DIY capsule Q to Bob, Harvey or anybody who cares

[Marik]

DIY capsule. Q for Bob, Harvey or anybody who cares

I had an idea of push-pull capsule for a loooong time--since I was into building electrostatic speakers and repairing Quad57 and Martin Logans. Today I made a search on RAP and stumbled into this thread:

http://groups.google.com/groups?threadm=8clajc$fsv$1@fcnews.fc.hp.com&rnum=1

Some common sense tells me that the first reply was from Arcanemethods. Bob is that right?
So here some questions. First, I have always been wondering why mic diaphragm should be metallized? Once again, from my electrostatic speakers experience, graphite can be rubbed into diaphragm to make it conductive. Its high resistance will be only of benefit for the mic. The only potential problem I could see is a humidity, but high Ohm resistors would take care of this problem.
A few times I posted about it on TT, but nobody picked the idea.
The biggest problem with this design, as I see it, is a tuning resonance of the diaphragm. In this configuration (like in ribbons), because of the forces on diaphragm, the whole system should be mass controlled, so it should be tuned at the low frequency of the band. In the electrostaic field the diaphragm can become unstable; or without having enough tension-stiffness, with 35 micron or so spacers, start bottoming backplates. Increasing width of the spacers will take care of this, but will reduce sensitivity, which can be taken care of by increasing polarizing voltage.
Another solution would be tuning the diaphragm to the mid band, making it resistance controlled, and then EQ it.

Gentlemen, what do you think? Any fresh ideas?
I am tempted to try it... then add an omni capsule and get a nice
variable pattern.

 

[arcanemethods]

Re: DIY capsule. Q for Bob, Harvey or anybody who cares

I had an idea of push-pull capsule for a loooong time--since I was into building electrostatic speakers and repairing Quad57 and Martin Logans. Today I made a search on RAP and stumbled into this thread:

http://groups.google.com/groups?threadm=8clajc$fsv$1@fcnews.fc.hp.com&rnum=1

Some common sense tells me that the first reply was from Arcanemethods. Bob is that right?

Ah, we all live on now. Yep, that's me.

So here some questions. First, I have always been wondering why mic diaphragm should be metallized? Once again, from my electrostatic speakers experience, graphite can be rubbed into diaphragm to make it conductive. Its high resistance will be only of benefit for the mic. The only potential problem I could see is a humidity, but high Ohm resistors would take care of this problem.
A few times I posted about it on TT, but nobody picked the idea.

I completely agree with you. The other option that might make more sense is to use an electret diaphragm. In any case, one problem is to prevent charge migration from center to edge as the diaphragm moves because it causes non-linearity and disortion. Either an electret or a highly resistive coating will accomplish that for the mic just as for the speaker.

The biggest problem with this design, as I see it, is a tuning resonance of the diaphragm. In this configuration (like in ribbons), because of the forces on diaphragm, the whole system should be mass controlled, so it should be tuned at the low frequency of the band. In the electrostaic field the diaphragm can become unstable; or without having enough tension-stiffness, with 35 micron or so spacers, start bottoming backplates.

Yes, and just before that instability you have a situation where the compliance of the diaphragm is nulled. This will lower the resonance to subsonic very nicely and leave you mass loaded over the entire frequency band.

Increasing width of the spacers will take care of this, but will reduce sensitivity, which can be taken care of by increasing polarizing voltage.
Another solution would be tuning the diaphragm to the mid band, making it resistance controlled, and then EQ it.

Gentlemen, what do you think? Any fresh ideas?
I am tempted to try it... then add an omni capsule and get a nice
variable pattern.

I think you should give it a try! I don't think Keith took his to completion.

I had a great little book once upon a time that took the analysis of that kind of system to full detail but it was stolen in a briefcase of mine 25 years ago. I've forever missed that book. It derived expressions for the static and the dynamic stability of the system as a function of voltage, separation and mechanical compliance. I would just love to find it again.

I think it was titled "Electromechanical Transducers" and it covered all the audio types in exquisite detail with fully parameterized equivalent circuits. It was at least ten years old when I got it.


Bob

 

[Harvey Gerst]

The idea of a low resonance condenser capsule has been around for a while, but I don't think anybody has pulled it off sucessfully to date. Most, as you both well know, are tuned somewhere between 5 and 10,000 Hz.

