diytube.com

I'm working on the phono amp now (I got lots of hum) and the cathode follower I'm using is a 6N3P. When I check the voltage on top of the 100K with the other 2 tubes pulled, it varies a little, 150-151. When I check the grid, the voltage starts at 130, and goes down to 60 volts. It will do it over and over again. I guess my multimeter is loading the grid down, but I though the 1M would be able to keep it up. I tried a 470K, but it still goes down. My multimeter is a Fluke 87.

Leaky coupling cap from previous stage? Try a new tube?

Both sides are doing it, and I tried another 6N3P with the same results. The capapacitor wouldn't be leaking, because the voltage would be too high, and the capacitor is a new panasonic ecwh polypropylene.

That coupling cap is a 0.1uf, I guess I could lower that 1M resistor to 300K with no loss of fidelity.

Try a grid stopper in there? I had a CF like that go into UHF parasitics.

Snap - Geek posted as I was about to.
AC Coupled Cathode Follower info:
http://www.valvewizard.co.uk/accf.html
Tube Cathode Followers, MOSFET Source Followers, Transistor Emitter Followers all need grid/gate/base stopper resistors.
For 6N3P try a 4K7.
If the circuit is on a Printed Circuit Board then cut the track to the grid right at the tube grid pin pad and solder a surface mount resistor across the track cut.
With 150 odd volts at the cathode (seems high) you will need to elevate the Heater Supply to say +80 or +90 Volts so as to not exceed the heater cathode maximum voltage spec.

Tube Datasheet for reference:
http://www.tubes.ru/techinfo/HiFiAudio/6n3p.html

Your measurement of 150V suggests 1.5mA tube current. 6N3P like more than that. Are you sure this circuit was'nt intended for 6N2P.

Cheers,
Ian

I have a scope on it, and there's only a jagged 60 120Hz. I removed the preamp tubes and there's about 1mvpp of AC ripple. The schematic is a copy of Morgan Jones' cathode follower, except he used a ECC88. I have to make another circuit with a 6BN8 as a buffer, I'll see how well that one goes and compare.