So, I was talking to a friend of mine about an idea I had to use a microcontroller to automatically adjust bias of output tubes on a fixed bias amp. I'm not sure he actually knew what he was talking about, but he said, "Hey, I think I heard you can use an opamp for that." That sounded vaguely like it might actually work, so I did some Googling to try to see what other people were doing. I couldn't find anything. Google "using an opamp to bias a tube amp", and nothing relevant comes up. So I decided to poke around some on my own.
This actually works in simulation:
Pretend that the mosfet is actually a power tube. R1 is a current sense resistor. It does the same basic thing as the 1 ohm resistors you guys put in your amps to check the bias and really is there for the same basic reason. It's chosen to be small enough that it doesn't contribute significantly to the bias itself, but large enough that the reference voltage doesn't have to be manipulated with millivolt precision. V2 is the reference voltage. You set the reference voltage to the same voltage you want to see at the junction of cathode (drain) and R1. Given the voltage at that junction you know the current. So, let's say you have a pair of 6L6s you want to run at 40mA/piece for 80mA total. If R1 is 10R, then your target reference voltage is 0.8V. The output of the opamp is referenced to ground through R4 and fed through to the grid (gate) as the bias voltage. The opamp compares the voltage at the cathode with the reference voltage and uses negative feedback to make them very close to the same number. For a high gain opamp, you probably can't measure the actual difference. As you can see, if cathode voltage rides up, then the output of the opamp goes down, which biases the tube (mosfet) cooler, so the cathode voltage rides back down. That's the negative feedback action. C4 is there to look like a short at audio frequencies, so your signal doesn't wind up trying to change the bias. Similar to a cathode bypass cap. The cutoff frequency (I think) is f = 1/(2 * pi * R2||R4 * C4). But I'm not confident in that analysis. Works in sim with current values, though. Smaller grid leak should mean bigger C4. The net effect is the bias is controlled solely by your reference voltage and is independent of the characteristics of your individual tubes, so you can change power tubes at will and always get the same bias current.
Anyway, obviously not actually tried in a tube amp. But I don't see why it couldn't work. As a practical consideration, you have to make sure that your opamp can actually get to the target bias voltage. Which might mean adding a mosfet to amplify the opamp. The mosfet would probably invert the opamp output, so be sure to swap the +/- inputs to the opamp. Negative feedback makes this thing work. If you accidentally swap something and get positive feedback you could completely meltdown your power section.
Anyway. Just thought it was a cool idea.
Topic: Using an opamp to automatically adjust fixed bias on a tube amp (Read 4085 times)