I wish Stephen Paul was still around. He could answer this question a whole lot better than I can.

 

[Marik]

Thank you Gentlemen for your responses,

I like trying new things...

Bob,

I will try to find this book. Meanwhile... Since at least once upon a time you have read this book, what do you think would be a good start--low tuning frequency and thicker spacers (and how thick?), or at midband and EQ? I cannot deside myself...
I had in mind rectungular shape (like Pearl) for its better resonant properties. Also, I think with this shape for PP, the forces on diaphragm would be more even. Although the Pearl looks like a 'true' PP, it has an acoustic chamber on rear, and I heard that the diaphragm is tuned to 8K. Also, I think Shure has EQ in MKH RF mics.

So, what is your verdict?

 

[arcanemethods]

Thank you Gentlemen for your responses,

I like trying new things...

Bob,

I will try to find this book. Meanwhile... Since at least once upon a time you have read this book, what do you think would be a good start--low tuning frequency and thicker spacers (and how thick?), or at midband and EQ? I cannot deside myself...

I don't see any advantage to spacing it more than is necessasary for the maximum excursion and I'm not sure how to predict that. I'd say that whatever is used on existing condensers. The spacing just acts, as does the charge, as the turns ratio that appears as a transformer in the equivalent circuit which affects sensitivity.

You want the smallest possible spacing.

If you knew the expressions for the point of instabiliy as a function of voltage and diaphragm compliance then you could start with the bias voltage, calculate the minimum compliance and stretch it to meet that compliance.

I really do hope you can find that book. It makes all these factors explicit and calculable. There is no doubt that it is out of print and I'd think that Alebris would be about the only source. If I can find the time I'll see what I can find but at the moment time is short.

What I meant with regard to the low resonance, is that at that point of instability the positive compliance of the diaphragm just equals the negative compliance induced by the bias voltage and it nets to zero. This would eliminate all resonance, or make it at DC which isn't too meaningful. In practice you would make the compliance such that there is a resonant frequency but it is subsonic. This also gives some margin for error in the stretching.

I had in mind rectungular shape (like Pearl) for its better resonant properties. Also, I think with this shape for PP, the forces on diaphragm would be more even. Although the Pearl looks like a 'true' PP, it has an acoustic chamber on rear, and I heard that the diaphragm is tuned to 8K. Also, I think Shure has EQ in MKH RF mics.

So, what is your verdict?

I really don't know much about square vs circular diaphragms. It would seem that calculation of compliance would be easier for circular. As far as resonant properties, I think it less likely that there might be resonant modes above the subsonic with circular but that is just an intuitive feel not really backed by analysis. The other reason I would prefer circular is that there are no corners for refraction of high frequencies.

To my knowledge there is no PP microphone that has been marketed.

Another big advantage it would have is that since there are no ports of any kind involved, the self noise of the mechanical portion would be as low as theory allows. I would make the grills of as fine a wire as possible in as large a mesh as seems realistic. It should be stretched tight to avoid its motion.


Bob

 

[Harvey Gerst]

Are we also talking about an omni here, or a pressure gradient design? If pressure gradient, then the enclosed space behind the diaphragm, the distance between the elements, and the vents all contribute to the sound. With an omni, just the spacing and tensioning seem to be the critical elements. Of course, with large diaphragm omnis, the acoustic shadowing and difficulties in tensioning the larger surface would contribute to acoustic non-linearities.

I had a long discussion with Stephen Paul about using a freestanding one micron thick piece of coated glass as a diaphragm.

 

[arcanemethods]

Are we also talking about an omni here, or a pressure gradient design? If pressure gradient, then the enclosed space behind the diaphragm, the distance between the elements, and the vents all contribute to the sound. With an omni, just the spacing and tensioning seem to be the critical elements. Of course, with large diaphragm omnis, the acoustic shadowing and difficulties in tensioning the larger surface would contribute to acoustic non-linearities.

I believe we are talking a pure pressure gradient, a figure 8. No vents, no chambers. A diaphragm exposed on both sides. An electrostatic equivalent of a ribbon mic.

I had a long discussion with Stephen Paul about using a freestanding one micron thick piece of coated glass as a diaphragm.

I've thought about that too with another material that may be even more suitable but two things bother me. First, why? Second how would you suspend the thing?


