For what scale reeds does the tongue pass completely through plate?

[Dana Johnson]

On 2/9/2020 at 5:55 PM, ttonon said:

On 2/4/2020 at 12:59 AM, Dana Johnson said:

I originally tried it because the reeds on my old Bb/F Jeffries had all the reeds carefully belled by hand.

Just how do you do that?  Do you use a ball mill cutter?  I can't see that you're cutting the sides with some kind of file or scraper, so maybe I really don't understand what you mean by a "belled" relief angle.  

I use specially made scrapers in three sizes and curved cutting edges.  I scrape from the bottom of course and the tips of the scrapers touch the steel jig before the bell is cut too deep and wrecks the top side of the reed window.  The Jeffries was belled with a file.  The scraper works quite quickly, and I mostly just do the long sides of the window and leave the regular draft at the tip.  
   I do have some issues with the blower ( axial duct fan ) speed pitches, but I don’t find the kind of interference with reed vibration that the tube resonances have.  It is also easy just to change the speed slightly to move it away from the reed pitch.

 

On 2/9/2020 at 5:55 PM, ttonon said:

On 2/4/2020 at 12:59 AM, Dana Johnson said:

As far as mechanisms go, I can’t see a curved draft is really any different than a straight one except that whatever is happening at the side clearances sees a gap that increases at a accelerated rate.  

I agree.  Would you agree that the effect is a little like that of a thinner shoe (plate)?

I would agree it is a little like having a thinner reed plate, but unlike a thinner reed plate, the reed efficiency increases noticeably on the oscilloscope amplitude up to a point of no further improvement, so even as the reed enters ever wider sections as the pressure goes up,  some effect on the sound creation requires some degree of containment.  I don’t really know what creates the actual sound amplitude.  If it were just pressure of the bellows, then  physical reed amplitude would make little difference. Perhaps it is pressure at a certain air flow. In electricity, that would be watts which is a power term.  ( pardon my daughter’s 8th grade physics class ) so that sounds reasonable, but increasing the pressure past the point where the reed no longer swings farther doesn’t make it louder.  Maybe that is because the reeds start to choke up and you can’t really force more sir through.  

 

On 2/9/2020 at 5:55 PM, ttonon said:

On 2/4/2020 at 12:59 AM, Dana Johnson said:

When concertina reeds have too low a draft angle, they require more pressure to get to max volume.

This surprises me because the major amount of energy addition to the vibration is during the downward movement of the tongue through the slot.  With high draft angle, there's a lot more air flow between the edges of the tongue and the walls of the vent, lowering the pressure driving the motion. 

I think we might need to distinguish between reed amplitude and sound amplitude.  The reed is after all primarily a mechanical valve perhaps it is a bit like the base in a transistor (or grid in a vacuum tube ) that controls the gain , ( dating myself ) between the emitter and collector operating at some larger voltage.  I don’t know... probably just more bad analogies .

Best Wishes,

Dana 

[ttonon]

On 2/11/2020 at 10:05 PM, Dana Johnson said:

It is also easy just to change the speed slightly to move it away from the reed pitch.

Do you have a speed-controlled DC motor on the blower?  That's a good feature.  I should change to that.  Right now, I'm using a "Windjammer" regenerative blower, which is one-speed AC, but it's capable of 18 inched WC, and I need it because of all the valves and small diameter tubes in the system.  

 

On 2/11/2020 at 10:05 PM, Dana Johnson said:

I don’t really know what creates the actual sound amplitude.

Neither do I; I haven't studied it carefully, though I have an inkling from other researchers.  If you recall, several years ago, I posted a link to a paper by Ricot, et al, in France, where they did analysis of the acoustic sound field produced by an accordion reed.  The results look good, with good agreement with measurement, so I assume the mechanisms they uncovered can give us a good idea about how the sound is produced.  It's quite mathematical, as acousticians often are.  I believe there are intuitive explanations yet to be discovered.  You're right in pointing out that the tongue vibration is not the same as the sound.  As I mentioned, I'm focused now on the vibration, and I hope the understanding of it can help in understanding the sound.  Ricot didn't solve the vibration problem, merely substituted the actual tongue vibration for a simple valve.  Perhaps a strict vibration analysis such as mine can contribute towards a better understanding of the sound.  

 

On 2/11/2020 at 10:05 PM, Dana Johnson said:

If it were just pressure of the bellows, then  physical reed amplitude would make little difference.

It's an interesting hypothesis.  I can't say for sure.  Certainly the bellows pressure fixes the maximum energy available for all aspects of the motion: both tongue vibration and acoustic sound field.  Bear in mind, the acoustic energy is a small fraction of the total energy of the vibration, so a subtle change in energy transfer from the vibration to the sound field might produce a lot of sound.  

 

On 2/11/2020 at 10:05 PM, Dana Johnson said:

but increasing the pressure past the point where the reed no longer swings farther doesn’t make it louder.

I'm not sure I agree, simply because higher bellows pressure makes more energy available for the sound field.  So more bellows pressure, more energy available for sound; thus, higher volume.

 

On 2/11/2020 at 10:05 PM, Dana Johnson said:

Maybe that is because the reeds start to choke up and you can’t really force more sir through.  

I believe you can always force more air through.  The higher the bellows pressure, the higher the air flow rate through the vent.  But I agree that there appears to be a maximum amplitude the tongue could vibrate at, and if that's really true, it's a clue to how the whole mechanism works.  The mechanism for that still puzzles me, though it must have something to do with the dissipation that must grow in relative importance to the vibration.  

