Elevated heaters, cathode bias, and trem

[Sonny ReVerb]

Hey Gents,

Well, I had my first electrolytic cap explosion. Sounded like a dud firecracker with some of that magic smoke coming out of the input jacks. I immediately powered down the amp. See pics below for what I found when I pulled the chassis out.

Sure made a mess. It was a Sprague 50uF 50V bypass cap in parallel with the power tubes' cathode resistor from the Tonebox amp. Cap C25.

This happened when I had the trem engaged. I decided to scope the trem signal at the top of the cathode resistor and it turns out to be about 20V AC. The DC level at that point was 33.4V. The sum of the two signals was >50V and pop. I replaced the cap with a 33uF 100V that I had on hand. No damage done that I can tell.

My question: Does the trem signal mess with the heater voltage in a bad way, since I am using that same point to elevate the heater voltage? Should I just ground the heater supply center tap or create another DC source to elevate the heater voltage (similar to this)?

Thanks

[shooter]

20V AC

an Ecaps job in life is to smooth out AC and make it more DC, so I'm wondering why inject trem at the PA(?), especially a fully bypassed one? 


did you have hum and decided on elevated heaters?
schematic might help

[Sonny ReVerb]

schematic might help

Schematic is available via the "Tonebox amp" link

[PRR]

55V peak-tips should not harm a 50V e-cap this year. Maybe not a "forever solution" but not an immediate burst like that. I suspect ATOMs have gone downhill. Use fresh Illinois or any of the mass-produced cans.

[shooter]

+1
don't believe elevated heaters and trem wabble are a negative, and now that point H and X are clarified I'm happy 

[Sonny ReVerb]

55V peak-tips should not harm a 50V e-cap this year.

I didn't think so, but it did go pop. The replacement was a Nichicon 33uF/100V.

+1
don't believe elevated heaters and trem wabble are a negative, and now that point H and X are clarified I'm happy 

Thanks, Shooter. The trem is injected on the grid of the power tubes via the grid leaks, similar to a Gibson GA-19RVT. The signal is reflected on the cathode.

[Tony Bones]

My understanding of popping electrolytic caps might not be complete, but I think that the usual cause is heat. Heat can come from

a) reverse polarity resulting in massive DC current
b) excess AC current
c) high ambient temp

I don't see any way that (a) could happen in your case. (b) is unlikely, but (b) plus (c) together might be a possibility. I would make sure that the bypass cap is not in contact with the cathode resistor. Give it as wide a berth as you can.

EDIT: I see in a photo in the Tonebox thread that there is some clearance between the cap and resistor. So, I don't know...

[PRR]

I once blew-up a MUCH bigger cap by putting a 120V plug in a 170V outlet. The PA system sounded extra good. For about 4 minutes. Then FWOOM! mushroom cloud of smoke to the high ceiling. The main power cap, size of a soup-can, burst.

That's one reason I doubt blips to 55V on a 50V cap, fed through a mere tube instead of right off the PT, would ever throw guts. I think that cap was bad. Atoms used to be Good Commercial/Repair parts (not supers); and I suspect they aren't even that now.

[Sonny ReVerb]

Tony - I thought heat might have been the culprit, but I'm usually careful about giving them some space next to a power tube cathode resistor. I wouldn't think a 10W resistor would get very hot in this application, though [(33.4V / 270R =) 123.7mA x 33.4V = 4.1W?].

PRR - Thanks. It does get your attention when they go. I thought Sprague had a better rep than Illinois, but I guess not these days. Do you see any problem adding the trem AC to the heater CT?