Jump to content

Łukasz Martynowicz

  • Posts

  • Joined

  • Last visited

Posts posted by Łukasz Martynowicz

  1. @ rlgph: Yes, I have been reading all your posts carefully, and I'm trying as much as I can to point out to you where you make some false assumptions on how physics of free reed works...


    I have simply been trying to understand why not, from the point of view of physics. In order for oscillations to occur, the reed must be subject to time-varying forces. The only things that can exert forces on the reed are air molecules (external force) and the molecules that make up the reed (internal force). (Gravity is too small to have any significant effect.) My current guess as to why the bent reed works is that the air flow around the end of the read while the reed is still open on the high pressure side causes the air on that side to exert a significantly larger force than in the flat reed case, probably due to turbulence on the underside created as the air flows around the asymmetric reed configuration.



    One other thing, which is relevant to both this thread and our second physics discussion, which may lead you to understand free-reed physics and wasn't stressed enough previously: air pressure force acting on the reed is much lower than needed to bend tongue to it's maximum position - internal spring force is much greater than maximum bellows pressure force. It is the effect of resonant amplification of small tongue movements that makes this amplitude so large. The pressure in the bellows is high enough only to bend a plastic/leather valve to the roughly same order of magnitude of displacement as an oscilating metal tongue.


    If you would make an experiment with a typical reed and measure tongue displacement, then your physics is a good description of what can be measured with a pressure acting in the silent direction of the reed. The tongue would bend a little, releasing high pressure - the higher pressure the more tongue displacement - but this displacement would be measured in fractions of a milimeter. And all forces would be equalised at all times, there is no need for aditional "missing time varying force" here.


    With sounding airflow direction, what happens is this (in as elaborate form as I can think of): at first, when you squeeze the bellows, the air starts to flow through the gap. The suction effect of the gap is signifficant enough to move the tongue a little towards the shoe. Then when there is no more airflow (when reed sinks a bit into the shoe) and suction effect stops, the internal spring force (which is greater than pressure force acting on the tongue) is accelerating the tongue back towards it's resting position, but due to momentum it swings a bit higher than resting position. At this point the suction effect starts again, and acts towards drawing a tongue into the shoe, but this time there is a bit of added momentum from the spring force going back from it's highest position. A cycle closes, but with a bit of added energy. With each cycle the swinging motion increase a bit and after a few cycles the tongue goes through to the other side of the shoe, at which point airflow increases, making each cycle even more influential and within a few cycles the tongue achieves a stable oscilation, proportional to air pressure in the bellows, which now translates to volume (an oscilation amplitude) and frequency.


    If you want to learn more about the strenght of suction effect in fluid mechanics, then read (and watch) about suction effect of ships passing each other in harbours as this is the most visual illustration of this effect. Or you can take two sheets of paper, hold them down close to eachother and blow gently between them and observe how they will be drawn to each other by a dynamic airflow.



    When the switch is turned on in an electrical circuit the light bulb does not have to wait for electrons to travel from the switch to the bulb for the light to come on -- fortunately, because electron drift velocity is very slow in comparison to the speed of the electric field change (which is at the speed of EM waves within the wire). Rather, electrons that are already in the filament are directly affected by the electric field gradient (voltage), and heat generated by the resistance to their movement causes the filament to glow.

    Not entirely true. While the filament doesn't have to wait for individual electrons to travel to it from the switch, the change in the electric field gradient isn't instantaneous. It may seem so to us, but it can't be faster than the speed of "light" (electromagnetic waves) in the wire.


    In the same way, the response of of a reed to a pressure gradient does not depend on air physically moving through the whole pathway, which is quite slow in comparison to the speed of the pressure changes. Air molecules already at the reed site are those responsible for affecting the reed when the pressure field changes.

    But as with the electromagnetic wave above, the pressure change is transmitted at a finite speed... the speed of sound. This happens through collisions between molecules, even though for individual molecules to travel the full distance takes a much greater time, if it happens at all.


    Of course you are right that changes in some aspect of the design of a musical instrument can have substantial effects on the instrument's performance. But that doesn't mean that an experienced or inventive instrument maker can't make other changes to compensate.

    Nor is there any guarantee that they can... or even that if such a solution does exist, it's easy to find or construct.



    Jim is right, but this is even more complicated than that, because...