 

On 2/11/2020 at 10:05 PM, Dana Johnson said:

I think we might need to distinguish between reed amplitude and sound amplitude.

Very good point, and I sometimes lose sight of it.  You're focusing on volume and I'm looking at vibration.  Not the same.  I don't think the mechanism for energy transfer to sound is very efficient, if you define efficiency as sound energy out divided by muscle energy in.  It would be interesting to compare it to the efficiencies of other instruments.  

 

Nice talking,

Tom

[Dana Johnson]

13 hours ago, ttonon said:

Do you have a speed-controlled DC motor on the blower? 

I never considered the output from the speed control. Since AC is going in, I thought it might just be a switching type speed control, like a lot of dimmers that simply vary the time the power is going to the filament.  Regardless, I just looked in a catalog for in line duct fans and picked a speed controller in the right power range.  Sounds like you have a monster blower, more like a vacuum cleaner to work at those pressures.  All I know is that my blower motor bears no resemblance to my 2 hp

 

On 2/14/2020 at 3:24 AM, ttonon said:

I believe you can always force more air through.  The higher the bellows pressure, the higher the air flow rate through the vent.  But I agree that there appears to be a maximum amplitude the tongue could vibrate at, and if that's really true, it's a clue to how the whole mechanism works.  The mechanism for that still puzzles me, though it must have something to do with the dissipation that must grow in relative importance to the vibration. 

you may be able to force more air through the reed though physically that becomes increasingly difficult past a certain point.  At some  point the reed simply stops oscillating (what I call choking). The less stiff s given reed is relative to its pitch and length, the sooner this happens.  It seems possible though that the greater the reed amplitude, the less time it spends in the power zone.

On 2/14/2020 at 3:24 AM, ttonon said:

 

13 hours ago, ttonon said:

Bear in mind, the acoustic energy is a small fraction of the total energy of the vibration,

That is an interesting concept.    I know it all gets turned into heat,  but is that in the air cooled reed or in the muscles / blower that does the driving?   It is funny to realize since our experience is all about the exercise we get and the resulting sound out.  On the other hand, since sound pressure can sympathetically drive a reed I wonder if the natural feedback mechanisms don’t increase the reed’s physical amplitude.  

[Chris Ghent]

The idea for my constant pressure tuning device came from Dana, so its a 6” duct fan, in my case speed controlled by nothing more than a domestic 240v lighting dimmer. 

[ttonon]

On 2/14/2020 at 3:24 AM, ttonon said:

Bear in mind, the acoustic energy is a small fraction of the total energy of the vibration, so a subtle change in energy transfer from the vibration to the sound field might produce a lot of sound.  

Yeah, well, I think I might have to retract that statement.  I knew I was sticking my neck out by saying it because I never verified it myself and I hate saying things that seem like common knowledge without my own direct verification or without documentation.  So I did look into this issue and found some things that may be of interest to some here. 

 

We want to compare the total energy input to the total sound energy output.  Power input to the process is the result of our muscle power, and of course we can measure that by measuring our force on the bellows and multiplying it by the speed at which the bellows closes or opens.  We don't have to do quite that, because that power is very closely equal to the product of the force the bellows pressure exerts on the airflow through a slot (consider only one sounding reed) times the velocity of that airflow.  The bellows pressure force is equal to the bellows pressure times the area of (slot) flow. 

 

Bellows pressure is easy to know, but what is the slot flow area and the air velocity through the slot?  The fact that there's a vibrating tongue obstructing air flow through the slot adds complexity, but we can simplify things considerably, first by saying the actual air flow area is about half the slot area and second by taking the average air flow velocity, which, thanks to Mr. Bernoulli, is easy to calculate.  That air flow velocity in fact accurately represents the power input available to the vibration (power input without vibration).  Here's our reasoning: bellows pressure forces the motion, exciting tongue vibration, which leads to acoustic power output and dissipation, or heat.  In turn, of course, the acoustic power output is also dissipated.  Of course, acoustic power output is generated in conjunction with tongue vibration and does not result at a time after vibration, but all these details don't prevent us from getting a good estimate of the power input. 

 

Separating the mechanical energy of tongue vibration from acoustic energy would be extremely difficult if you'd want to do that with a complete analytical understanding of the entire phenomenon.  But there are easier ways than that, so stay tuned.  

 

The product of bellows pressure, slot area, and airflow velocity is very simple, and for a bellows pressure of 6 inches WC, slot dimension of 2 by 0.1875 (using only half that), I get about 1.5 milliwatts of input power.  

 

What's the total acoustic power output?  A simple estimate is to consider how loud the reed sounds when you're a certain distance away, converting that to acoustic power intensity, then multiplying that by the area of a sphere of that distance.  Here again, we make rough estimates.  If you go to the web and find typical sound levels produced by various sound sources, you learn that conversational speech is about 60 decibels at one meter and a vacuum cleaner is about 70 dB at one meter.  Let’s assume we have a particularly muscular concertina player who’s an extrovert and take 70 dB at one meter.  This gives a total of 0.13 milliwatts, or about 1/10 the input power calculated above. 

 

So it looks like my statement above can be accurate, but bear in mind the simplifications we made.  We are definitely in the realm of order-of-magnitude reasoning.  I thus would not be surprised if in some cases, say for different pitched reeds, that statement may not be too accurate.  I come away from this thinking, “Okay, that’s a rough rule of thumb, but…”

 

Best,

Tom

Edited February 17, 2020 by ttonon