SF Twin blowing fuses and output tubes

[spy]

I have an early silverface Twin that keeps blowing fuses and I lost one pair of output tubes.

The amp is fully serviced, new el. caps, new rectifier diodes, new grid stoppers and screen resistors, etc etc but after a few days I lost one pair of 6L6s. I replaced them, replaced the fried screen resistors plus the bias pot, diode and resistors at the bias circuit just in case, biased it properly around 60-70% but after a few rehearsals mains fuse fried again! Now, it's on my workbench with a new fuse and it seems fine! I have it for an hour with a signal generator and a resistive dummy load and I have full output power.

Coincidence? I don't think so as when I was rehearsing I heard the same crackling noises as the previous time before the fuse fried so something is going on there. Any thoughts? The only difference it that this time I didn't loose another pair of 6L6s.

I suspected the output tube sockets but they look clean and fine without any traces of carbon. What else can be the problem?!

[spy]

Ok, after a few hours with a resistive load it blew the HT fuse (for my tests I've put a HT fuse) and I saw flashes on one of the power tubes. Unfortunately, I don't remember if it's one of the new pair or the old pair, damn.

[Leo_Gnardo]

Hi Spy good to see you again. I'd be looking for a bias failure. Not the whole supply, more like a poor connection across the output socket with the 1K5 resistor, the way Fender does, cracked/drifted resistor or solder come unstuck. Or loose connection of the wire feeding combined bias & signal to output grids, or loose 220K at the bias distro point. Possibly even a failing cap between output drive plates and bias distro points. Those old silverfaces, some of 'em still look nice and new, but they're getting on to nearly 50 years old (!) and often need help.

[spy]

Yes, I thought about bias failure, too that's why I replaced the bias pot with a new one and every part at the bias circuit. I've also replaced those 1K5 so no problem there but I'll check the connections around those 220K and every other place I missed.

Maybe it's a good idea to replace those sockets, you never know, what do you say?

[Leo_Gnardo]

Maybe it's a good idea to replace those sockets, you never know, what do you say?

You've been very thorough to replace much of the bias network. If that hasn't improved the amp, it's still sparking inside output tubes & popping fuses at random intervals, then yes I'm thinking the output tube sockets, or at least one of them is suspect. There may be a hair-thin carbon track left over from a long ago arc. Hard to see and just the right thing for your complaint. I'd put two fresh sockets in place on the side where you lost two tubes previously. Since, in each pair of output tubes, the two grids are only separated by a pair of 1500 ohm resistors, when one tube fails or is forced into failure, the other is pretty much in parallel and also fails.

[Jazz P Bass]

My question is: are the tubes that failed new?

If the tubes where just jockied around, then they may have all been on the edge.

[Drewline]

I agree about the tube sockets. Check for carbon arcing between the pins. If it is there replace them. If there is no carbon present, clean them & retension the pins.

[spy]

Jazz P Bass, I think it's one of the old ones!

I decided to replace the sockets, you never know. :/

[MandyWeidemann]

yep, try to replace the sockets...seems that this kind of problem is seldom related to other parts...