cathode for plate ccs ?

[jj2]

I have finished a 12B4 pre with plate ccs 10m45 running at 24ma.
intalled cathode resistor and cap (395 ohm , 100uf).
here is my problem, have 300v going into ccs but only 65v going into the tube and how do I calculate the value of the cathode resistor and cap ? or do I need it at all ?
Thanks John

[Brian Beck]

A triode has a fixed relationship between its plate voltage, its plate current and its grid voltage. You can force any two of these parameters to specific values, but the third must be allowed to find the value that the tube “wants” (this is called a “degree of freedom”). In your case, you’re forcing plate current to be 24mA with the plate CCS, and that’s a good value. Since that same current also flows through the cathode resistor, we know that there must be about 9.5 volts on the cathode, by Ohm’s law. Since the grid is presumed to be referenced to ground, the grid bias is equivalent to -9.5 volts. So you’ve now forced two out of the three parameters, and the tube gets to pick the third - the plate voltage - to satisfy its internal “requirements”. If you were measuring 65 volts on the plate referenced to ground, then the real plate voltage (which must be referenced to the cathode because that’s all the tube can “perceive”) is 65-9.5 = 55.5 volts. This seems a bit low according to the standard plate curves for this tube, but units do vary. The tube will work at that low plate voltage, but it may be better to make it go higher. I run mine at about 120 volts. Keeping the plate current at a reasonable 24 mA, you’ll need to increase the cathode voltage (which is the negative grid bias) to somewhere around 15 volts, which will let the plate then go higher. Since you’ll have about 15 volts on the cathode, and if you want 120 plate-cathode volts, you’ll be looking for around 135 volts on the plate referenced to ground. But it’s not critical. You might try something around 620 ohms in the cathode and see what happens (keep the 100uF bypass in the circuit). You can then tweak that value a bit to get closer to what you want. Anywhere above 100 volts is probably OK.

[jj2]

Originally posted by Brian Beck
A triode has a fixed relationship between its plate voltage, its plate current and its grid voltage. You can force any two of these parameters to specific values, but the third must be allowed to find the value that the tube “wants” (this is called a “degree of freedom”). In your case, you’re forcing plate current to be 24mA with the plate CCS, and that’s a good value. Since that same current also flows through the cathode resistor, we know that there must be about 9.5 volts on the cathode, by Ohm’s law. Since the grid is presumed to be referenced to ground, the grid bias is equivalent to -9.5 volts. So you’ve now forced two out of the three parameters, and the tube gets to pick the third - the plate voltage - to satisfy its internal “requirements”. If you were measuring 65 volts on the plate referenced to ground, then the real plate voltage (which must be referenced to the cathode because that’s all the tube can “perceive”) is 65-9.5 = 55.5 volts. This seems a bit low according to the standard plate curves for this tube, but units do vary. The tube will work at that low plate voltage, but it may be better to make it go higher. I run mine at about 120 volts. Keeping the plate current at a reasonable 24 mA, you’ll need to increase the cathode voltage (which is the negative grid bias) to somewhere around 15 volts, which will let the plate then go higher. Since you’ll have about 15 volts on the cathode, and if you want 120 plate-cathode volts, you’ll be looking for around 135 volts on the plate referenced to ground. But it’s not critical. You might try something around 620 ohms in the cathode and see what happens (keep the 100uF bypass in the circuit). You can then tweak that value a bit to get closer to what you want. Anywhere above 100 volts is probably OK.

replace cathode resistor with 400 ohms and plate voltage went up to 102V , cathode voltage is 9.65v.

Thanks again.
John

[poynton]

Brian,

Well done and thanks for a patient, well mannered explanation.

So often on the forum, the answer to a question like this is unhelpful, usually along the lines of " Use the search button !!".


Andy