got a gremlin in a 102 hybrid build

[Luthierwnc]

Hi All,

I'm spark-testing a redo based on 102 with an internal loop. I've attached a gut shot and the schematic. Everything works fine until I get to the output of the loop (circled in red). On the scope, the signal is very strong and symmetrical. The input (drive) trimmer knocks that down quite a bit but it goes into the recovery grid of the loop in good shape.

Past that, nothing. I get no signal on the plate or the other side of the .1 coupling cap. I checked the wiring, cap for value, leakage and shorts and the recovery pot. It all seems fine. Changed tubes too. The cathode reads the correct resistance to ground and the bypass cap isn't leaking. I pulled the LNFB loop .22 cap to see if that cap was leaking signal but that wasn't the issue.

The one thing that did seem odd is that the voltage on either side of the 100k supply resistor was only one volt different. 354 to 353. It should be knocking down more than that.

If anybody spots something I missed, I hope you'll mention it.

Thanks, sh

PS the drive trimmer is on a terminal strip directly behind the .47 cap and in front of the send jack. It is hard to see in this picture. sh

[martin manning]

The recovery stage doesn't have any current flowing through it. Since you have voltage at both ends of the plate load, the problem is between there and ground. Either the ground is missing or there is an open connection somewhere.

[Luthierwnc]

Thanks Martin, I'll have a look over the weekend and report back, sh

[Luthierwnc]

The weekend was too far away.

Rechecked all the grounds. Didn't discover anything. The send supply is after the return supply on the rail and it works fine. Both the power supply and the resistors for the loop ground at the same point near the power tubes.

Took the .1v pic on the recovery grid. Took the 5mv pic after the coupler. Tried a third tube too. Same story. Also checked the connections on the socket to see if there was a broken pin sleeve. Nope.

No loose wires or funky solder joints either. Thankfully, it is time for bed.

Skip

[martin manning]

What's the DC voltage at the cathode of the recovery stage?

[Luthierwnc]

zero

I disconnected the cathode resistor on the board and attached another between the #8 pin and the return jack. Still nothing.

I did check the socket pin sleeve for continuity but I might have another look.

[martin manning]

I should have asked for the resistance to ground from the cathode too, but it looks like there is something wrong with the socket.

[Luthierwnc]

1.5k

[martin manning]

If it's an open circuit, then it down to the pin connections at the plate and cathode, or, no filament current. How about the filament voltage? is the heater on?

[Luthierwnc]

3.28VAC on both sides. The first half of the tube is working fine. I do have some DC on the heaters that I neglected to put on the schematic. I've updated that now.

[MKB]

Make sure your plate resistor is the correct value.

When you run into a weird problem like this that defies reason, something very strange is happening. I had a similar problem awhile back when repairing a Fender Blues Deluxe, the PI was not developing signal on one side, even though the DC readings were spot on. Replacing caps didn't help, nor did checking resistors or swapping tubes. The problem was a hairline solder splash on the PCB that shorted the signal but left the DC voltages intact. Removing that splash fixed it and it worked great after that. It took me hours to find that gremlin.

In your case you may have a plate or cathode pin that opens when hot and doesn't allow current to flow. Either that or the cathode resistor ground lead isn't grounding, check its resistance to ground as well.

[Luthierwnc]

MKB;

Yeah, sometimes you just have to redo stuff because it all checks out fine but it doesn't work. The next time I sit down at the bench I'll probably replace the socket. I've got some spares, I know I have signal on pin 7, juice on pin 6 and the correct resistance to ground on 8. And, I know the tube works.

There have been times when the rework worked but I still didn't know what I fixed.

sh

[Luthierwnc]

Bad socket.

It had a hairline crack on one of the pins that the solid wire held in place. All's good although I've only heard it through the singing of the OT -- but the waveform looks very clean.

Tomorrow I'll get into the tweaking phase. Since this is being rewired along similar lines as before, that shouldn't be too dramatic.

Thanks to all for chiming in and hope your team does well tomorrow. Mine will have to wait for next season. sh