How does condenser mic capsule load affect THD?

[MicUlli]

Hello All,
I wonder in which way electrical loading of the condenser mic capsule affects THD..
Some thoughts:

1. Because signal voltage is generated by changing the capacitance of the capsule a little bit, it changes also the electrical capsule impedance.
2. The electrical input impedance of the internal impedance converter interacts with the (non stationary) capsule impedance (voltage divider).
3. Hence it should show THD..

"Good" impedance converter solutions would have very high input impedance throughout the audio frequency range,
"Bad" solutions (like some tube amps or the KM84ish negative feedback via capacitor circuits) would show significant THD.

What are your thoughts?
BR MicUlli

 

[MicUlli]

Okay, seems to be of lower interest..

I solved the calculation by myself, find attached the PDF.
Based on the formulas it is very easy to figure it out for other mic capsules.

And indeed, load matters! Never ever use a charge amp for mic capsules or be aware for a high THD..

 

[rock soderstrom]

Okay, seems to be of lower interest..

On the contrary, I find it interesting. I am certainly not the only one. The problem is, I can't contribute anything clever to the subject and I think you know much more about it than I do.

I solved the calculation by myself, find attached the PDF.
Based on the formulas it is very easy to figure it out for other mic capsules.

And indeed, load matters! Never ever use a charge amp for mic capsules or be aware for a high THD..

Your result proves my last sentence impressively. Thank you so far.

 

[kingkorg]

The problem is, I can't contribute anything clever to the subject and I think you know much more about it than I do.

+1

Could be the reason why poeple don't like how simple cap in parallel with capsule pad sounds? Or did i get this wrong?

 

[molke]

Weren’t you the one who linked to the German PhD thesis of Holger Pastillé (pp. 45–67)?

The problem is, I can't contribute anything clever to the subject and I think you know much more about it than I do.

Same here.

Could be the reason why poeple don't like how simple cap in parallel with capsule pad sounds? Or did i get this wrong?

That should be Fig. 35a for a microphone with stray capacitance, where THD increases from 0.5 to 3.5 % with the pad (at 140 dB SPL). There is also an English JAES paper for theoretical background, but I don’t have access to it anymore.

Those 0.5 % are a rather arbitrary engineering standard (from a perceptual point of view). The listening test I conducted once upon a time showed a just noticeable difference at well above 110 dB SPL (and no relation to any particular THD level). And sometimes, I tend to like a little distortion.

 

[MicUlli]

Weren’t you the one who linked to the German PhD thesis of Holger Pastillé (pp. 45–67)?

Same here.

That should be Fig. 35a for a microphone with stray capacitance, where THD increases from 0.5 to 3.5 % with the pad (at 140 dB SPL). There is also an English JAES paper for theoretical background, but I don’t have access to it anymore.

Those 0.5 % are a rather arbitrary engineering standard (from a perceptual point of view). The listening test I conducted once upon a time showed a just noticeable difference at well above 110 dB SPL (and no relation to any particular THD level). And sometimes, I tend to like a little distortion.

Yes, I mentioned this paper. But (to be honest) some formulas are a bit complicated for me. Therefore my short document. It reduces the THD estimation to only two variables: polarisation voltage and sensivity.
In case of a capacitively loaded capsule the third variable (load cap) can be easily incorporated.

 

[MicUlli]

+1

Could be the reason why poeple don't like how simple cap in parallel with capsule pad sounds? Or did i get this wrong?

I think so. With a -10dB pad via parallel cap THD rises really considerably and couteracts for what the pad shall provide..

 

[kingkorg]

I think so. With a -10dB pad via parallel cap THD rises really considerably and couteracts for what the pad shall provide..

Which brings me back to thinking if designing (modding) a capsule by increasing the distance between the backplate and diaphragm in order to reduce parasitic capacitance of the capsule itself, and increasing the pol. voltage might be something worth trying.

 

[ccaudle]

(modding) a capsule by increasing the distance between the backplate and diaphragm

Doesn't that also change damping, and so modify the behavior more than just a little?

 

[kingkorg]

Doesn't that also change damping, and so modify the behavior more than just a little?

Absolutely, that's why there was designing in front of the modding in the original post. On some capsules though that extra space might bring something to the table. Sound wise. Those cheapie mxl603 type capsules could be great for this kind of experimentation.

 

[Gus]

listen to a 84 with the pad on vs pad off and adjust the playback level

That posted some tube microphone designs do have plate to grid capacitance.

Some well liked microphone have added plate/drain to grid/gate capacitance.

The tube is vacuum capacitance vs a JFET drain to gate.

 

[rogs]

Be interesting to know whether the THD calculations are completely different for a capsule used in an RF bias configuration, where the capsule is essentially part of a low impedance system?
And I imagine the calculations would be different yet again, when applied to Sennheiser 'push/pull' capsules -- which are symmetrical in construction, and used as part of a low impedance RF system....

