3 amps, same fault in one week: arced-over power tube sockets

[jamesmafyew]

I need some help understanding possible causes for this mess.

A respectable recording studio in NY had a Hiwatt DR504 clone I built on loan for a while. It was basically brand new, about 50 hours on it. It blew the HT fuse and they just replaced it without checking the amp out (no redplating though, they had the back off and were watching the tubes)... after that it worked for 6 minutes and then blew the mains fuse (but NOT the HT this time).

Weird thing is, that same week, they had an Orange OR-80 and an Ampeg V4 that also blew fuses and went down. Both those amps have been reliable for years. I went up there this week to check all three of them out and see what was up.

All three amps have the same visible problem: one power tube socket in each amp has arced between pins 2 and 3. In the Orange, it's nasty: melted wire, crumbling screen resistor, bad smell. In my clone build, just a mild arc on the top side of the socket between pins 2 and 3. In the Ampeg, barely any arc trace but it blew both the 6A mains fuse AND the 10A internal "idiot fuse" in series with the mains fuse.

All amps power up with no tubes and voltages look ok (including heaters).

The head producer says that the power company has been digging stuff up for a few weeks and businesses in the neighborhood have lost power periodically. I measured the wall voltage at the studio and it was 95VAC in the control room and 130VAC in the live room (where the amps were plugged in)! The studio's electrician has no explanation for this. Still waiting to hear from the guy who did the wiring. Worse, they were using effects on the front end of these amps that were plugged into the control room mains at 95VAC.

So,
1. What in the world could have caused this? How (un)likely is it that it was just coincidence?

2. How certain is it that all these power tubes are toast? The V4 had NOS 7027A's in it.

[Enzo]

You pull up to a traffic light in your red Toyota Corolla, and look out the side. There is another red Corolla just like yours. You are wearing a ball cap, and so is the other driver. WHAT COULD HAVE CAUSED THIS???


Unrelated dissimilar amps have similar problems. To me that is coincidence. AMps are not designed so close to the edge that 120v mains will blow them up.

[jrfrond]

In order for a socket to carbonize, you need heat. For heat to be generated with electrical current, you need resistance.

Pin 2 is one side of the heater, which is drawing app. 1.6A. There's your current source. If the heater pin has poor pin contact i.e resistive, there's your heat source. That initiates carbonization. For arcing to subsequently occur, you need HV. That's at the next door neighbor, Pin 3 (plate). It's a slow, downward spiral, and has nothing to do with the power grid.

Sounds like unrelated, catastrophic failures to me (nod to Enzo), caused by tube sockets that developed resistance over time. Actually, this is more likely with the V4, which originally used excellent Amphenol or Cinch tube sockets. The typical crap we get nowadays, even the "best" Belton sockets, don't hold a candle to the old stuff, and their contact tension usually sucks, so intermittents and resulting high resistance can develop rather easily.

On all of your new builds and even replacements, use ceramic sockets, and save yourself some headaches.

[Pedro Vecino]

Bad speaker cable?

[Prairie Dawg]

You pull up to a traffic light in your red Toyota Corolla, and look out the side. There is another red Corolla just like yours. You are wearing a ball cap, and so is the other driver. WHAT COULD HAVE CAUSED THIS???


Unrelated dissimilar amps have similar problems. To me that is coincidence. AMps are not designed so close to the edge that 120v mains will blow them up.

Tony Hillerman's fictional Navajo detective Joe Leaphorn always says "If you believe in coincidence you're not looking close enough."

[Prairie Dawg]

Every once in a great while somebody'll uncover a case of the old Amphenol micalex military surplus sockets and they are as John says, the good stuff. I bought about forty when I had the chance, dude didn't know what he was selling. I've been getting pretty fair bakelite sockets out of China but I don't use them in real high plate voltage applications. There is where the ceramics go.

Once you get carbon tracking there's no getting rid of it either-ask anyone who works on airplane magnetos. I have been known to put lacquer on new covers and bake them, works great but don't tell the feds I did it. You probably could put lacquer on sockets and bake them if this is a regular issue.

[jamesmafyew]

Thanks all.

The new clone build already had ceramic sockets and they were CLEAN and well-tensioned. I had played it dimed for hours with my own gear before it went to the studio, no problems. The two older amps have been used regularly at that place for 2 years with no issues. If it had been just two amps that went down, I probably wouldn't have thought much of it, but three in one week with the same arc seemed unlikely. I didn't think that just the 130VAC mains level would be solely responsible, but if the power company is turning transformers on and off, I wondered if bad line transients could cause arcing like this. Is that possible? Would a spike on the line also spike the HT enough to cause an arc on the tube socket?

