C12 Grid Voltage

I'm working on building a couple of C12s from scratch, but I am a little bit confused about the grid voltage. People seem to make a big deal about it, but is it as simple as sending a certain amount of voltage into it? I've been trying to follow the schematic, but it doesn't say anywhere the amount of voltage being sent into it. It looks like it's coming off of pin 4 in the psu. How much is actually going into it?

For anybody else who is wondering, my Telefunken 251 build came out amazing! It sounds absolutely fantastic.

If you are asking about the grid-cathode bias, yes the grid has a small negative voltage applied to put the tube in a operating point - since the cathode is connected directly to ground. (The more negative the grid with respect to cathode the more the tube conducts)
The bias voltage for the grid comes from connection 4 like you say, going back to the connection at R8 (1k) which creates a small voltage change between two grounds. The cathode is connected to the ground before R8 (1k). The returning current from the mic (say ~.0012Amps) goes through R8 which drops a small voltage (~.0012*1k=1.2v) This voltage is then the bias on the grid.

I just built a C12 inspired mic, but used the polarization of a C24, which uses a pulldown to ground for the grid and a standard cathode RC. There a 2.5k resistor is used on the cathode. If the tube current were the same (.0012A), the grid-cathode bias would be -3v.

dmp said:

(The more negative the grid with respect to cathode the more the tube conducts)

Click to expand...

The more negative the grid with respect to cathode the less the tube conducts.

Easy confusion - as the grid goes negative with signal modulation, the anode current drops, producing a rise in signal voltage at the output cap, and vice versa. 

Right - thanks for the correction

I've been wondering if it makes more sense to convert a heater supply to a full wave rectifier, then pull a negative reference and regulate it with a LM337 and send that out for tube bias.  This decouples the grid bias from the tube operating point (which is a dependent parameter), and lets the user set the tube idle/operating point to the "best" value from tube to tube.

The MK47 design by Max (ioaudio) use the heater current to 'stiffen' the cathode bias and omit the standard bypass cap. Meaning the changes in the cathode current due to the tube operating point changing with signal do not affect the cathode voltage as much, due to the lower R. Much of the cathode voltage bias comes from the heater current.
It is interesting and something I'd like to experiment with. In the MK47 the B+ is used for the heater.

I was speaking more to the C12 circuit which inserts a resistor between ground and the negative half of the rectifier.  In a sense, this is similar to cathode bias:  plate current dumps to ground, and drops across this resistor on it's way to the transformer, which creates the negative bias.  As plate current increases, negative voltage increases which lowers plate current.  It's a similar negative feedback mechanism of cathode bias.

However if we derive a constant negative voltage (independent of plate current, much like what would be used on a guitar power amp), then the bias point could be explicitly set by the user rather than by the tube.

I'm more wondering if there is an advantage to doing so, rather than arguing that it is the right answer.

As plate current increases, negative voltage increases which lowers plate current.  It's a similar negative feedback mechanism of cathode bias.

Click to expand...

R10 and C6 should dampen this significantly, yes? I see what you are saying though.
I have not built nearly enough mics to contribute an opinion

Matador said:

However if we derive a constant negative voltage (independent of plate current, much like what would be used on a guitar power amp), then the bias point could be explicitly set by the user rather than by the tube.

I'm more wondering if there is an advantage to doing so, rather than arguing that it is the right answer.

Click to expand...

Interesting point. I suppose it depends whether you are really talking about controlling the stiffness of the bias voltage itself, or about setting an ideal nominal bias point for the type of tube in use.

Bear in mind that in the U47 circuit a quite substantial signal level is required (I haven't measured it) before the bias will drift significantly - although this is not to say it is not a legitimate consideration if the mic is to be put to use at high SPL.

At the end of the day, your ears will decide, no?  ;D

Matador said:

I was speaking more to the C12 circuit which inserts a resistor between ground and the negative half of the rectifier.  In a sense, this is similar to cathode bias:

Click to expand...

The C12 bias scheme is EXACTLY the same as cathode bias!

It's just that the cathode resistor is in the power supply.

AKG probably intended to make a power supply with fixed biased at some later stage but found the 'secret cathode' bias so effective that they never bothered.  In C24, they let the secret out.

Cathode bias is much less critical than fixed bias and compensates for different tubes and aging to a large extent so your mike lasts longer before you need to tweak it or change tubes.

Then there wouldn't be a problem if I copied the C12 circuit component for component, but biased the tube using the same method that is used in the C24?

Edit 2: Ricardo, I don't see how it could be the same. On the C24, the cathode is connected directly the ground with the resistor and capacitor in parallel. It doesn't even touch the grid.

On the C12, the grid and cathode are connected with a resistor and capacitor in series.


Edit 3: If I understand this correctly, the easiest way to bias the grid would be to wire a 1k trim pot in between ground and the grid, and adjust to -3V?

I agree totally with Ricardo.  But you have to look hard (and perhaps long!) at the C12 schematic to see that, although topologically different, the method is electrically the same as that used by the C24, or by the Aurycle tube mics. http://www.aurycle.com/manual/a5500_schematic_clean.jpg

If you have two resistors in series with a d.c. voltage source, the voltage drop across any one resistor is the same whether it has one end connected to the +ve terminal of the source, or to the -ve terminal.  Of course, the absolute voltages are different, but the resistor only sees the relative value from end to end.  In the case of the tube, the potential difference between cathode and grid can be made exactly the same as in the original C12 circuit by adopting the method used in the C24 circuit.  As with the resistor, the only thing that matters in the C12/C24 circuits is what the tube sees, not the absolute voltages.

I explained the configuration thus in an earlier thread:

Look carefully at the C12 circuit, particularly at the 1k resistor, R8. It is true that the cathode is connected directly to ground, but the grid resistor, R15 is not: it is connected to R11 and thence, via R10 to the negative end of R8. The other end of R8 is connected to ground. There is current passing through R8 -- actually the return from the HT or B+ line to the bottom end of the lower winding of the transfomer. This makes the left hand end of R8 negative with respect to ground. Thus the grid is also negative with respect to the cathode, as is normally achieved by means of a resistor in series with the cathode. (See the C24 circuit.) The method of achieving the bias is not unlike that in the U47 circuit which also relies a on a voltage across a resistor (R3) fixed by a larger current also flowing through it -- in that case the heater current.

The connection between R10 and R11 may be obscured if you are looking at the version of the C12 circuit on which someone has kindly drawn a fat vertical line with arrow heads on it.  This one is cleaner: http://www.groupdiy.com/index.php?action=dlattach;topic=44737.0;attach=8315

The method used in the C12 just places the resistor that would normally be connected between cathode and ground to the position of R8.  See http://www.hans-egebo.dk/Tutorial/biasing.htm, where it is described as "Power supply backoff".

David

I agree:  the tube doesn't care.  The negative feedback is still there as I explained.

However the question remains:  does splitting off and providing a low impedance fixed bias voltage (that DOESN'T move with plate current) provide any sonic advantage?  I would guess that gain/distortion might be slightly higher since the NFB is essentially removed.