New amp with high voltage on the transformer

[Mick Bailey]

I was checking over the reverb on a newly-delivered amp from a custom builder in the US and while the owner was playing it he got a shock from one of the mains transformer mounting bolts. The MM transformer has a little outrigger bracket to support the weight and a bolt through the top of the cabinet - the chassis is a longer version of a Tweed Deluxe. Anyhow I checked and there's 180VAC riding on the bracket with respect to ground. On the bench the transformer shows 180v right along the laminations (I'm in the UK with 240v mains). The end-bell and bracket aren't grounded. Why such a high voltage? Usually there's just a volt or so. There's no significant ground leakage to the chassis and grounding the bracket with a clip lead 'fixes' it but the laminations still show 180v.

[Leo_Gnardo]

There's no significant ground leakage to the chassis and grounding the bracket with a clip lead 'fixes' it but the laminations still show 180v.

Am I reading this right? Your ohm meter shows you infinite ohms from any winding to transformer lams, bells & bracket? It might take a "megger" insulation testing device to reveal the problem, a rare item in build/repair shops. I sure don't have one. Maybe it's time. Problems that don't show up with an ohmmeter that uses a couple of volts, can show up with some big voltage present, that's what a megger does.

Transformers are (presumably) factory tested, to prove no connexion, insulation breakdown, or arcing between windings, lams, bells etc. Mercury must have missed the test on this one. I'd demand a replacement from them, one that HAS passed a proper hi-pot test to replace it. Good luck. You may want to alert the builder in case he has any more dodgy MM iron, for his safety and his customers, it would be a good thing to make sure there's no more like this.

[Mick Bailey]

That's right - no reading and everything looks in good order when unpowered. My trusty Megger is down right now (broken lead to the meter coil). I need to get this repaired and then do an insulation test.

This seems really unusual to me. The end-bell is powder-coated, the laminations oxidised iron and varnished. That bracket is insulated from ground, as is the end-bell.

I may pull the transformer and run it out of the amp to isolate the fault and work from there

My leakage tester shows a ground leakage current of 15uA at the IEC inlet under operating conditions, which is pretty low. I would expect a fault within the PT to give a much higher reading, but maybe the Megger will reveal more information.

Anyhow, two bands, beer and a late night have put the lid on today. New start tomorrow.......

[J M Fahey]

This rises many doubts.

1) WHY aren't lamination and endbells and brackets grounded?
That's automatic as soon as the transformer is bolted through the chassis, and the bolts pressing laminations, sandwiching them between a pair of endbells (or a single one if horizontal mounted) also short everything together.

2) many transformers have an extra "blind" wire to separately ground a copper band surrounding the windings but outside the laminations (inside the bells though) to cut down on stray fields; the guy who mounted that transformer may have miswired it, wouldn't surprise me in the least.

3) in any case I trust way more MM , who *are* transformer makers with a long background, than many a "custom maker" out there, if you know what I mean.

Looks like anybody who can order a kit and barely solder now is a "booteek maker" ; I have seen *horrors* made by such guys.

To boot, they cover their backs by stating : "I use the most expensive parts" which is a guarantee of nothing.

[Chuck H]

I'm going to back Juan on this (*with one exception ). The idea that induced current is putting 180VAC onto any part of a PT is sort of insane. I've tested many PT's on the bench without the bells mounted to a ground and I've never seen this. Sure, there is a possibility the PT suffers a manufacturer defect. I would rule out any possibility of a wiring error first before suspecting that from a unit MM supplied. The fact that the PT is mounted in an isolated fashion would indicate to me that there is cause to suspect the builders awareness and intentions.

*The laminates should be individually varnished and therefor isolated.

[Mick Bailey]

I pulled the transformer and bench-tested it, plus did an insulation test at 500v (Megger now working). No problems whatsoever. There were quite a few construction issues with the amp that at least in the UK aren't viewed as safe (but will be conveyed back to the builder). I installed the PT and rebuilt the PSU how I'd do it myself on a new build and there's now zero voltage on the metalwork.

Have to say I've never come across this type of fault before and it's taken longer than usual to track down because it was built that way. Jack Darr used to say that when you get a fault on a piece of equipment you are not there to re-design it. It once worked before it broke - Just fix it. Well........ turns out not always.