Bob

 

[Marik]

<Are we also talking about an omni here, or a pressure gradient design? If pressure gradient, then the enclosed space behind the diaphragm, the distance between the elements, and the vents all contribute to the sound.>

Harvey,

As Bob has said, it is a pure pressure-gradient fig. 8 capsule. If I were to put a chamber behind of it, then my dilemma with tuning frequency would be easily resolved--added and tuned acoustical resistance of the chamber would take care of the peak resulted by tuning the diaphragm at midband. But I don't want this chamber--I want to 'hear' sound of diaphragm only. The vents are going to be pretty large--probably 3 mm diameter.


<Of course, with large diaphragm omnis, the acoustic shadowing and difficulties in tensioning the larger surface would contribute to acoustic non-linearities.
I had a long discussion with Stephen Paul about using a freestanding one micron thick piece of coated glass as a diaphragm.>

<First, why?>

Bob,
Then you don't need to tension the diaphragm for omni. Omni is a stiffness controled system with tuning resonance at high band. The tensioning is one of the ways to make the diaphragm stiff. With glass you don't have this problem. Also, it gives you very uniform characteristics and good 'repetative' results.
There is one mic I know of, which used this approach. I think it was Altec. Harvey, what was that?

<Second how would you suspend the thing?>

Imagination Bob, imagination and inventiveness.

BTW, I am looking for source for 0.8-1micron thick nickel foil. Anybody knows where I can find it?

 

[arcanemethods]

<Second how would you suspend the thing?>

Imagination Bob, imagination and inventiveness.
[/B]

I tried that and it didn't work. Insufficiently compliant.


Bob

 

[crazydoc]

I had a long discussion with Stephen Paul about using a freestanding one micron thick piece of coated glass as a diaphragm.

Excuse an ignorant interjection into this interesting discussion. How would a freestanding diaphram respond to SPL's? It would seem that since the pressure levels would be the same on the front and back (for most audio frequencies, depending on the size of the diaphragm), that there would be no movement and it would not work.
To quote Harvey in the big thread:
" The thing is, if you just hang a microphone diaphragm out in space, it will be pushed around by wind or by air currents of any kind (including if you just blow on it) but it won't pick up much in the audio frequency band because it's a thin element and the pressure from sound waves will tend to be identical on both sides of the diaphragm, at least until you get up to the high frequencies (which we'll talk about some other day), and when the pressure is the same on both sides of the membrane there is no net movement and no output."

 

[arcanemethods]

Excuse an ignorant interjection into this interesting discussion. How would a freestanding diaphram respond to SPL's? It would seem that since the pressure levels would be the same on the front and back (for most audio frequencies, depending on the size of the diaphragm), that there would be no movement and it would not work.

There is a distance from one side to the other as you go around it. What the diaphragm ends up responding to is the difference between the pressure on one side and that on the other. When you measure the difference in pressure between two closely spaced points what you get is a function of the air velocity. That is a vector quantity which means it has direction.

To quote Harvey in the big thread:
" The thing is, if you just hang a microphone diaphragm out in space, it will be pushed around by wind or by air currents of any kind (including if you just blow on it) but it won't pick up much in the audio frequency band because it's a thin element and the pressure from sound waves will tend to be identical on both sides of the diaphragm, at least until you get up to the high frequencies (which we'll talk about some other day), and when the pressure is the same on both sides of the membrane there is no net movement and no output."

I'm not sure what that is apropo of. While it may be true (I need to think about that), a diaphragm free floating in air is not the same as one pinned at the edges and responding at its center to the difference in pressure on either side.


Bob


Bob

 

[Marik]

<I had a long discussion with Stephen Paul about using a freestanding one micron thick piece of coated glass as a diaphragm.>

Harvey,

Not to get into further confusion, could you elaborate little more (if it is not a secret). Also, I would be very interested to know Stephan's ideas about it.

Bob,

Please check your PM box.

 

[guitarlover]

been reading with interest

I've been reading this thread with a lot of interest. Not so much the existential aspects of hanging a diaphram in space but rather the physical process of building a condenser unit and diaphram.

This is something I have pondered since I got interested in homerecording. I was the kid who built crystal radios and capacitors made from tin foil and wax paper. The thought of making a mic is something that has whetted my appetite.

