Marshall Valvestate 8040 Quiet Boost Channel

[Helmholtz]

I asked about heater voltage because I remember a former thread where JMF reported that this special (and stupid) Marshall heater supply can't supply enough heater current with some ECC83s.

[chilidawg]

I didn't read all, but did you check heater voltage and supply?
Plate voltage has no effect on heater brightness.

Yes I did. 16V AC at both heater pins, same as the main transformer secondary taps, and the two 68 ohm 5W ceramic power resistors that I put in to replace the factory default 3W are also hot to the touch.

What do you mean with AC mode and DC mode?

My multimeter measuring mode.

[chilidawg]

You can't really 'prove' the diode D3 is not bad. It can measure fine with your diode check function, yet still be intermittent in circuit. Or the traces measure continuity but some solder joint will not pass more than the tiny current your meter tests with. So either the diode or it's connections seem to be responsible.

But it can't be the diode right? because it does rectify the AC voltage, proven by the fact there is DC voltage presents at the HT pin, albeit only a few hundreds mV and the high voltage cap is charging painfully slow.

It is only when I'm measuring AC voltage (negative probe is clipped to the chassis and the positive probe is touching the diode anode leg from the component side) then that same cap will be charging very fast, and when it has enough DC voltage, the tube turns on and passes the audio signal to the next opamp stage.

And I didn't apply pressure at all to the diode anode leg with the probe.

[g1]

When it is only showing mV DC at the HT pin, it is not conducting current, so not really rectifying. This comes up occasionally with rectifier circuits, the diode measures ok but just doesn't work when the circuit voltage is applied.
It may not seem like you are applying much pressure, but it's not really possible not to. Or somehow the meter's resistance is making the diode start working.

[The Dude]

Diodes are cheap. Just change it and see, if you suspect it might be the cause of your problem.

[chilidawg]

When it is only showing mV DC at the HT pin, it is not conducting current, so not really rectifying. This comes up occasionally with rectifier circuits, the diode measures ok but just doesn't work when the circuit voltage is applied.
It may not seem like you are applying much pressure, but it's not really possible not to. Or somehow the meter's resistance is making the diode start working.

Got it. Learned something new today that a diode can fail this way too.

[chilidawg]

Diodes are cheap. Just change it and see, if you suspect it might be the cause of your problem.

True. I will do just that.

[Helmholtz]

Yes I did. 16V AC at both heater pins, same as the main transformer secondary taps, ...

Hardly that much with the tube in. 16V seems to be the no-load voltage.
Heater voltage should always be measured between the tube heater pins and not to ground.
.

[chilidawg]

Hardly that much with the tube in. 16V seems to be the no-load voltage.
Heater voltage should always be measured between the tube heater pins and not to ground.
.

My bad. I measured those with the negative probe clipped to ground.

And all of my measurements are done without the tube in the socket.

Anyway, I have changed the 1N4007 diode. Same behavior. I timed it. It took 3 minutes for DC voltage at 10uF 350V cap to charge up to 500mV and it will only charge up 10x faster if I touch the diode's anode leg with my multimeter probe while in AC voltage measurement. If I remove the probe, the cap will start to discharge.

I just can't seem to replicate the result that I got when I had the board on my bench at the time I tested it with my 2x10VAC power toroid transformer and there was 40V DC measured at HT pins (which I took photos and attached in my earlier post). The only variable that is different from then and now: the board was not grounded to the chassis by the green wire because I didn't bring it home, but this time it is.

Is there a better circuit to get that high DC voltage for the tube plates from the onboard auto transformer output? it seems this single rectifier solution is not very reliable.

[Helmholtz]

My bad. I measured those with the negative probe clipped to ground.

And all of my measurements are done without the tube in the socket.

With this heater circuit the tube must be in to see the correct heater voltage.

Regarding the plate supply, there must be an intermittent connection somewhere. Could be a bad solder joint or a broken PCB trace.

[g1]

Regarding the plate supply, there must be an intermittent connection somewhere. Could be a bad solder joint or a broken PCB trace.

Maybe a cracked leg at the HV transformer? Or something else about the autotransformer that the meter resistance is correcting?

[chilidawg]

Regarding the plate supply, there must be an intermittent connection somewhere. Could be a bad solder joint or a broken PCB trace.

Maybe a cracked leg at the HV transformer? Or something else about the autotransformer that the meter resistance is correcting?

No cracked leg or a bad solder joint. Broken PCB trace, yes, but it has been corrected when I put in a new 1N4007.

Last night, I pulled out the autotransformer to inspect it, physically it looks fine and almost like new considering its old age. Measured the resistance of the 3 pins.

From pin that connects to GROUND and to the middle pin, resistance is approximately 43.5 ohm.

From pin that connects to GROUND and to the pin that goes to the diode, resistance can't be measured (meter reads OL). Same result from this pin to the middle pin.

Since I don't know anything at all about it, I resoldered it back.

No change in behavior. The plate supply cap is still charging (and discharging) at a steady rate of 1mV DC/second. Left it for an hour, came back and meter read 8V but it's already discharging, so I have no idea if the cap voltage ever reached more than that.

[chilidawg]

And here's 5 minutes later.

[x-pro]

From pin that connects to GROUND and to the pin that goes to the diode, resistance can't be measured (meter reads OL).

This is where your problem lies. The winding resistance should be tested, especially since you have a device with auto-selection of the measuring limit.
Or still a crack in the printed conductor from the transformer to the diode anode, as pointed out earlier.

[chilidawg]

This is where your problem lies. The winding resistance should be tested, especially since you have a device with auto-selection of the measuring limit.
Or still a crack in the printed conductor from the transformer to the diode anode, as pointed out earlier.

I measured the winding resistance of the auto tx after I pulled it out of the PCB.

I don't know how many times I have to repeat this: there is no cracked PCB trace.