Troubleshooting fuse blowing

[Sherman]

Last night I "finished" wiring the first of my KT88 monoblocs. After double checking all component values (ticking them off on my component list) and double checking all connections (ticking them off on the schematic as I stepped through it from the input) I decided to turn it on.

Following my standard procedure for turning on anything electric/electronic that I've built myself I connected it to a heavy-duty (15A) extension cord plugged in to a switched socket across the room. Standing across the room I flipped on the switch. No squeal from the cheap speaker attached, no bang, no fire not even a wisp of magic smoke! In a moment the filaments began to glow and appeared normal on all three tubes.

However... after 30 to 40 seconds the 2A slo-blo fuse blew. Checking all components revealed nothing was overheated, no smoking or burning anywhere. I checked all component values again and all connections again. I tightened down my chassis ground post, removed all the valves and blew out the tube sockets with compressed air in case a bit of wire got caught in there. I replaced the fuse and turned it on again with the same result.

Nothing was connected to the input in either case. I did have an 8 ohm speaker connected at the output.

I am using tube rectification (5U4G) so it seems that the fuse is blowing after the rectifier begins passing current. I've attached a Tubepad drawing of my components and how I've grounded things. The first cap is 40uF, the choke is 8 henries, the second cap is 100uF. PSUD II says this should work for 400 VDC, my target B+.

Any suggestions on where to start? My current ideas are-

1- disconnect the rectified high voltage from the CLC filter and check the filament voltages and see if the fuse blows with only the filaments connected

2- check the rectified voltage before the caps (with it still not connected to the CLC filter)

3- reconnect the rectified high voltage and check its value

4- ????

Any help greatly appreciated!

[SY]

Let's do some basic troubleshooting. Disconnect the signal circuitry from the power supply- substitute a power resistor as a dummy load, maybe 5Kohm at 50W. See if the fuse still blows. If not, your problem is in the signal cicuitry. If so, your problem is in the power supply. And either way, you've cut the possibilities in half.

[Jax]

I cannot see any resistor between 6N1P grid and ground. The grid would float if the input is left unconnected.

[SY]

Hmmm, looking at the drawing, it's not clear which way the cathode bypass cap is oriented. Hopefully, positive to cathode...

[Sherman]

Originally posted by SY
...Disconnect the signal circuitry from the power supply- substitute a power resistor...See if the fuse still blows. If not, your problem is in the signal cicuitry. If so, your problem is in the power supply...

Thanks for that advice. I'll have to see what power resistors are in my bin. I'm sure I don't have 50W but I do have a handful of 500 ohm at 25W. I could put some in series to get 5K.

Originally posted by SY
Hmmm, looking at the drawing, it's not clear which way the cathode bypass cap is oriented. Hopefully, positive to cathode...

Actually the bypass caps are non-polar electolytics so orientation there shouldn't be a problem.

[Sherman]

Originally posted by Jax
I cannot see any resistor between 6N1P grid and ground. The grid would float if the input is left unconnected.

You are right! The original schematic had a 100K pot between the input and the 6N1P. When I decided to do monoblocs I just put a 100K resistor in line with the input and forgot about putting one to ground.

What value would you recommend there? 500K or higher? I checked a schematic for a different SE amp and the designer recommended replacing the 100K pot in that design with a 100K resistor in series between input and grid and a 1M to ground.

[SY]

The missing grid leak resistor's value isn't very critical. 100K will do fine.

[jan.didden]

.... but that 100k should NOT be in series with the input, it should be from grid to gnd. There may be a need for a series input resistor as grid stopper, what would you recommend SY, I'm not a tubey person, a couple of k?

Jan Didden

[SY]

I thought the connection was clear, but maybe not- thanks for the backup, Jan. A grid stopper of 4.7K will take care of all but the most stubborn cases; it's important to keep its body near the physical grid pin.

[Sherman]

Originally posted by SY
I thought the connection was clear, but maybe not- thanks for the backup, Jan. A grid stopper of 4.7K will take care of all but the most stubborn cases; it's important to keep its body near the physical grid pin.

Thanks,
I'll solder in the resistor before doing more tests. The arrangement of parts should allow me to solder the leads directly to the pin and to the ground point with no additional hookup wire.