I read the link to that "other" forum and I'm going to check the local library today for those books (I am blessed with a good library system here in NW Ohio). My question is simple. Are you folks aware of anymore resources to get me started? Such things as drawings, instruction, etc. The amplifier is actually the easier portion for me as I have experience in making PC boards. I'm more interested in the physical aspects of the condenser itself, materials (especially the diaphram), size etc.

Thanks for the help,

Steve

PS. If I do find the book in question, I'll let you know.

 

[Harvey Gerst]

Excuse an ignorant interjection into this interesting discussion. How would a freestanding diaphram respond to SPL's? It would seem that since the pressure levels would be the same on the front and back (for most audio frequencies, depending on the size of the diaphragm), that there would be no movement and it would not work.
To quote Harvey in the big thread:
" The thing is, if you just hang a microphone diaphragm out in space, it will be pushed around by wind or by air currents of any kind (including if you just blow on it) but it won't pick up much in the audio frequency band because it's a thin element and the pressure from sound waves will tend to be identical on both sides of the diaphragm, at least until you get up to the high frequencies (which we'll talk about some other day), and when the pressure is the same on both sides of the membrane there is no net movement and no output."

It was never intended to just hang. It was an idea to have the traditional mylar diaphragm replaced with a piece of thin 1 micron glass. I reasoned that the stiffness of the glass would put the resonance above the audible range. And the need for proper tensioning would be eliminated.

 

[Harvey Gerst]

I had a long discussion with Stephen Paul about using a one micron thick piece of coated glass as a diaphragm.

Harvey,

Not to get into further confusion, could you elaborate little more (if it is not a secret). Also, I would be very interested to know Stephan's ideas about it.

No secret. After the usual 3 hour phone discussion with him, Stephen basically said (in his usual kindly manner), "Look dummy, the diaphragm has to be able to move."

"Ohhh", I said. End of discussion.

I really miss Stephen; he would have really enjoyed this thread. (He would have made everybody here miserable, but he would have been able to contribute so much.)

 

[crazydoc]

While it may be true (I need to think about that), a diaphragm free floating in air is not the same as one pinned at the edges and responding at its center to the difference in pressure on either side.

I assume this is how a ribbon mic works? I guess I don't understand, with the very short distances involved in relation to the wavelenth of sound (as opposed to ultrasound), how there could be a significant pressure difference between the two sides. It seems that without a baffle of some kind to at least delay the wave to the backside that it would have no movement.

Maybe someone could point me to a good explanation of the physics of ribbon or diaphragm movement in an acoustical environment - that is, why does the ribbon move? I can't seem to find one. Thanks.

 

[crazydoc]

BTW, I found this interesting link in my ramblings, for DIYers. There's a link there where you can buy 1 micron aluminum foil for cheap.
http://www.lkmusic.co.nz/ribbonfix.htm

 

[Harvey Gerst]

I assume this is how a ribbon mic works? I guess I don't understand, with the very short distances involved in relation to the wavelenth of sound (as opposed to ultrasound), how there could be a significant pressure difference between the two sides. It seems that without a baffle of some kind to at least delay the wave to the backside that it would have no movement.

Maybe someone could point me to a good explanation of the physics of ribbon or diaphragm movement in an acoustical environment - that is, why does the ribbon move? I can't seem to find one. Thanks.

Physics? PHYSICS? WE DON' NEED NO STINKIN' PHYSICS !!

Actually it's pretty simple. Ribbon mics are velocity mics; they depend solely on the air motion hitting the diaphragm. That's why they have the proximity buildup when you get in close. Omnis, at the other end, depend on changes in the air pressure in the room.

All gradient patterns use a combination of pressure and velocity to create the desired patterns. Pressure omnis at one end of the scale, figure 8 velocity at the other end of the scale. In between, is the cardioid pattern, a 50/50 mix of pressure and velocity. Cut the velocity part by half again and you get a wide cardioid with less proximity effect. Cut the pressure part of a cardioid in half and you get hypercardioid, with more proximity effect.

There's basically just two components: pressure and velocity. Mixing them in different ratios produce all the in-between pressure gradient patterns. Look at the markings of any multipattern mic; you'll see omni at one end, and figure 8 at the other end of the scale. Everything else is in-between.

