Capsule polarization resistor/capacitor values.

[vmanj]

I welcome everyone.

To polarize the capsule, (after the divider) different values of the resistor and capacitor are used.

For example, in U-47 it is 100M and 10nF.
In ELA-M 251 it is 8M and 4.8nF.

If, for example, in ELA-M 251, instead of 8M and 4700pF, change the values U-47 to 100M and 0.01mF, what changes can be in the sound?

 

[kingkorg]

It's a rc lpf, so the cutoff frequency matters, it affects the noise, doesn't affect the sound in any way. Lower value resistor just before the capsule will give you faster capsule recovery time in case of plosives or moisture shock, but as it forms lpf in combination with the capsule capacitance it might result in higher noise floor. The high value grid resistor is more important to the sound.

 

[vmanj]

It’s clear, but why in Sony С800G, after the divider, there is no resistor at all ..
It turns out that with such a scheme there should be a very high noise level.
It's actually a low noise microphone.

Also, in the ELA M-251, the resistor value is low, so it must have an increased noise level?

If so, are the developers of such expensive microphones really just saving on high-resistance resistors?

 

[abbey road d enfer]

The actual value of the resistor does not matter much, because there is no current in normal operation. The R and C determine how long it will take to properly charge. If the time constant (R.C) was too long, the microphone would take too long to be operational. Since the tube takes about 30 seconds to operate, it is just adequate to make the bias time constant similar.
In the C800G, there is a 47uF capacitor, coupled with a 180k and a 47k resistor, so the time constant is about 2s. Time-constant is the time the voltage takes to reach 71% of its final value. So in practice, the nominal voltage will be reached after about 10-12s.
In the M251 and the U47, the time constants are much shorter; it means that the bias voltage will be ready before the tube is operating normally. It does not matter much.
The values are chosen to filter out the noise from the PSU, so they are determined in relation with the PSU noise, which depends on many things, the ESR of capacitors, the inductance of the power transformer, the speed of rectifiers...

 

[oscarhuang]

I'm not an expert, but I also do a lot of mics. In terms of hearing, the lower the resistance value, the less the low frequency of the sound. I think the capsule of the original ELA M251 has a lot of low frequency, so it uses 30M and 8M resistors. The resistance values of the new TELEFUNKEN are 100M and 500M, and the other version also has 10M and 300M.

 

[kingkorg]

This could be the effect of the lower value grid resistor, which is usually the same value. I cant see how polarisation resistor could have this effect.

 

[vmanj]

Those. the value of the polarizing resistor can also affect the bass response ?

 

[abbey road d enfer]

In terms of hearing, the lower the resistance value, the less the low frequency of the sound. I think the capsule of the original ELA M251 has a lot of low frequency, so it uses 30M and 8M resistors. The resistance values of the new TELEFUNKEN are 100M and 500M, and the other version also has 10M and 300M.

Of course, for a given capsule capacitance, the higher the grid resistor, the lower the LF cut-off. A typical LDC, at about 80pF, with a grid/gate resistor of 1 Gigaohmgives a LF cut-off frequency of 2Hz. Nobody needs 2Hz. But with 100Megohms, it becomes 20Hz, which has some audible consequences, in particular handling noises will be somewhat attenuated. ELA251, with only 30 Megohms, put the LF cut-off at 60Hz, which should be definitely audible. However it must be noted that all cardiois capsules are inherently lacking LF response, but corollarily have a noted proximity effect. The combination is the very well known "chest voice", which happens at about 80-100 Hz, so the lack at 60Hz and below is not a problem.
But for distant micing, as in movie stage, I think it has been recognized as an issue.
As a result of the low value of the grid resistor, the intrinsic noise is not so good, which is another reason for not liking it for distant micing.

 

[abbey road d enfer]

Those. the value of the polarizing resistor can also affect the bass response ?

No.

 

[oscarhuang]

No.

Dose it change the sound? I used 10M ohm and 1G ohm. I can heard the difference.

 

[abbey road d enfer]

Dose it change the sound? I used 10M ohm and 1G ohm. I can heard the difference.