    …before there is sound produced by the reed, which then travels at a speed of sound, because it is pressure gradient change, there is need for air movement at "wind speeds". When you press your bellows with air button pressed, you aren't struck by a sonic boom, just a mild wind blow. This is this "wind speed air movement" which starts reed oscilation by a transfer of momentum between air molecules and reed tongue and suction effect of the gap and then feeds reed resonance for a stable oscilation and nice concertina sound.

  3. Beside the reeds themselves, the most time consuming parts of the concertina (and hardest to do in an amateur workshop) are levers and buttons - even some proffesionals outsorce lever making to a laser cutting workshop.


    From my experience, any wooden part can be home made with satisfactory quality and tolerances with reasonably cheap equipment and enough patience. But metal working requires heavier equipment, heat generation handling, better tools with tighter tolerances which are usually more expensive etc… And don't forget, that there are dozens of levers and buttons in a concertina, and those sheer numbers make making them very time consuming. My 66 buttons took me the same amount of time, as the whole endboxes and reedpans (without the fretwork cutting, which is another time consuming part, but depends heavily on fretwork design and material used).

  4. It is not the reeds that rattle, but the unbushed buttons of your Jackie. And this was so annoying for me after purchase of my Elise, that I have modified it within two weeks of delivery…


    Unfortunately there is not much you can do about it except for entirely replacing the buttons. The problem is caused by those thin and light metal strips that form the base of the buttons, which tend to vibrate on the levers due to air vibration and rattle on the levers. This is a big design flaw of concertina-connection entry level instruments, strangely happens only to some batches of those instruments or only some of buyers find it annoying. But it is definately button-related, as it has vanished completely after button replacement.

    [A quick edit] What you could try to do besides replacing buttons is to cover the end part of levers with some sort of silicone paint - a layer thin enough to easily go through a hole in button posts, but thick enough to dampen some of the vibration. Or you could try to bush only endplate holes with a thin felt, but this may require redrilling the holes to a bit larger diameter - the felt bushing holds buttons a bit and this may be enough to dampen any unwanted vibration.
  5. By "static" I don't mean stationary, but "equalized at all times" - in this case, in your approach, high pressure is always compensated by internal spring force. It does not mean that the tongue does not bend, only that forces are equalised in any given "time slice". I may use the wrong english word to describe it. "Dynamic" in my comments mean that the process does not equalize in such manner, and there is no such "missing force" in any given "time slice", but you have to consider the whole temporal evolution of a system consisting of high pressure reservoire, tongue spring, gap suction, airflow etc...


    The reed is not a physical pendulum, it is a spring oscilator. The "missing forces" (if we must name it as such) that you choose to neglect are dynamic effects of gap suction, initial airflow increase, pressure gradient "momentum" and resonant oscilation effects in the first cycles of the reed movement and I have given you a very extensive and detailed description of that process.

    And I'm not debating on how the reed work, I'm refering to you a well established knowledge, which for some reason you refuse to acknowledge… And you have even mentioned earier, that you have read an extensive essay on the matter on concertina-connection site, but you still refuse to agree, that symmetrical or flat reed simply won't work...


    So I have one question for you: Your description does not differentiate sides and you assume that flat reed would work. In your model, what is the difference between the physics of sounding and silent direction of airflow through traditional reed? If there is none, and gap and asymetric thickness is irrelevant and flat reed should work, then why the reed sounds only in one direction and flat reed does not speak at all?


    As i said before, the physics that i have described is correct. However, my description of the operation of a reed under the influence of a perpendicular pressure gradient does not depend on whether there is an initial gap or not.




    Then it is not correct, as the gap and asymetry are crucial in free-reed, accordion/concertina like instruments which is proven by almost 200 years of experimentation and you can replicate this in a 1 minute experiment involving a single reed and pair of lungs…




    It will only find a stable position if the pressure difference does not change, so that the force due to the air pressure exactly balances the restoring force internal to the spring. In this case, however, air is flowing around the end of the reed causing a decrease in the pressure difference. At the maximum downward swing, the restoring force is greatest, and the force due to air pressure difference is getting less due to the air flow. The net force is toward the equilibrium position and the reed acelerates in that direction. And that is true regardless of whether in equilibrium the reed is bent upward or not.




    You are describing a spring pressure valve here, not a free reed oscilator. What you are (correctly) describing here is the physical process involved in a static tongue bending caused by pressure gradient and air leaking through an unvalved reed in it's silent direction, until all bellows is compressed and there is no more higher pressure reservoire. At which point (and just then) the net force you describe is in fact accelerating the tongue back to it's resting position with some tiny amount of oscilation at the end of it's movement.


