Marshall Superbass with mystery short in power supply. Exorcist required.

[The Dude]

What resistance do you measure from the rectifier output to ground?

[Jazz P Bass]

Oops.
I did not take note that the issue is with the HT fuse removed.
The way I see it is you have 2 .022 caps, 2 100uf caps and 4 diodes.
Build a bridge on the bench and try to power that up.
No caps.

[Greg_L]

Check that those big caps aren't shorting to the chassis.

[glebert]

I think by removing the HT fuse, I have eliminated all the tube sockets. Everything from V1 to OT. The short is happening with no HT fuse AND with the Standby Switch closed. Standby switch open shows no short. Filament and bias supplies (replaced 2 leaky caps and its diode) are On with Power ON, and don't cause any issues.

Anything else you can think of Glebert, or can suggest I check? I have a couple of other suggestions to check on when I get back to my workshop.

Of course removing the fuse isolates the sockets, my lack of attention there.

So the standby switch is before the rectifier on this? That seems weird. You mentioned bringing it up on a variac with standby ON. Does the HV come up at all when you do this or is it basically 0v?

Well, you could pull the rectifier that you just put in and see what that does and just step it back towards the standby switch.

[xtian]

I think you've confused me, xtian. With 2 tubes out and the same speaker load do you use half or double the output tap? I thought it was half. . if I understand you correctly, if he had an 16 ohm speaker he would need to use a 32ohm (non-existant) tap, and not the 8ohm tap that I thought.

I am, in fact, confused. I thought if four tubes are happy pushing against X impedance, two tubes would be happy with X/2 impedance. Is that not correct?

[glebert]

I am, in fact, confused. I thought if four tubes are happy pushing against X impedance, two tubes would be happy with X/2 impedance. Is that not correct?

It is about matching impedances. With four tubes you have output impedances in parallel, so the output impedance is cut in half relative to two tubes. If you start with four tubes and remove two you have doubled the output impedance (on the primary side of the transformer) so you should double the impedance on the secondary side of the OT as well. You can do that by using a load with 2x ohms relative to the output tap.

[Enzo]

The impedance thing is an important and worthwhile discussion, but mismatching speaker to transformer here is not going to short out the B+ or blow fuses.

[christarak]

What resistance do you measure from the rectifier output to ground?

Hi Dude. Using Fluke 87V, the rectifier output (HT) to ground was 74K, which is also the top of the totem capacitors, so I measured the mid-totem to ground: 1M !!! I discharged the caps and tried again. Similar. I measured the capacitance of both 100uF caps (50+50 dual cap cans) and they both read at around 130uF. I'm not sure if I can trust my readings while in circuit.

So I thought "bugger it. I'll put two 220K totem resistors across the capacitors". Same. Short circuit appears when closing the Standby switch.

I'm stumped Dude.

[christarak]

Check that those big caps aren't shorting to the chassis.

Good suggestion, Greg. I tested all the terminals with the Fluke, and only the one at the bottom of the Totem showed an earth connection. Thanks mate

[christarak]

Thanks for that idea, doombass. It must be something weird like that. I'll check that when I get home in 5 hours and report back.

Sort of good news Doombass. We have cleared the switch. I removed the "out" connections of the switch to rectifier and caps and left the transformer connection. Turned the switch ON and NO short. I built a rectifier bridge with the same diodes and connected that to the "out" connection of the StandBy switch. NO Short when the switch was closed.

[christarak]

Nothing has been accidentally grounded at the center of the totem-pole filter caps? (cap shells insulated if required)

I am knocked out by the suggestions. After a whole bunch of tests recommended by our colleagues, I seem to be coming back to my brand new F&T 50+50uF caps. I'm not sure what you mean by cap shell insulation...but its starting to sound good. The resistance to the HT/B+ side of the totem cap to ground was 74K. The resistance from the centre of the totems to ground was 1M.

[HaroldBrooks]

I am knocked out by the suggestions. After a whole bunch of tests recommended by our colleagues, I seem to be coming back to my brand new F&T 50+50uF caps. I'm not sure what you mean by cap shell insulation...but its starting to sound good. The resistance to the HT/B+ side of the totem cap to ground was 74K. The resistance from the centre of the totems to ground was 1M.

I just had a can cap short recently on a much more basic and older amp, and it was at first very preplexing to me, but easier to diagnose than in your particular case, because the old 1957 Zenith amp was so very simple in design. The transformer on my amp gave a good 660 volts with no tubes and no rectifier, and I switched rectifier tube to a known good one, again still the same short appeared on my light with no other tubes in the amp.

In my case what pointed me to the first filter cap being shorted was the big carbon resistor coming right off the Rectifier and just before the first filter cap. It started cooking and crackling even at reduced voltage with a variac, and with no other tubes in the amp. I replaced the can cap with conventional single caps and all was then well.

Hope this is your issue and g1 got it. I'm in this game of fixing my own amps only for close to three years now, and already I've gotten two bad JJ tubes and one bad cap, supposed to be brand new !

For more information read to quote at the bottom of my post...

[Chuck H]

Well there's just not that much to inspect and the usual method of discovery hasn't been explored for some reason. Just start disconnecting things one at a time until the problem stops. When the problem stops you have found the culprit. I would disconnect things as shown below. A possible exception being a miswire that could have existed before you got the amp. It was reported that it kept blowing fuses and the fuse holder needed to be replaced.?. That seems like an odd order of thinking. If a lamp keeps blowing light bulbs I typically wouldn't suspect the socket.

[Enzo]

Just a thought. The stacked caps are wired in series watching polarity. In other words negative of one cap to positive of the other. The generic dual 50+50 cap has them in parallel, not series. So both negative sides are wired together. There is no way to wire a single dual 50+50 cap for series. Could this be something we did here?

[g1]

Yes, a picture of the main filter caps wiring would be helpful.