level devil triode diode trouble

[njm]

So I've been given a level devil to fix up. Compressor section was working but the expander wasn't. It still had a lot of original lytics in there and leaky caps in the signal path/control section. So I recapped the lot.

Here's the full schematic
http://www.harmonycentral.com/t5/Craig-Anderton-s-Sound-Studio/Gates-Level-Devil-Restoration-Blog/td-p/34460437

In the control section, the signal gets as far as the input to V6a, but nothing comes out of this diode. The heaters are working. Valve has been swapped. Valve socket tightened and cleaned. Voltages checked. Ground connection checked. R37/38/39 checked (all ok but replaced anyway just in case!) Still faulty though.

So I lifted C11 and R39 to isolate that stage as much as possible and it's just doing nothing. FFS!!!!

I'm stumped. Any ideas?

 

[PRR]

You do realize the diode output is _DC_, right?

Zero signal AC at cathode should make near-zero DC at plate.

1V AC at cathode should make a part-Volt DC at plate (more or less with or without C11).

If plate is resolutely ZERO, bet on a short at that node.

 

[njm]

haha yep I know what the circuit is doing, half wave rect of waveform to get the cv etc etc

You'd think a short or open circuit component, it's only one semi with not even a handful of passives. But nope, I've checked that. This circuit's made on a PCB, gave that a good clean and checked the resistance from the anode and cathode to all the surrounding traces. No shorts anywhere. It's not a densely populated board, the spacing between the traces is pretty good. No components touching eachother that shouldn't be. And there's plenty of signal level getting coming in through C13.

I'll come back to it tomorrow. Right now, I need a drink

 

[PRR]

The 64V bias at the input means you need a BIG signal level in before the diode starts to conduct.

I'd try that as zero V and at 3V just to be sure it isn't insufficient signal.

 

[njm]

Aha so yes there's not enough level for V6a to conduct. Needs at least 50v peak to start doing it's thing.

I eventually found the fault, bias problem with V7b was preventing enough signal level through. I found a couple of little errors on the lo-res redrawn schematic around this stage. And then went on to find someone had very kindly replaced R53 with a 3k instead of a 1k6 at some point and this was putting too high a voltage on the cathode. They'd done such a neat job of it that it looked like an original part!

Never assume anything!

Thanks for your help PRR

And here's a corrected control section schematic