Aaron wrote: ↑Wed Jan 17, 2018 1:22 am
I don’t think it would be a voltage spike with a valve rectifier, silicon yes.
But I had a quick look at the specs for the Kemet capacitors and they have a tolerance of -10% - +30%.
So even though it’s rated for 450v, the capacitor you have might have been only rated to 405v and still be in spec.
That tolerance spec is for capacitance not voltage. Whatever excitement doubtless exceeded 450V. With 384VAC math shows ~500VDC average with expected transients closer to 600V.
I don’t think it would be a voltage spike with a valve rectifier, silicon yes.
But I had a quick look at the specs for the Kemet capacitors and they have a tolerance of -10% - +30%.
So even though it’s rated for 450v, the capacitor you have might have been only rated to 405v and still be in spec.
Thanks,
Aaron
ahh crap, well it runs well under 450 normally I think, but on first power up before tubes conduct, they do spike up above that I'd guess, I think I recall seeing it in the 490 range on first power up. After conducting the power all drops into the high 300's if I recall. I'll have to replace that one and give it a look.
I think the failing cap was probably the reason for the arcing. I’m not sure how having back-up diodes would impact that. If during the arc there is any reverse voltage on the caps (if that happens, I don’t really know), they would help contain the damage. Certainly they will protect against a short in the vacuum rectifier.
Replace them all? I've heard good things about KEMET caps, I'm sure there can be occasional duds from the factory, and this is pretty much my only failed new cap ever, so I figure I was due for one eventually. I was just going to buy another identical cap and replace it.
pompeiisneaks wrote: ↑Wed Jan 17, 2018 3:35 pm
Replace them all? I've heard good things about KEMET caps, I'm sure there can be occasional duds from the factory, and this is pretty much my only failed new cap ever, so I figure I was due for one eventually. I was just going to buy another identical cap and replace it.
Measured ~490V and engineering formula suggests peak could well exceed that. KEMET capacitors top out at 450V specification. Cannot say what they have for temperature derating curve but expect it's similar to other electrolytics. Really do want high quality with 500V or even better 600V rating. First in the chain gets transient brunt due to downstream impedance. You could replace it, and leave the rest. Personally lazy, so I'd spend the money and time changing all now. That, or a power transformer down closer to 330-340VAC.
I don’t think it would be a voltage spike with a valve rectifier, silicon yes.
But I had a quick look at the specs for the Kemet capacitors and they have a tolerance of -10% - +30%.
So even though it’s rated for 450v, the capacitor you have might have been only rated to 405v and still be in spec.
Thanks,
Aaron
I think those specs are for the capacitance rating, I think the caps can take surges up to like 500+ without major problems, so someone else said
Oh I may have failed to update the group here, I had a nice catastrophic failure debugging it further, I just powered it back up to see with an old rectifier that had been arcing to get voltages again, and the first cap blew up in spectacular fashion. Here's the gut shot.
IMG_20180116_151901.jpg
and man was it a shock.
I ended up thinking it through, and posting on hoffman and fb threads (Not sure why I forgot to here) and think the most logical idea here is that the first tube that arc'ed in that first video was bad (it came from the radio with the bad power transformer). And that may have damaged the first filter cap. That or the fact that the power transformer is a bit higher voltage than the tweedle dee deluxe is supposed to be, may have just been too much voltage with first power on and killed it.
I'm going to be replacing that 4.7k 2W with a 5k 5W that will also drop voltages a tad down the line, and then get a replacement F&T 500V 30uF for that first stage so I can give it the room it needs.
~Phil
You do not have the required permissions to view the files attached to this post.
Well to put this one to rest, I got in a 500V 30uF cap from Doug Hoffman, put er in, and then put in a replacement 5k (vs 4.7k) 5 watt resistor between B+1 and B+2, and fired it up. The voltages at B+1 after I slowly brought it up on the variac, even with warm tubes sat at around 440 to 450 at idle. This jives with the theory that the new transformer is a touch higher voltage and was causing too much voltage there. I started playing it and it sounds great again, and the voltages drop more in the 390 range with active play going on. I put in the old RCA rectifier and it's doing fine, do I need to swap it if it arced that much?
The complete story is that exceeding the voltage rating on the reservoir cap caused it to fail and draw excessive current, which in turn caused the rectifier arcing. From what I read, arcing can locally strip the cathode of its coating. I suppose that debris, which may be trapped between the cathodes and the anodes, might lead to more arcing or shorting in the future. Have a look for flakes of coating material in the envelope. It's working now, but I think I would feel better with back-up Si diodes when using a suspect vacuum rectifier.