12AT7 LTP PI

[Torquil]

Hi!

I think I'm having a problem with an Electroharmonix 12AT7 LTP PI on a homebrew amp. The voltage difference between grids and cathode is very large, around 20V. Still, the amp works fine at low volumes. I havent turned it up.

I'm using the circuit (and notation) here:
http://www.freewebs.com/valvewizard/acltp.html

HT/B+ = 460V
Ra1 = Ra2 = 47k
Rt = 10k
Rb = 270
Rg1 = Rg2 = 470k

Voltages:

Plate ~ 280V
Grids ~ 60V
Cathode ~ 80V
Utb ~ 78V (between Rt and Rb)

Anyone know why I get such a large negative bias, around -20V? There are around 18V across the 470k grid leak resistors, so there is 40uA of grid current. I don't understand why this happens... I was expecting to get a few volts of negative bias, not -20V.

[Merlinb]

Because your voltmeter is dragging down the voltage at the grids. Measure the voltage directly between grid and cathode instead.

Everyone makes this mistake.

[Torquil]

Oops, I should have caught that one...

Thanks, you have probably saved me from a couple of hours searching for a non-existent defect!

[jjman]

Because your voltmeter is dragging down the voltage at the grids. Measure the voltage directly between grid and cathode instead.

Everyone makes this mistake.

So the meter allows enough current thru it to drop the voltage thru the grid resistor? I guess one should use the voltage on the junction of the "4 resistors" as the grid voltage?

I also noticed a large alleged bias voltage on my 12at7 LTPI. Now that I look at my notes I see that the 13volt difference between my cathode and alleged grid would actually be 1.1volt if I use the other node for the grid figure.

Excellent info to know !

[Alexo]

I think most multimeters have a 1M impedance, but LTP grids are "bootstrapped" and have a much higher impedance than you'd expect (think about it - the grid is not just referenced to the tail voltage, it's referenced to an anti-phase node); high enough that 1M will load them down.

...'course I learned this by making the same mistake!