If you start with figure 8, and gradually bring in the omni pattern, the back side of the figure 8 goes down, and the front side gets louder. When you get to cardioid (a 50/50 mix of the two patterns), the + output of the omni in back exactly cancels the - output of the figure 8 in back and the output drops to 0 as you get around to the back, creating the familiar cardioid pattern.

Beyond cardioid, the backside starts to open up again as you use less figure 8, till you finally get to the circular pattern of the omni. And of course the velocity component (which creates the proximity effect) is now gone, making the omni free of proximity effect.

Does this make a little more sense now?

 

[Marik]

<No secret. After the usual 3 hour phone discussion with him, Stephen basically said (in his usual kindly manner), "Look dummy, the diaphragm has to be able to move." "Ohhh", I said. End of discussion.
I really miss Stephen; he would have really enjoyed this thread. (He would have made everybody here miserable, but he would have been able to contribute so much.)>

Thank you Harvey. So, as I understand, by 'freestanding' you mean--'not tensioned', is that right?
As I wrote earlier, Altec once made a glass diaphragm capsule commercially (I have to dig the model#). I am wondering why it was the only attempt.... I would think it is pretty fragile.



<I assume this is how a ribbon mic works? I guess I don't understand, with the very short distances involved in relation to the wavelenth of sound (as opposed to ultrasound), how there could be a significant pressure difference between the two sides. It seems that without a baffle of some kind to at least delay the wave to the backside that it would have no movement.>

Crazydoc,

Even with this short distances there is a time delay between front and rear. The interesting thing is that the baffle will improve LF, but lower HF cut off point. Have a look at this:

http://cara.gsu.edu/courses/MI_3110/mic1/5.htm

and this:

http://home.vicnet.net.au/~macinc/news9.htm



< read the link to that "other" forum and I'm going to check the local library today for those books (I am blessed with a good library system here in NW Ohio). My question is simple. Are you folks aware of anymore resources to get me started? Such things as drawings, instruction, etc. The amplifier is actually the easier portion for me as I have experience in making PC boards. I'm more interested in the physical aspects of the condenser itself, materials (especially the diaphram), size etc.>

Guitarlover,

While you are in the library, try to find old (I mean OLD) books on acoustics. Also, check if they have 'Microphone Engineering Handbook" edited by Michael Gayford, ISNB 0 7506 1199 5.
None of these books, however, will give you any practical information about capsules. Also, try to search for patents. In my opinion the best resource on the whole net for DIY traditional (Neumann style) is a thread which is still going on TT. You will find there a lot of links, practical designes, tips, etc. Have a look:

http://www.recording.org/postt1597.html

 

[arcanemethods]

Maybe someone could point me to a good explanation of the physics of ribbon or diaphragm movement in an acoustical environment - that is, why does the ribbon move? I can't seem to find one. Thanks.

From the acoustic wave equation you can show that the bulk velocity of air in a sound wave in a particular direction (the vector component in that direction) is the time integral of the difference in pressure between two closely spaced points in space. The axis of those two points determines the direction in which the velicity is projected and the spacing determines an upper cutoff frequency beyone which the relation no longer holds.

Thus the two sided diaphragm or ribbon measures the pressure at two points in space, one at one side and one at the other. The distance around from one side to the other is the separation between the two spatial points that I described above. Since the pressures are measured at opposite sides the diaphragm/ribbon moves in proportion to that pressure difference; they are effetively subtracted from each other by measuring the diaphragm/ribbon motion.

What remains is to integrate that with respect to time to get a signal that is porportional to the component of bulk air velocity in the direction of the microphone axis. If you don't integrate, the output will have a respnse that starts low at low frequencies and goes up at 6 dB/octave as you go up in frequency. The integration flattens that out.

In real microphones that integration is accomplished by reaction of the system to the mass of the diaphragm and by air moving in various chambers that may or may not exist in the design. This is what is meant by mass loading.

Also in a real microphone there is a compliance or spring constant involved due to the stretched diaphragm. With that in place, the system only acts as a mass loaded one and does the needed integration above a certain frequency determined by the resonance of the diaphragm due to its combined mass and compliance. This is why directional microphones exibit rolloff below the resonance frequency and why we are thinking of ways to get that resonance as low as possible. If we can get it subsonic then the frequency response curve will be integrated to flat over the entire audible band.

Does this help?


Bob