 

[MicUlli]

Be interesting to know whether the THD calculations are completely different for a capsule used in an RF bias configuration, where the capsule is essentially part of a low impedance system?
And I imagine the calculations would be different yet again, when applied to Sennheiser 'push/pull' capsules -- which are symmetrical in construction, and used as part of a low impedance RF system....

In a LF approach the capacitance change is the PARASITIC effect. But in a RF approach capacitance change is the DESIRED effect. Therefore THD calculation is completely different.
Sennheiser's push/pull capsules make usage of a simple math formula: (C+dC)*(C-dC) = 1 for small d. The two Cs are driven as part of a bridge configuration in a resonance tank and therefore small d doesnt change the tank resonance frequency.

 

[MicUlli]

Which brings me back to thinking if designing (modding) a capsule by increasing the distance between the backplate and diaphragm in order to reduce parasitic capacitance of the capsule itself, and increasing the pol. voltage might be something worth trying.

I don't think so. Increasing the air gap leads to a smaller C_capsule. Hence the other parasitic C's get larger in relationship and THD gets larger too.
But of course it pushes the SPL limit higher because the mic sensivity decreases...

 

[rogs]

In a LF approach the capacitance change is the PARASITIC effect. But in a RF approach capacitance change is the DESIRED effect. Therefore THD calculation is completely different.
Sennheiser's push/pull capsules make usage of a simple math formula: (C+dC)*(C-dC) = 1 for small d. The two Cs are driven as part of a bridge configuration in a resonance tank and therefore small d doesnt change the tank resonance frequency.

The distortion advantages derived from the push/pull concept are mentioned by both John Willett in his paper HERE and by Manfred Hibbing in his work HERE
I'm guessing Sennheiser have the concept pretty much tied up commercially with copyright and patents?...
I would love to see one of the mic hobbyists who actually make capsules have a go at trying one out

 

[MicUlli]

The distortion advantages derived from the push/pull concept are mentioned by both John Willett in his paper HERE and by Manfred Hibbing in his work HERE
I'm guessing Sennheiser have the concept pretty much tied up commercially with copyright and patents?...
I would love to see one of the mic hobbyists who actually make capsules have a go at trying one out

Yes, very impressive. The papers address some additional aspects of THD in mics. Unfortunately I doubt that a hobbyist is capable to produce such capsules

 

[kingkorg]

The distortion advantages derived from the push/pull concept are mentioned by both John Willett in his paper HERE and by Manfred Hibbing in his work HERE
I'm guessing Sennheiser have the concept pretty much tied up commercially with copyright and patents?...
I would love to see one of the mic hobbyists who actually make capsules have a go at trying one out

This is pure gold! Have you by any chance had the opportunity to hear the CD mentioned in the first document?

One other thing which came to mind is that at least for low diaphragm damping part, the existing capsules are easy enough to mod and eq correct in the post. You could actually test this by just separating the halves of capsules in your RF mic. Basically just remove one backplate of rk12. The midrange and signal will go through the roof. Not sure what kind of polar pattern you get, but good enough for testing noise.

 

[rogs]

This is pure gold! Have you by any chance had the opportunity to hear the CD mentioned in the first document?

One other thing which came to mind is that at least for low diaphragm damping part, the existing capsules are easy enough to mod and eq correct in the post. You could actually test this by just separating the halves of capsules in your RF mic. Basically just remove one backplate of rk12. The midrange and signal will go through the roof. Not sure what kind of polar pattern you get, but good enough for testing noise.

I've never heard the actual CD he mentions, although I'm pretty sure THIS is the Lenehan/Satie recording he describes as his second piece on page 6.

Interesting thoughts on the backplate removal... I must rummage around and look for a 'sacrificial' capsule...

 

[kingkorg]

Now i have to try that 2 speaker distortion measurement system, as soon as i wrap my head around it.

 

[kingkorg]

I have tested this two speaker method in virtual environment (daw), simply to check if i am on the right track. I must say it is truly amazing, however calculating the THD values in real environment with or without anechoic chamber would be a nightmare, for me at least. Supposing i got the paper right.

Instead, i read the Amir's (ASR forum) detailed analysis of Audio Technica m50 headphones

Audio Technica ATH-M50X Review (Closed Headphone)

They have below 0.5% thd above 500hz at 114db!

So i tested some of my mics and capsules at high spl even above 114db with properly sealed m50. And then even above 120db using my KRK's tweeter which seems to have also very low thd (double checked with several mics, sm57 included)

To me it seems this whole Sennheiser THD thing is quite a bit overplayed. I have no idea which mics they used against their MKH, but my mics with regular suspect capsules always start clipping the circuit before the capsule shows increase in THD. I even have a k67 mic with just 15v polarization. I see no increase in THD at HF. Came out very clean.

It could be they just used some crapy mics to shoot against MKH, or that there might be something wrong with their measurement method.

There is most certainly something to it, but i don't think it will ever be audible. Reminds me of c800g overengineered cooling system.

I might just have the MKH800 capsule heading my way, fingers crossed, so something interesting might be happening soon.