"U67"-ish microphone

[vmanj]

I welcome everyone.
I do not have the values of resistors R5.R6 - 100M, but there are 200M.
If they are used in this circuit 200M, then I also need to increase the value of the capacitors C4 / C5 to 20n?

 

[abbey road d enfer]

I welcome everyone.
I do not have the values of resistors R5.R6 - 100M, but there are 200M.
If they are used in this circuit 200M, then I also need to increase the value of the capacitors C4 / C5 to 20n?

The value of these resistors is almost irrelevant. You don't need to change the value of capacitors.

 

[vmanj]

Thank you.

Just for an example: in the Sony c800g circuit the capsule voltage divider, these are small resistor values and large capacitor values.

In the circuits of most other microphones, on the contrary, these are large resistor values and small capacitor values.

I thought that an accurate calculation of these values is needed here.

 

[abbey road d enfer]

Just for an example: in the Sony c800g circuit the capsule voltage divider, these are small resistor values and large capacitor values.

In the circuits of most other microphones, on the contrary, these are large resistor values and small capacitor values.

I thought that an accurate calculation of these values is needed here.

The role of this combination of resistors and capacitors is to filter out any residual noise on the bias voltage. This noise is a combination of AC mains frequency residues (100/120Hz and harmonics) and some low level wide band noise.
The typical combination of 100Meg and 10nF gives an attenuation of 50dB at 50 Hz, which is largely enough if the PSU is half-decent.
A combo of 1 Megohm and 1uF would give the same result.
Nothing very complicated here.

 

[vmanj]

So the combination of 10K and 100uF will be the same?

 

[RuudNL]

It is the R x C product that counts.

 

[vmanj]

Thanks, I'll try to figure this out.
And I tried to combine Sony c800g and your filter with feedback.
I drew 2 conditional simplified versions of the schemes.
Can you please tell me if this will work or not?

 

[abbey road d enfer]

So the combination of 10K and 100uF will be the same?

Your 100uF capacitor would probably be an electrolytic, innit?
You have to know that 'lytic capacitors have a much higher leakage current thna film or even ceramic caps. This leakage current can surely load the voltage multiplier, and result in a loss of bias voltage.
What's more, leakage tends to increase with age.
So I would not recommend this combonation.

 

[vmanj]

Your 100uF capacitor would probably be an electrolytic, innit?
You have to know that 'lytic capacitors have a much higher leakage current thna film or even ceramic caps. This leakage current can surely load the voltage multiplier, and result in a loss of bias voltage.
What's more, leakage tends to increase with age.
So I would not recommend this combonation.

It's clear, but I'm looking at the s800g circuit, and it works fine ..

 

[kingkorg]

The role of this combination of resistors and capacitors is to filter out any residual noise on the bias voltage. This noise is a combination of AC mains frequency residues (100/120Hz and harmonics) and some low level wide band noise.
The typical combination of 100Meg and 10nF gives an attenuation of 50dB at 50 Hz, which is largely enough if the PSU is half-decent.
A combo of 1 Megohm and 1uF would give the same result.
Nothing very complicated here.

I haven't tested this yet, but could there be a difference in case of capsule overload, hard plosives and or moisture from breath (k67). Recovery wise. @abbey road d enfer @RuudNL

 

[abbey road d enfer]

I haven't tested this yet, but could there be a difference in case of capsule overload, hard plosives and or moisture from breath (k67). Recovery wise. @abbey road d enfer @RuudNL

The lower the resistor values there, the less they are sensitive to overload effects.

 

[vmanj]

RuudNL -​

Thanks for the schematic.
I simplified it a little (on the image), but it works!
I have a question about the R1 10K resistor.
If you increase or decrease its value, how can this affect the operation of the feedback filter?

 

[abbey road d enfer]

I have a question about the R1 10K resistor.
If you increase or decrease its value, how can this affect the operation of the feedback filter?

The amount of HF negative feedback is directly proportional to the value of this resistor, so if you increase it it decreases treble.

 

[RuudNL]

What is the anode voltage with +230V? (It could be that you will have to adjust R7 (1K).
Is there a special reason for a polarisation voltage of 47 Volt?
(If you would raise it to 60 V. you would gain 2.1 dB signal to noise ratio.)

 

[vmanj]

I tried the U67 (EF86) circuit with a fixed bias but without high frequency compensation and Sony c800g (6AU6) and 47v on the capsule.
Based on my capsule and transformer (cad trion 8000), I liked the sound more according to the Sony scheme. I tried to raise the voltage on the 60v capsule, there is a little less noise, but I liked the sound less than 47v.
I decided to dwell on it, but add a filter from your circuit to the Sony circuit ..

 

[kingkorg]

Try to place the microphone at exact same position, preferably not move it at all. Record same exact sound source with two different pol.voltages. I play pink noise and some music through monitor or raw guitar tracks through a cab. Then blind shootout the files, do null test with both recorded files.

My experiments show the files cancel out perfectly at these voltage ranges ->exactly the same. Noise and max spl of course differ. The catch is not to change anything besides pol.voltage. i don't even breathe, nor move during the test. Go frim lower to higher voltage.

 

[vmanj]

Try to place the microphone at exact same position, preferably not move it at all. Record same exact sound source with two different pol.voltages. I play pink noise and some music through monitor or raw guitar tracks through a cab. Then blind shootout the files, do null test with both recorded files.

My experiments show the files cancel out perfectly at these voltage ranges ->exactly the same. Noise and max spl of course differ. The catch is not to change anything besides pol.voltage. i don't even breathe, nor move during the test. Go frim lower to higher voltage.

I did not conduct such tests, this is on a subjective level .. It's just that 47v on the capsule did not seem as sharp as 60v.

 

[vmanj]

And if you connect the capsule according to a different scheme (C12), then it will be possible to use this feedback filter as in the image?

 

[LevinGuitar]

I have tried your U67ish de-emphasis in a similar mic circuit but if I connect the 120pf after the output capacitor I get noise (probably from B+ because if I turn off the psu, the noise dissapears). If I connect the cap to the anode, is ok but the result is different, I like it more after the capacitor, as you have in your schematic.
I have the 10K +1uf filter before the divider and anode resistor, how could I filter it anymore?
For now I found a solution placing a small resistor between 120pf and 10nf, but guess I have a bit different effect. Even this way it works nice.

Thank you!

Just for fun, I recently put together a "U67"-ish microphone.
It is a kind of mix between a U67 and a U87.
The microphone works as expected, sounds good and I don't hear any difference between this microphone and a U87ai.
Measured self noise is 2 dB higher than a U87ai. Not too bad, but maybe a better EF86/EF806 could improve this.
I used an original Neumann capsule that I had as a spare.
The microphone is powered from an existing tube power supply, delivering +120 V for HT.
De-emphasis is through feedback from the output. (15 KHz is 5.5 dB down referred to 1 KHz.)

For those who are interested: the schematic is in the attachment.

 

[Aleguitarpro]

Hi.
I think this is a great project.
I'm going to build this with a body without switches on it so I'm thinking to use REMOTE PATTERN from PSU.
I modded schematic. I don't know if it's ok.
I'd like to build it but I'd like to remove FILTER and PAD, too.

Could it work like this scheme attached?
To test it without feedback (to hear Flat capsule response) can I only wait to install C9 or with this pattern arrangment C5 need to be grounded to work without C9 installed?

Thanks