[Chuck H]

Way to manage it I know it would have taken more time to diagnose the original wiring error that caused the problem but I was interested to find out what it was.

[drewl]

Something of mine I had built for myself had a bad input and cathode ground in the first preamp stage.
This would cause the guitar cord ground to float up to the plate voltage.

I'm guessing a bad ground also caused your problem.

[Mick Bailey]

There were a few grounding issues. When I metered the transformer in the amp every component was isolated from ground. So the first thing I did was to make sure everything was grounded when tightened up. The first B+ node also had a bad ground (a badly-flowed eyelet) and the bias circuit daisy-chains through the first node ground. So there was an additional fault of the B+ collapsing under load and the bias votage shifting to -168v. This took extra time to investigate.

In addition the mains ground needed work; the earth tag hangs off one of the transformer bolts and wasn't as secure as I would like. I moved this to a dedicated terminal on the chassis. It has a universal-primary PT and one of the rectifier socket unused pins was used as a terminal for the mains strapping for 240v. That's pretty bad if the socket shorts out and a trap for anyone working on the amp.

The mains was also fused in the neutral side, meaning that in a fault situation the amp would still be live. One of the output tube sockets was partially filled with solder so the tube couldn't make proper contact with the rest of the pins and arced when moved slightly.

All now fixed and ready for action.

[J M Fahey]

Jack Darr used to say that when you get a fault on a piece of equipment you are not there to re-design it. It once worked before it broke - Just fix it. Well........ turns out not always.

Jack Darr's (also Enzo's) quote usually applies to a properly built, specially a Factory product.
Although sometimes "I would have done it a different way", a Fender, Marshall, Laney, Crate, etc. product can be trusted.

Now a homebuilt one ....... think again (and this amp is living proof).

Not "re engineering" by the way since not much Engineering was used in the first place
Expand that to include proper build techniques which any competent Tech handles very well.

For variousreasons that amp was a horror, or worse, a Death Trap.

Your customer owes you his life, and I'm not stretching it.

Don't think he would be much relieved knowing his relatives could sue the original builder into ruin.

Just curious, what brand or maker did this amp come from?

[Chuck H]

Just curious, what brand or maker did this amp come from?

A bold but fair question. It's reasonable to expect discretion from comrades in the face of an embarrassment. Except when your ignorance endangers people and your only justification is arrogance. The builder must have had the amp on a bench of SOME kind and should have been aware of the danger, OR, even if they didn't know just exactly what they were seeing, that there was a problem. And they sold that amp anyway. Bad form.

[Mick Bailey]

Well, it came from a highly-regarded builder that has a who's-who celebrity client list. It goes against my personal philosophy of openness here, but I'd prefer not to name and shame at this stage unless he turns out to be an arrogant pr*ck. My customer doesn't want any bad publicity by association, so it puts me in an awkward spot.

I'm hoping my feedback gets noted and improves the build quality.

[Tom Phillips]

Well, it came from a highly-regarded builder...

Interesting how that happens. I've seen some very bad workmanship from well known builders. One was due to farmed out work. Nothing wrong with hiring help but the individual with the well known name is still responsible for QC. In another case the owner proudly proclaimed that the amp had been worked on by "the man himself" but wanted to know if II would take a look at the continuing problem so he didn't have to ship the amp back to the big name. It was a repair of a vintage Fender Harvard that still had an intermittent problem and it was a mess inside. It turned turned out to be far less effort to locate and repair the original problem than it took to clean up the mess the "guru" created as he cut wiring and indiscriminately replaced parts during his flawed troubleshooting.

[g1]

We can speak of what should have been, but can anyone recommend a QC test that would have detected this issue?
Does anyone here routinely check for voltage on the PT itself?

[Chuck H]

We can speak of what should have been, but can anyone recommend a QC test that would have detected this issue?
Does anyone here routinely check for voltage on the PT itself?

Good point. But if your of a mind that iso mounting the PT and grounding at a PT bolt is a good idea then you have no credited basis for it and therefor should question everything you do. Lest you forget to breath or something.