    The process involved in producing sound by a free reed is entirely different, involves a gap and asymetry of the reed, and I have described it earlier. If you don't see your mistake at this point I have no more ideas on how to point it out to you...



    Thanks for your thought provoking comments.



    You're welcome :)

  7. Reeds from Wim have (at least had few years back, when I wrote to him about it) separate waiting list of a few weeks only. But they do cost more than a half of a finished Wakker concertina...


    One other thing - why we assume, that exchangeable reedpans mean "two reedpans with a full set of reeds each"? If we just want a non-destructive conversion it can be perfectly done with a single set of reeds that covers both layouts and two sets of wooden parts. There will be of course need for tuning after each conversion, and some form of "placeholder reeds" for keeping reedpan dimensions stable, but I really cannot imagine, that one would want to swap reedpans on regular basis, as even with separate reed sets it takes too much time to swap "on the fly" at a gig.


    And personally I don't understand a need for playing two different duet systems, other than different sound or range of available instruments, which makes such "dilution of practice hours" a necessity.

  8. Sory, you have mislead me into believieng you have it backwards by this sentence: "As i understand it, the slight bend in the reed (toward the lower pressure side) allows air to flow around the end." as it is exactly oposite to reality.



    My description of the physics is right (though turbulence, which i've ignored here, may play some role) -- a flat reed will oscillate when subjected to a pressure gradient set up perpendicular to the reed by a push or pull of the bellows, and the oscillation will cause periodic cutting of the air flow. What i don't understand is why this cannot be used to make an acceptable concertina sound.




    Because it does not work like you describe… A reed without a gap does not speak, flat reed mounted perpendicular does not speak.


    Perpendicular reed needs to be assymetric and require a gap and nothing can be done about it. It is just how reed physics works. Please make a simple experiment - take one of your bass reeds, bend it back flat into the shoe so there is no gap and try to make a sound. It won't speak at all and will be just poor pressure valve. [bass one is best for this purpose, because high tone reeds require only a slight gap to start, so even when flatted out, the tolerance around the tongue is a gap sufficient to start the movement at low pressure levels]



    The more it moves down, the lesser the pressure difference and the greater the (internal) restoring force of the spring, causing the motion of the reed to stop, then accelerate back toward the equilibrium position.



    This is the point where you get this wrong - as long as you have pressure it won't accelerate back, it will just stop bending more and will find stable bent position depending on the pressure gradient.


    In more elaborate form: in the flat, gap-less reed the pressure is just released as through valve - spring just bends proportionally to the pressure pushing on it [and because tongue thickness is so small it happens with very minute pressure gradient levels]. As long as you squeeze your bellows with same force, the pressure inside is constant, the reed/valve opens and air just flow outside with a rate depending on reed stifness. This is exactly what happens when you blow air through reed in oposite direction - it just leaks air, it does not oscilate and in this case the initial gap seting is irrelevant (it can even be "negative", i.e you could set the tongue to be initially inside a shoe, this will just slightly increase "releasing pressure"). Of course there is some oscilation when you release the force on the bellows, but this is not resonant oscilation, just a depleting momentum of an elastic valve returning to equilibrium, making no audible sound at all.


    One other thing: pressure gradient changes are not instantaneous. When you press your bellows or open the pad on the airhole it takes some small amount of time to start the airflow. This is the moment when reed oscilation "jump starts" through gap mechanism on pressure levels lower than needed for stable oscilation, so when that pressure is achieved the reed is already in it's swinging motion (the gap is a leak which makes effective pressure gradient lower at the start of the motion than when the tongue is inside the shoe and blocks the airflow). With fast enough, high pressure pump instantly and constantly feeding high pressure air through a reed you can choke even the properly set assimetric concertina reed and effectively turn it into a mute pressure valve. This is one of the reasons why different free-reed instruments require different reeds depending on operating pressure levels, as this can even occure on my Elise lowest bass note, when I squeeze a bellows the hardest I can, as it has accordion reeds, designed for lower pressure levels.

