V76 gain control working principle

V76 type of gain control is used for many tube diy projects and pro gear. Like Tele V76, Lorenz V241 nad revision here, Ian Bell's EZ preamps, etc.
Gain is often only partially variable NFB like in UA, EMI, Pultec, although their sound can change to better through small range...
http://www.wagner-microphones.com/images/v76.gif
To understand it better i took R55, 57, 58  and C12 on schematic of Tele V76. Are R57 and R58 biasing resistors and divider for R55, how is formed dominant pole? Looks like C12 closes NFB at DC.

V76's gain control changes closed loop gain and gain of tube. I find positions sound more even, with mods pad can still be used.
Ian's EZ manual shows another example on p.7, where it is very handy for ECC83's low distortion because output stays low. Many of us thought this tube isn't good enough for inputs, while at least his tests and implementation as here show it is.
https://drive.google.com/file/d/0B_n67A1hN3qtWlo2ZEpxWnAxQUE/view?usp=sharing

R57 sets the dc bias for the first stage. R57 + R58 set the minimum stage gain. C12 ensures the transformer is connected between the grid and 0V.

Stage gain is set roughly by the ratio of the plate resistor (R59) to R57+R58 plus whatever is in parallel with them at ac which is mostly the gain control pot. At low gains this looks like about 30K so it only increases the stage gain a little. At maximum gain this drops to 90 ohms so the first stage gain increases significantly.

Overall gain is set by the ratio of the negative feedback resistor, which is the top half of the gain switch plus R53 (40K) and R57+R58 and the bottom half of the gain switch.

The basic idea is to keep the loop gain, which is the open loop gain plus the gain of the feedback network, relatively constant because this is a major factor in determining the stability of the circuit. The other factor is the number of poles and zeroes in the loop

In practice it is a little more complex than that but the basic principle is as I have described it.

Cheers

Ian

Thanks, very good explanation.
C13 50uF (with pot) is in parallel with cathode resistors. Does it frequency compensates gain positions when pot changes resistance?
R55 100k  is typical value and there in many gain controls of this type. Value seems too high for doing anything for cathode bias or gain, is it related to grid 1?
Next i will dive into iz EZ preamp. Looks simpler when V76 is understood, while showing how ECC83 works as fine and cheap input tube.

My3gger said:

Thanks, very good explanation.
C13 50uF (with pot) is in parallel with cathode resistors. Does it frequency compensates gain positions when pot changes resistance?

Click to expand...

Frequncy response will change a little with gain but the worst case 90 ohm resistor has a 3dB point of about 35Hz with 50uF and I suspect they were not too much worried by that

R55 100k  is typical value and there in many gain controls of this type. Value seems too high for doing anything for cathode bias or gain, is it related to grid 1?

Click to expand...

Its job is to connect the grid to the bias point at dc. C12 shorts it to ground at ac. It only needs to be high enough not to make much difference  to R58 which is part of the NFB network.
quote]
Next i will dive into iz EZ preamp. Looks simpler when V76 is understood, while showing how ECC83 works as fine and cheap input tube.
[/quote]
The basic principles are the same, just the particulars of the implementation are different.

Cheers

Ian

Rather than start more topics...

I was reading RDH4 today and the V76 R55, C12, R56 and R57 seems like what is described to give a partial bypassed cathode resistor.  The bias equation checks out for 7.5 uA at 11V  to  a 15K cathode resistor.    Anyone have any thoughts?

guavatone said:

Rather than start more topics...

I was reading RDH4 today and the V76 R55, C12, R56 and R57 seems like what is described to give a partial bypassed cathode resistor.  The bias equation checks out for 7.5 uA at 11V  to  a 15K cathode resistor.    Anyone have any thoughts?

Click to expand...

See my original reply to the first post.

Cheers

ian