar and it might be easier to try out quickly to see what it sounds like (in case we hate the result!).
In that case I fear I've misunderstood your 'reeds-out' idea; are no reedblocks involved? 
Logged
Pages: 1 2 [3] 4 5 6   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