470K Grid Leak Resistors

[RossH]

What happens when they are not there? Well, I found out. I swapped out my non-power resistors from CF to CC this weekend. This amp has had a couple of faces now so there are a few extra turrets on the board (I should remove them but hey...I'm not completely done yet. )

Anyway, I connected the 470K grid leak resistors ahead of the EL-84's to the wrong turret, effectively connecting them to nothing. They were removed from the circuit.

A couple things happened:
1) more signal noise in general (would have expected that)
2) the plate voltage dropped ~50V.
3) the cathode voltage raised ~15V.
4) there was a small voltage (under 1 volt) on the grid.

The amp was way more distorted (less plate voltage) and more noisy. It wasn't a good thing. This is mostly an FYI type thing. I don't quite understand why it effects the plate and cathode. It was interesting to note that when I would check the voltages the DMM would flash 330V on the plate first and then immediately follow with the 275V and remain there. The same with the 10.6V-to-25V cathode variation. When I tied the resistors to the right turret it all went back to the great 18 watt.

I get to try the CC's out with the band tonight. It will be interesting to note any differences.

~ Ross

[zaphod_phil]

hose resistors are there to tie the grids down to 0 volts DC, so that the voltage at the cathode will provide the correct bias. Without them in there things can go haywire as the bias becomes indeterminate, and if the grids floated towards the same voltage as the cathodes, you could redplate your EL84s...

As far as using CCs in that position goes, there is almost no voltage differential across them and so it's unlikely there will be any improvement in tone you can hear. To get some results you should make the PI plate resistors carbon comp.