Is it possible and/or likely for a preamp tube failure to cause a mains fuse to blow?

[mort]

Welp, Mercury has determined that there was no fault or error in the construction or materials, and that the cause was "from outside the transformer".

I have the amp chassis but I do not have the tubes that were in it. Should I have the customer send me the tubes so I can check them out? How can I figure out what happened? I certainly don't want it to repeat the failure. This is kinda discouraging. I've built the same circuit several times without issue so it's puzzling to me how a whole friggin PT could just go poof like that with no apparent cause. And this is not a real good time for me to have to eat an expensive part like that

[mort]

And just got this additional explanation from the owner... does this ring any bells?


What I can tell you is that the first time it made a noise I had a short cable patching the low ch 1 to the high ch 2 mixing the two channels. It started crackling and I pulled the jumper and it stopped. But the amp sounded different after that - very clean. Not much break up at any volume. It worked for a few more days after that.

[Gaz]

I think the main issue here is that there was no B+ fusing (unless I missed that somewhere), and my guess is a tube short. You can't rely on mains fusing to protect the PT - that's not what it's there for. And did you build this amp per the schematic without screen resistors? - that's just asking for trouble with modern tubes in my opinion.

I'm not sure about the customer's comment with the patch cable, but I'm 99% sure that that was just a coincidence, or that they noise was simply set during signal conditions.

[mort]

The point is well taken about the B+ fusing. I've seen it on Marshall drawings but never on old Fender drawings, to my knowledge. It's cheap insurance so yeah, fuse it.


As far as the customer's comments, I wasn't concerned about the jumpering. I was interested in the crackle, then the change in amp dynamics, and finally the all out failure. Can a tube short behave that way? Or only cause partial damage to the PT but still work for a while and then go poof? I now I have the tubes back and I'm not reading any shorts on the tubes but I'm scared to put them into an amp for testing. How do I go about determining if the tubes (power and recto) are good?

[pdf64]

Build a light bulb limiter.
Use a test amp that has fusing and series silicon diodes on the B+ winding. Fusing can be quick blow type if standby (ie normal implementations) are avoided; quick blow fusing should greatly reduce the stress on the PT etc if a tube shorts.
The heater line to have a robust ground reference.

Power up via the limiter with just the rectifier in place.
If ok, fit the power tubes.
If ok, run some signal; light bulb brightness should track the signal level.
If ok, try it without the limiter.

[R.G.]

And this is not a real good time for me to have to eat an expensive part like that

There's a business lesson here.

If you buy parts for ... um, say $400 just to pick a number, then build and sell an amp for $1200, you make $800 for your labor, right?


Wrong. You make $800 to cover everything that went into the amp past and *future*. The past stuff not accounted for was supplies, tool wear, advertising, your time doing any accounting, taxes, insurance, all that stuff that goes into overhead. The future stuff includes, as Mother Nature is delicately teaching you right now, the cost of warranty repairs.

It's not widely appreciated, but businesses issue warranties as much to limit the time they're going to have to support the goods as they use them for advertising. And they account for this future cost by calculating the per-item cost of parts, labor and shipping for fixes by multiplying the failure rate of the parts times the length of the warranty, and including this statistical future cost in the price of goods sold as a reserve for warranty costs.

It is a smart idea to put back a portion of each sale's "profit" as what it is - a reserve against future warranty costs. It's not a profit to be spent, and the cost that the reserve covers will come back and bite you whether you've made the reserve or not - as you're finding out now.

[mort]

Ok here's what I've done. I powered up the amp with a regulated power supply and started from zero vdc on the B+ and monitored the current as the voltage was raised. Everything seems to be working perfectly with the original power tubes installed. With all tubes installed (except the rectifier) it's pulling 87mA at idle and peaks at around 100+mA while playing at near full volume at 350vdc. No unwanted noise, pops, anomalies or anything.

This leaves me thinking one of two things ... either the PT just ate itself adn Mercury doesn't agree, or the customer inserted the recto tube incorrectly(the guide post broke off the recto tube at some point) when he first got the amp. He says he didn't but that he had to be very deliberate about inserting it correctly. But I'm not sure that even makes sense because he played the amp alot for two weeks before it failed. If he had inserted it incorrectly, wouldn't it pretty much fry something right away [if it were going to damage anything at all]? If it just happened to get HV on the heater then wouldn't it pull enough current to cook those windings in about a second?

[R.G.]

the customer inserted the recto tube incorrectly(the guide post broke off the recto tube at some point) when he first got the amp. He says he didn't but that he had to be very deliberate about inserting it correctly. But I'm not sure that even makes sense because he played the amp alot for two weeks before it failed.

A rectifier with a broken-off insertion key is by far the most likely explanation for the stuff you're listing.

If he had inserted it incorrectly, wouldn't it pretty much fry something right away [if it were going to damage anything at all]? If it just happened to get HV on the heater then wouldn't it pull enough current to cook those windings in about a second?

Maybe, maybe not. It depends on exactly what wrong orientation he inserted it. I'm thinking that it's possible that he got it in a way that caused a "soft short", a low resistance across the HV winding.

A soft short is one of the ugly possibilities for failures. Something not "short enough" to pop a fuse, but high enough to just cause too much heating could cause this. Unchecked heating, perhaps in only one of the halves of the HV winding could well overheat one place in the winding and eventually cause degradation of the insulation.

It may not happen in seconds because if it's only heating, not a fault current that would open the AC mains fuse, then you're in the range where the thermal mass of the copper and iron around the hot spot keep things going for long enough to heat up and soften the insulation over most of the winding - until the softened wire insulation gives up somewhere, and nearby turns short, then arc, then weld, then...

The thermal time constant of a 300-500W mains transformer could well be several hours, if it's heating all over a winding, not in one hot spot. We used to test insulation temperature rise by running a transformer at maximum current with dummy loads, then measuring the heat rise versus time by measuring the copper resistance in a winding. Copper has a thermal coefficient of resistance of 1.0039 per degree C, so you can use the copper wire itself as a temperature measuring device if you have sensitive, accurate, and repeatable enough meters. (Those three properties are different.)