TheWinterSnow
Den Mørke Natt
Well both ways can be correct, but it depends on what value you want to use and how you want it to react. I would do it the way you have mentioned, since the capacitance from grid to anode will get multiplied by the voltage gain of the stage. So a lower value cap can be used. You could still put a cap from grid to cathode, but you would need a much higher value cap to get any reasonable difference.
I agree. In the rig talk thread they where eventually talking about using a 100pF cap from the cathode to grid or a 10pF from the plate to grid. Since the Plate to grid capacitance will be more extreme, you wouldn't want to use such a high value.
FYI for everyone going to try this, I had a thread about this over on a builders forum a while back. There was a big discussion on putting caps in these positions, and after messing with it on my breadboard, the consensus was that it's better to achieve the effects in other ways throughout the preamp. It just didn't sound right/do enough. In fact, I think a lot of the guys over on the 'other forum' that did this mod are just 'thinking' they are hearing a difference. Because that one cap, in that spot, at that value, shouldn't make to crazy of a change.
I was looking at the datasheet on my tubes and the capacitance from the plate to grid is 10pF, so adding a 10pF cap would increase the total to 20pF. Now unfortunately I forgot how to calculate the bandwidth of a gain stage from eddy currents and miller capacitance, but I can imagine doubling the capacitance would provide at lease something.
The only other means of decreasing the bandwidth without dramatically effecting the gain of that stage would be to increase the grid resistor. You have any other recommendations Wolfe?