The other (more pedestrian) possible commonality is that an intermittent speaker cable got used on all three amps during that week of recording, causing a flyback pulse and arc in each case when the load went open. Right now I think my money's on this.

[riz]

Unrelated dissimilar amps have similar problems. To me that is coincidence. AMps are not designed so close to the edge that 120v mains will blow them up.

The OP reported 130v mains. I might be a little uncomfortable with that.

[jrfrond]

Well, if the sockets are ceramic, then carbonizing is out of the question. You would clearly see an arc-over on the white ceramic substrate.

I'm not totally buying the 130V, but it IS what they all have in common. Flyback pulse from the OT with a bad speaker cable? Hmmm..... in a studio, where it would be painfully obvious??? Man, I don't know. If that IS the case, you might be looking at some latent OT problems from insulation breakdown.

How about socket contamination from the factory on the ceramics? Just a thought.

My suggestion: repair the amps as necessary, and add R3000 spike diodes to ground right on the tube sockets at Pin 3. Just as a precaution.

[Steve Conner]

My money would be on the bad speaker cable. Or a 4x12 with a blown speaker, causing it to be twice the impedance shown on the nameplate. Or a bassist with a Big Muff. Or some hotshot producer who thinks the amps can have their speaker output just plugged into a DI box with no dummy load.

Power line transients are unlikely to cause the problem, they won't get past the filter capacitors.

[jamesmafyew]

All really helpful!

The guys who run the studio in question are not dumb... I've worked with them before and they know what they're doing. The producer actually checked all the cabs they were using with a multimeter when this first started happening just to make sure it wasn't blown speakers, that's why I was thinking an intermittent cable.

Yeah, the arc trace on the ceramic sockets in the clone amp was a tidy little black track on the top of the socket, with a matching track on the power tube base. On the Orange and Ampeg, it was on the underside of the socket with no burn on the tubes. The Orange situation is a mess, it looks like somebody dropped a cherry bomb in there, with almost nothing left of the wires to pins 2 and 3. The Ampeg I didn't even see the tracks on the socket until I shined a flashlight on it.

Thanks John for the suggestion on the diodes. I've read conflicting things about them (was it RG who said they're "amulets against the voltage gods"?) but it probably can't hurt in this case. What about caps in the same positions, like Music Man uses in the HD130? I've seen those caps occasionally fail short so I can see the tradeoff there, but if the caps are rated high enough (say, 3kV ceramic or those "X" caps) is it any benefit?

[jrfrond]

Caps can fail, so can HV diodes. I don't like the cap treatment personally The diodes kick in only on HV spikes. Can't hurt, and there's no sonic difference. I've used them on all new builds for bomb-proofing, along with increasing SG resistor value and slightly cold biasing. So far, so good. Just don't buy garbage. I purchased the original Rectron type from Mouser for 39 cents ea./100. At bargain houses, most times you don't know what you are getting.

[Mark Black]

In my experience: 95VAC at one place and 130VAC at another = poor mains neutral bond in the AC panel or associated equipment. What happens is the heaviest-loaded hot leg is dragged down and forces the lighter-loaded hot leg up. It can get much worse depending on what is plugged-in/turned-on...I've seen 65V and 170V! The problem could be as far upstream as the utility company transformer wiring on the pole or as far downstream as a sub-panel breaker box inside the building.

I had that problem at a fresh sound system installation where the factory never properly torqued the neutral lug in some brand new feeder equipment (was supposed by the installer to be "good-to-go"), and also had the problem at an old installation (the building I worked in) where the neutral connection lug had oxidized at the utility/user interface box due to age.

If you can locate a known good ground check to see if there is any voltage on neutral at an outlet referenced to that ground. Millivolts up to a few volts are considered normal by electricians, but there shouldn't be anything substantial (like in the 10's of volts). Here in the USA the neutral is supposed to be bonded to ground at the service panel per NEC so other than the drop across the wiring there really shouldn't be much if any voltage between the two.

Edit: the installation electricians at the "new install" where I found the problem were mystified - I had to track down the source of the problem myself with logical troubleshooting techniques common inside an amp but not as often thought of as having to do with the bigger picture. After the problem was located they said something like "wow - we've never seen that before..." Just a FYI.

[Jazz P Bass]

"Never come up with reasons to Not check something."
An Enzo Quote.
True enough.

[Enzo]

If all these amps are heads, and the same cab and cord connected to each one, then the cab or cord may be involved, sure.


But as to the coincidence angle, be aware that more than one person has won the Lotto more than once.