Grounding the cathode in a tube microphone

GroupDIY Audio Forum

Help Support GroupDIY Audio Forum:

This site may earn a commission from merchant affiliate links, including eBay, Amazon, and others.

vmanj

Well-known member
Joined
Oct 3, 2017
Messages
322
Hello everyone.

Previously, I had already made different fixed biases in microphones, everything was fine and worked.

But this time I decided to just remove the cathode resistor.
For some reason everything works and sounds good.

Why does this happen and can it destroy the tube?
 

Attachments

  • СATHODE GROUND.jpg
    СATHODE GROUND.jpg
    42 KB
Destroy the tube, maybe not(?), but i've got a hunch it will definitely clip / distort earlier than before (grid-to-cathode voltage now being 0v, as opposed to minus-whatever-it-was-previously).

... But i could very well be wrong, i'm definitely no tube expert 🤷‍♂️
 
Because the 560MΩ Grid-leak resistor is way too large and develops Grid-leak bias (also known as contact bias)…

Now that you know “Why does this happen” put back in the cathode circuit the 2.366Ω Cathode resistor for the properly Cathode bias, before you promote blocking distortion & maybe destroy the tube…
 
Because the 560MΩ Grid-leak resistor is way too large and develops Grid-leak bias (also known as contact bias)…
I see, I read that such a leakage resistor can create a bias voltage of about -0.2 volts.
But this is not enough for the normal operation of the circuit.

Now I measured it and I have -0.5 volts on the grid.
This, of course, is also not enough.


Now that you know “Why does this happen” put back in the cathode circuit the 2.366Ω Cathode resistor for the properly Cathode bias, before you promote blocking distortion & maybe destroy the tube…
I have several microphone options with different types of bias according to the correct patterns.

I removed this cathode resistor for experimentation, and not to put it back in..

The thing is, I like the sound, it's similar to a regular fixed bias, but with no additional voltages on the grid or cathode..

Now I have only one question, how much damage can this do to the tube?
 
Perhaps I measured incorrectly.
I'll try with another tube later.
 

Attachments

  • CATHODE GROUND GREED VOLTAGE.jpg
    CATHODE GROUND GREED VOLTAGE.jpg
    43.4 KB
I suspected this, but nevertheless, now the microphone works and sounds good.
Just wondering how quickly this will kill the tube..
 
I guess it works with grid leak bias which was used quite a bit in the old days on small signal tubes like an ef86 combined with a 10Mohm grid resistor. A friend of mine rebuilt such a circuit (from an old Grundig tape machine, if I remember right) into a wonderful guitar amp.
The Royer mod for small diaphragm mics relies on grid leak bias for the tube and it works great.
I guess the u67 would also work with simple grid leak biasing.
 
I’ve measured the grid leak in a couple of different ECC81 in a C12 (with a high impedance voltmeter). Turns out it varies quite a lot between different brands. The tube that sounded best was a Valvo and it had a bias of -1.1V wich resulted in a similar current draw as 6072.

So If you design a microphone with grid leak bias you have to be able to measure voltage at high impedance and find a tube that gets a decent bias in that circuit or try to change the B+.
 
You can use a FET input op-amp (something like an OPA2134) wired as a voltage follower, to carefully measure the actual bias voltage at the gate. I have one expressly for this purpose that runs on a pair on 9V batteries so it can measure both positive and negative grid biases (at least within the +-9V common mode input range).
 
I continue the circus...

I removed the grid resistor, reduced the anode voltage to 60V, and tried different tubes.
Of course, I'm measuring incorrectly, but my multimeter shows -0.5V in different versions.
But the most interesting thing is that without the grid resistor, there is subjectively less noise...
The sound is good, I don’t understand why this “wrong” circuit works?
 

Attachments

  • CATHODE GROUND GREED VOLTAGE 02.jpg
    CATHODE GROUND GREED VOLTAGE 02.jpg
    42.2 KB
I continue the circus...

I removed the grid resistor, reduced the anode voltage to 60V, and tried different tubes.
Of course, I'm measuring incorrectly, but my multimeter shows -0.5V in different versions.
But the most interesting thing is that without the grid resistor, there is subjectively less noise...
The sound is good, I don’t understand why this “wrong” circuit works?
You are feeding the capsule half the voltage, too, that could explain the lesser noise.
And there ist the David Royer mod, which does not use a grid resistor in a cathode follower circuit.
 

Attachments

  • Screenshot 2024-06-08 at 01.53.09.png
    Screenshot 2024-06-08 at 01.53.09.png
    344.1 KB
Well, some (more) answers to your questions…

The Grid-leak bias (also known as contact bias) was an “ancient” type of biasing, way back in these “ancient” old days of electronics, when the tubes were too inexpensive and the electrolytic capacitors were too big and expensive to fit them in a microphone…

This “ancient” biasing method develops a negative voltage on the grid by using a very large grid resistance.

During normal operation, a few electrons will always strike the grid and bleed away down the grid-leak resistor.

If the resistance is large enough, the negative voltage developed on the grid due to this leakage current flowing in the grid-leak may be large enough to self-biasing the valve (tube)…

However this “ancient” method is almost never used in our newest -and more precision…- audio today, for many reasons:
  • It varies to much with valves samples.
  • As well as collecting electrons, the grid also collects positive ions and also emits electrons.
  • The rate of emission depends largely on temperature and contamination of the grid, over which the designer has no control…
  • It requires a very large resistance, which introduce too much unnecessary resistor noise and makes the stage susceptible to heater hum…
  • It is rather unpredictable.
  • Choosing the resistance involves little more than guesswork, and the voltage developed tends to wander as the valve ages.
  • As an “approximation”, most small signal valves will only develop -0.1mV per Meg-Ohm when the valve is new, which severally limits our biasing range…
  • The large Grid-leak resistance is liable to promote blocking distortion
  • It does not offer the frequency-shaping advantages of the cathode biasing…
 
I continue the circus...

But the most interesting thing is that without the grid resistor, there is subjectively less noise...
That is normal.
The noise spectrum is governed by the product of the grid resistor and the capsule's capacitance.
Although the noise energy increases with resistance (check KTC noise), its spectrum moves down in frequency. The net result is that most of the noise is moved down to infra sounds.

The sound is good, I don’t understand why this “wrong” circuit works?
There is a limit to the voltage developed by contact bias. Increasing the grid-leak resistor above about 10 Gohm results in no significant change.
That's because the cathode, being less negative than the grid, becomes an anode for the grid.
There are a few microphones that completely did without a grid resistor.
 
Back
Top