Actually it depends on the circuit. If the bias resistor ends up being in parallels AC-wise with the capsule, it changes the LF cut-off.
On LDC's theis arrangement is not desired, in favour of applying bias to teh fixed electrode.

 

[vmanj]

Well, the bass response is more or less clear.

But with the noise level - I want to clarify.
If I understand correctly, then the lower the value of the resistor, the higher the noise level ?

 

[kingkorg]

Do a lpf rc calculation, it will become more obvious what is going on.

 

[vmanj]

I don't have the knowledge to do the calculations.
For me, it's just a hobby, that's why I asked my question on the forum.
There are different answers here, so I just wanted to clarify ..

In any case, I am grateful to everyone who participates in this.

 

[kingkorg]

Yes basically you get lower noise, but to understand it fully it would be great to understand how rc lpf works. It is not hard, there are calculators online. The resistor on its own doesn't matter, the interaction between the resistor and following capacitor does. They depend on eachother, pure value of the resistor doesn't mean anything. In some cases 1Meg is used. It would also reveal why the value of the resistor has no effect on the sound.

 

[Khron]

It's a rc lpf, so the cutoff frequency matters, it affects the noise, doesn't affect the sound in any way. Lower value resistor just before the capsule will give you faster capsule recovery time in case of plosives or moisture shock, but as it forms lpf in combination with the capsule capacitance it might result in higher noise floor. The high value grid resistor is more important to the sound.

Considering the replies / discussion I've seen here, i feel there's a slight possibility there's some misunderstanding / confusion, between:
- the resistor in series with the bias voltage for the capsule (in the Mohm range usually), that has a capacitor to ground after it, on one hand
- on the other hand, the (often) high value resistor (hundreds of Mohm up to 1-2Gohm) going from that above-mentioned bias voltage node, to the capsule itself

The former is quite trivial to include in an RC low-pass filter calculation, since the capacitor value after it is known or readable; the latter includes the capacitance of the capsule itself, which is less known / less easily determined.

 

[vmanj]

I tried the online calculator (pictured).
How can you determine the noise level from these graphs?

 

[kingkorg]

Excellent! Look at the graph's curve (bode plot) as eq low pass / high cut filter. The cut off frequency is shown, and moved via change in capacitance/resistance. You can imagine the noise as audio signal coming from the psu, and you can see how it gets filtered by the "eq curve".

Important to mention this is just the last filter in series of several filters, so it might not be even necessary if the supply is clean. One thing to take into consideration is the time constant of rc (resistor/capacitor) Abbey mentioned.

Phase shift here is not important as it has nothing to do with audio signal picked up by the capsule.

Edit:
Forget this part, for reasons explained by Abbey in the following post.
However phase could be important in cr (not rc) high pass filter that is formed by the capsule (and or coupling cap depending on topology) and high value grid resistor. This filter might mess with low frequencies in somewhat unpredictable ways in multi mic setup while recording the same source. Some people say it is inaudible, and i might agree if you use only one mic, but once you use another one this could create some issues.

 

[abbey road d enfer]

I tried the online calculator (pictured).
How can you determine the noise level from these graphs?

You can't because you don't know the noise level coming from the PSU. And even if you knew it, it depends on the actual PSRR (Power Supply Rejection Ratio) of the circuitry, which is not too difficult to calculate, but involves the circuit, teh tube and the transformer.

 

[abbey road d enfer]

However phase could be important in cr (not rc) high pass filter that is formed by the capsule (and or coupling cap depending on topology) and high value grid resistor. This filter might mess with low frequencies in somewhat unpredictable ways in multi mic setup while recording the same source. Some people say it is inaudible, and i might agree if you use only one mic, but once you use another one this could create some issues.

In most condenser mics, there is one dominant pole, which defines the amount of phase-shift at low frequencies. The phase-shift at audible frequencies related to a single-pole cannot be more that 90°. So two mics will combine in the same quadrant, making phase cancellation impossible, unless polarity is reversed somewhere.