Crate Blue Voodo BV6212 blows fuses, testing power supply component quesitons

[stoneattic]

So three songs into the gig my Crate Blue Voodoo BV6212 makes a weird "humming" sound for a split second before going quiet. The power light is off and the thing appears completely dead. I quickly swap it out for the back up I always bring to gigs (a Peavey Bandit 65 from the 80s that would survive a nuclear blast, old Peavey amps and cockroaches are all that will be left).

I finally got around to troubleshooting it the other night. I pull the chassis and do a visual inspection which doesn't show anything obvious so I replace the fuse and turn it on with my current limiting light. Light is off on standby but as soon as I take it off standby the light glows BRIGHT. I pull the tubes one by one and get the same bright light when taken off of stand by (even with no tubes installed). So I unplug the connector from the power transformer B+ secondary and the current limiting light stays off when taken off of standby.

As this point I believe the power transformer is okay and the problem is something in the B+ power supply. Based on what I've read the most likely culprits are a shorted filter cap or shorted rectifier diode. Does this sound right? Besides pulling at least one leg of each diode/cap is there a simple way to test these in circuit? It seems me that to cause this issue one of the two diode connected directly to ground must have failed. Can I just measure between them to ground? I don't have a way to measure the value of the caps but I assume a shorted cap would be a "real" short and measure 0 ohms. I assume (hope) that l can measure these to ground also to detect a short. Am I wasting my time even testing these and should I just replace them all? Any other suggestions?

Schematic attached for reference.Crate Blue Voodoo BV6212 Schematic.pdfCrate Blue Voodoo BV6212 Schematic.pdf

Thanks!

[Leo_Gnardo]

Two more possiblities. Check for arcs on output tube sockets, also on bottom of output tube bases. Also are there flyback rectifiers from output tube grids to ground? If yes check these for shorts.

[Enzo]

Agree, way more likey a shorted flyback diode. Power off, measure resistance to ground from power tube socket plate pins, either look low? Or disconnecet center tap of OT primary from B+ and see if fuse no longer blows.

[stoneattic]

Thanks guys, you are spot on!

I had a couple of minutes to look at it last night and it looks like one pair of flyback diodes (two 1N4007 diodes on each EL34) is shorted. I have some spares but didn't have the time to swap them. Hopefully I will get it back up and running over the weekend.

I'm glad I posted yesterday or I would be chasing my tail looking at all the B+ components!

Now the question is why did the flyback diodes fail? From what I've been reading about flyback diode failures likely causes are:

bad tube
no load
wrong load
bad cable

The speakers are ~ 2 year old WGSs (1 ET65 and 1 Veteran 30, it's a 2 x12) with enough hours on them that they are broken in, but never abused.
I built the speaker cabling myself when I installed the WGS speakers and feel confident that they are good and they are connected per the Crate manual.
I assume that it would most likely be a power tube failure to cause the flyback diodes to fail, more so than preamp tube (maybe the PI?) and they are only a couple of years old (went in just before the WGS speakers) with pretty low hours.

I'll put a set of known good EL34s in when I first fire it up and assuming that goes well I will use the current limiting light and put the "original" EL34s in to test them. Is that a valid way of testing for bad power tubes?

[Leo_Gnardo]

I had a couple of minutes to look at it last night and it looks like one pair of flyback diodes (two 1N4007 diodes on each EL34) is shorted. I have some spares but didn't have the time to swap them.

I'd go three-in-a-row when replacing those flybacks. Can only guess Crate was trying to save a couple pennies. For amps with a fairly low B+, say 350 or less, might use two.

[Enzo]

Don't spend a lot of time on OMG WHAT HAPPENED in there. Sometimes it is just a random component failure.

Tubes generally short in parallel with them, so no stress on the tube.

A wrong load won't hurt anything.

A broken connection? Sure.

Something else? Sure.

[The Dude]

I agree with Leo. Cobble 3 in each spot instead of 2. The fact that they failed is possibly evidence that 3 are needed.

Edit: Another option would be to use a single 3KV rated diode on each side like this and put a jumper in the other spots (if PCB).

[stoneattic]

She lives!

I had spare 1N4007s laying around so I put 3 in series on each tube and tested it with the light. I was happy to see the light quickly go down to just a slight glow. I then reinstalled each tube one by one using the light and everything looked good. I only played through it for a few minutes because it was getting late but it sounded and acted as it should. I'll give it a more thorough playing over the weekend and have band practice on Tuesday so I should know if there is anything else wrong by then.

Thank you everyone for the help!