Grounding the cathode in a tube microphone

[MagnetoSound]

Hi @rock soderstrom
Because we are both on a "Sela mod project" something caught my attention : on the schematic you posted it seems that polar. volatge is applied on backplate and signal taken from front diaphragm... But souldn't it be the opposite ? I'm just asking...

You can polarise the diaphragm, but then you need a blocking cap before the grid.

Opinions vary as to whether the difference is audible, but it adds to the parts count nonetheless, and is no longer the Sela design.

 

[Emmathom]

You can polarise the diaphragm, but then you need a blocking cap before the grid.

Opinions vary as to whether the difference is audible, but it adds to the parts count nonetheless, and is no longer the Sela design.

You mean it's when you polarize the backplate and take the signal from the diaph. that you don't need a coupling cap ?
So Sela design implied to polarize the backplate ?

 

[Emmathom]

So:
Pin 1 = signal from capsule
Pin 2 = R cathode to ground
Pin 3 & 4 = heater (6,3v@175mA)
Pin 5 = to anode (90v@3,5mA)
Pin 6 = connect to anode > output
Pin 7 = internal connexion with cathode (no further connexion)
Is it correct ?

 

[abbey road d enfer]

I think both is possible.

Of course.
Actually, biasing the backplate instead of the diaphragm prevents the diaphragm attracting dust.

 

[abbey road d enfer]

You mean it's when you polarize the backplate and take the signal from the diaph. that you don't need a coupling cap ?

Correct.

So Sela design implied to polarize the backplate ?

See how the "Kapsel" connects to B+ via R3, R4 & R5.

 

[Cqwet Dbdfte]

For fixed bias one could use an IR LED (~ 1.1V) in series with the cathode to ground
Very stable, low noise, low component spread.

 

[Emmathom]

For fixed bias one could use an IR LED (~ 1.1V) in series with the cathode to ground
Very stable, low noise, low component spread.

You mean an infra-red led ?
What about 2 x 1N4148 in serie (2*0,6v) ?

 

[tomas.borgstrom]

You can polarise the diaphragm, but then you need a blocking cap before the grid.

Opinions vary as to whether the difference is audible, but it adds to the parts count nonetheless, and is no longer the Sela design.

I’ve never checked if it’s true but I was ones advised to always have a DC blocking capacitor between the membrane and grid since the capsule can leak a small amount of DC and influence the grid voltage.

 

[abbey road d enfer]

I’ve never checked if it’s true but I was ones advised to always have a DC blocking capacitor between the membrane and grid since the capsule can leak a small amount of DC and influence the grid voltage.

A leaky capsule creates other problems than simply altering tube bias. It plays bad with sensitivity and increases noise. A capacitor may hide one issue but not solve the other.
And, haven't you heard? Capacitors in the signal path are EVIL.

 

[Cqwet Dbdfte]

You mean an infra-red led ?
What about 2 x 1N4148 in serie (2*0,6v) ?

https://www.diyaudio.com/community/threads/led-tube-biasing-pros-and-cons.324722/post-5482951
No bypass caps are needed.
Of course other diodes can be used.

 

[Cqwet Dbdfte]

A leaky capsule creates other problems than simply altering tube bias. It plays bad with sensitivity and increases noise. A capacitor may hide one issue but not solve the other.
And, haven't you heard? Capacitors in the signal path are EVIL.

It is hard to avoid using caps in the signal path, some argue that bypass caps are in the "path" too.
I designed and built a couple of p-p 200W 211 triode amps 30+ years ago with no coupling caps. Each stage voltage was stacked on top of the next, with shunt regulators providing a virtual ground for the subsequent stages.
Was it worth it, nah, too much power for domestic use, a sub does the heavy lifting. Lots of DC servos and protection circuits.
It sounded good no doubt.

 

[abbey road d enfer]

It is hard to avoid using caps in the signal path, some argue that bypass caps are in the "path" too.

I've always said that bypass caps are as much in the signal path than coupling caps, because signal propagates in a loop.

It sounded good no doubt.

I don't doubt it. Maybe a tad over the top but certainly an interesting exercise.

 

[MagnetoSound]

Yes, bypass caps are definitely in the signal path.

Coupling or decoupling, the signal is present.

 

[mjmajilton]

You mean an infra-red led ?
What about 2 x 1N4148 in serie (2*0,6v) ?

Yes. Two switching diodes works well too.

 

[Cqwet Dbdfte]

LED tempco would be better, but on this application not important.

 

[Khron]

For what it's worth:

https://groupdiy.com/threads/microphonics-in-vf14-and-u47.82644/post-1128559

 

[terry setter]

Yes. Two switching diodes works well too.

I've tested many LEDs between cathode and ground and they are noisy. NOISY. Use two UF4007's in series for a 1V bias and you'll be VERY happy.

 

[Cqwet Dbdfte]

I've tested many LEDs between cathode and ground and they are noisy. NOISY. Use two UF4007's in series for a 1V bias and you'll be VERY happy.

What tube, how much current?

 

[vmanj]

Well, this happened..

My experimental microphone without a grid resistor stopped working correctly...
Possible high or low humidity or magnetic storms on Saturn..

In general, I returned the grid resistor, but first experimented with its value.

I tried different values from 10 to 1000 Mohm.
This greatly affects low frequencies and noise.
I liked the 500 Mohm the most.

But the cathode remained simply grounded, the sound is very similar to a normal fixed bias but without additional external voltage on the grid or cathode.
It seems that -0.5 Volts on the grid is enough for this "strange" microphone.

I don’t want to switch to cathode self-bias anymore, this
I like the sound much better..

 

[terry setter]

I designed a popular mic that had a grounded cathode. It could use a 12AT7 or 12AY7/6072. The sound was great, the noise was low, and I never saw ANY issue with distortions associated with bias being insufficient. I asked David Bock and Oliver Archut about it and they both knew of mics in history that did the same thing. Try it, you might really like it!