  9. As i understand it, the slight bend in the reed (toward the lower pressure side) allows air to flow around the end.



    You think it backwards… Reed tongues are bent to the higher pressure side: pressure is working towards closing the reed (pushes the tongue inside the shoe, not away from) and is released when tongue reaches the other side of the shoe. Then the energy acumulated in the tongue swings tongue back against the airflow and opens the reed for half a cycle, when another portion of air draws the tongue back into the shoe with more force. And this initial gap works as a miniature version of a bladeless fan - it is essential to draw the tongue into the slot with initial pressure lower than needed to hold the tongue on the other side permanently. It is dynamic not static process and takes a few cycles to achieve resonant and stable vibration. Without this gap (and in the oposite direction) a reed tongue works as a pressure valve which won't vibrate - it will just open in a static way when pressure gradient is high enough to bend it and let the air out from the higher pressure reservoire.


    There are asian type free reeds, which are flat as the one you describe, but they work on a different principle and need a pipe resonator to sound and are mounted parallel to the airflow direction, not perpendicular as in concertinas or accordions.

  10. I do think, however, that your implication "seem to forget..." is a little bit unfair.



    Don't get me wrong, I was not intending to insult you or Jake in any way. In fact, I had some similiar thoughts about "pneumatically improving" concertina design when I first started designing my DIY Hayden. That was until I made this very experiment I have described earlier and realised, that sound produced by the reed is a very fragile thing and everything affects it, usually in the least desirable way. There are numerous long threads here on concertina.net about such minute changes to instrumenst as beveling airholes on chambers that are placed not on the edge of the reedpan or have additional 5mm of action board "in the way"of the sound. My "seem to forget" coment was only meant to point out, that concertina design should be s"ound-centric" and not "pneumatic-centric" and that many (otherwise great or intiguing) engeneering ideas simply do not apply to concertinas.


    I can see the last point (obstructing with the innards of a resonance chamber), thanks for pointing it out! But I don't understand your first point at all; why would the reed chambers need to be huge?



    This one is more true with hybrid concertinas and accordion reeds, but applies to a proper concertina reeded instrument also - you have to place your pad assembly out of the way of swinging reed and inner reed valve. Normally you can have your air hole directly above the reed (flat mounted accordion reeds have chambers rougly the lenght of a reed shoe, concertina reeds chambers can be and often are extended for acoustic reasons). In your design you have to place the button assembly "out of the way" - you cannot place your pad and spring over the reed.


    In my opinion, both yours and Jake's designs are valid only from a pneumatical point of view. They will both produce controlled pressure and airflow to feed the reed. But both of you seem to forget, that it is the musical instrument that you're working on. Every bounce of sound matters, every cubic inch of confined space that vibrating air has to travel inside a concertina will alter the sound. Make a simple experiment: build a variable volume reed chamber (a moveable one side, like a piston), apply airflow through a reed mounted to it and listen to the sound while moving a piston. And measure response time of the reed depending on volume (cubic space not loudness :D). Then put a tube (either straight or bent like in Jakes design) on the airhole and listen how sound produced by such setup changes drastically - depending on materials used the sound would be more or less muffled and altered in timbre. In Jake's design the amount of distance that the air has to travel inside a tube will decrease the loudness significantly, and the reeds will be very slow to speak because the sheer amount of air trapped between stationary reed and button, that needs to be moved/compressed before the reed starts to speak…



  12. There are so many problems with this design I can think of, that I don't even know where to start...


    First of all, such design seems valid only for linear keyboards, as you cannot freely move airholes around and you have quite space-demanding design...


    Then, your reed chambers would be HUGE, as you must effectively make them longer by a diameter of your air hole. Otherwise your pad will obstruct inner reed valve. This raises a numerous problems with reed response times and tone balance, as you will have to increase each chamber lenght by a fixed amount, thus for smaller reeds probably doubling the chamber lenght… And you create a LOT of unwanted bouncing of sound and obstruct chamber resonance properties with this "inner pad". And resulting sound will probably change substantially with button travel, as you're moving parts inside of a resonant chamber...


    Another problem with this design is a small effective airhole: with normal setup, you need only short pad travel, because you have a cylindrical gap between airhole and pad (in open position you need a height of quater of hole diameter to have equal flow areas). With this design you obstruct a lot of this cylinder with back and side walls of a chamber. Try to imagine a smoke traveling through reed and airhole in this setup to visualise what I'm thinking about.


