JCM900 Frequency Doubling Weirdness

[J M Fahey]

Weird.

[Ted]

Still didn't quite believe it until I unpicked the winding wires from the tags and swapped the sense of one of the primary windings. Stuck it back in the amp and all was well. Someone has QA issues.

[nickb]

It was a Dagnall, bought direct from Marshall.

That's absolutely tragic to hear. Words fail me..

[Jazz P Bass]

"all is rainbows and unicorns"

That is truly a good thing!

[J M Fahey]

That's absolutely tragic to hear. Words fail me..

Doesn't say much for the transformer manufacturer's final QA process, does it?

True.

That said, this is a "white fly"
manufacturers, and specially one established and massive one such as Dagnall or Drake do NOT test all products, one by one, by any means.

They test samples, say 5% or 1% , not to catch production errors which are assumed non existent, (since they would have been corrected long ago), but mostly to catch materials errors, such as poorer lamination quality, cracked or out of spec plastic bobbins, etc.

So if Joe Worker randomly was too hurried to leave on Friday late shift or still overhang head exploding on Monday first shift or was CRAZY because Manchester United lost 7:0 against Bolivia and so he mis wired one winding , QC wonīt catch that unique event unless by sheer chance itīs the 1% actually tested.

IF it were a regular I bet Marshall ninja team would be at Dagnall/Drake throat in a minute.

That it did not happen points it, in my mind, to an absolutely random affair, not indicative of *average* quality.

[Helmholtz]

Agreed, regarding functional testing.
No component/part manufacturer in the world will/ can guarantee zero functional defectives (shit happens). That's why large quantity purchasing contracts contain agreements on AQL (acceptable quality level) in terms of defectives per million (dpm).

But, FWIW, as an OT is a safety relevant component (isolating hazardous voltages from accessible parts), European safety standards require 100% insulation (hi-pot) testing.