    Last problem I can think of is the wobblines of this setup - in normal concertina construction, you have two guiding constraints for a button: an endplate hole and an actionboard hole. Together they make button motion almost linear. But with your design you only have endplate guide and a wobbly spring. You will have to either add an inner pole to guide a button and pad assembly or find another way to ensure perfect closing of airhole (like conical hole closure for example) and limit angular deviation from straight button travel.

  13. First of all, warmest welcome to another Hayden player! And now to the point...


    Chords on a Hayden duet and corresponding music theory is much easier than on non-isomorphic instruments. Virtually everything you should know to play chords on a Hayden is shown here: http://www.shiverware.com/musix/wicki/chords.html

    Those are chord diagrams on this keyboard. Since it is isomorphic, all chord types in every key have a single "shape". You must only know the root note and type of chord you want to play and that's it. You can READ the chord structure from the keyboard itself, so there is no point in chord wheels or similiar detailed chord charts…

    And when you'll learn chord types in your finger memory, it becomes natural to make your own accompaniments and countermelodies, as you'll be able to read harmony structure of a tune straight from melody line. Regardless of key, since the concept of a key on a Hayden is somewhat artificial and obsolete...

  14. This is very relevant and is close to what I have in mind form my DIY - in fact I have tested such wristrest back in at the very begining of my concertina playing (on anglo back then).

    They are very comfortable, but have several limitations and are tweaking ergonomics in a substantial way:

    - first of all, they prohibit wrist movement, which makes it unnatural to play in lower range on the EC and limits reach to high notes (especialy sharps) on the Hayden.

    - secondly, if done high enough for comfortable wrist position, you end up with completely different finger "attack angle" and different muscles are involved in fast retracting of fingers. It is less of a problem with EC, because it's "normal" ergonomics different finger positions and training of different muscles.

  15. @Jim: if such MIDI would be based on switches and not entire action (though it could be done in either case as there is no "reed placement imperative") it is possible to design "rotary endplates" to make variable slant. Even without such mechanism, Wicki version has no slant, so there is no reason why MIDI Wicki shouldn't have options for all possible orientation combinations (given symetrical number of physical buttons).

  16. Dave, I think that synth bass is a good idea and I have enjoyed your recording, but I think it could be executed in a better way. Twofold:


    - first of all, timing in multilayer rendition should be perfect, no matter what. With a single melody line, or even when playing melody&accompaniment on a single box, uneven rhytm or missed beats don't stand out as much as with multilayer recordings.


    - secondly, while bass line IMHO is good and fits this tune well, your rendition misses the "middle layer": you should add an accompaniment (drone or chordal harmonies, or a mix of both) inbetween your melody and bass line. Listen to what Tona is doing in his rendition of this very tune. He adds layers and changes them throughout the whole tune, to build up the tension and character, adding a "macro structure". Of course it requires great skill to play all such lines simultanously, but if you're already using postproduction mixing, why not go further with it and record a complete arrangement?

  17. Perhaps i am confused by what it means to play in parallel 3rds, 6ths, etc? My understanding is that, e.g., playing in parallet thirds, every note on the left side is a third above the corresponding melody note on the right side. For an idealized Hayden system this just means shifting your left hand a couple of buttons over, doesn't it? In which case you would be using the same corresponding finger on each hand. (Of course, the finite size of the button arrangement on a real concertina may make this not possible in some keys.)





    You have two 3rd intervals - major and minor and playing in parallel 3rds you'll have to switch them accordingly while moving up the scale, so each parallel pair stays in given key. It is easiest to learn on a single side of Hayden concertina - as you know single (major) octave on a Hayden is divided in two rows (3+4 butons). Playing in thirds means playing notes that are confined to this 3+4 shape and follow this rule: "you play a note and a third note that comes in sequence" [NOT "always third button to the right" - this will produce chromatic scale with four intervals being outside the key]. Geometrically speaking you wll move your fingers in two possible shapes - "C-E" and "D-F" that will fit in 3+4 octave.

    The same principle applies to any given parallel interval, so in both Wicki and Hayden layouts you'll have to move your fingers in non-mirrored way. Unless of course you play modern, chromatic music :)

  18. Because when playing a lot of drone accompaniment or using a lot of minor chords in oompah rhytms it is the fully depressed state that hurts my fingers most. I have my middle finger significantly longer than index and third, so playing minor chords on a Hayden makes me back bend my middle finger and press he button with a fingertip near the nail instead of a digit [i hope this vocabulary makes sense :)]. With sinking buttons finger rests on an endplate which is far more comfortable.

  • Create New...