Transformer issue

--
Sounds reasonable; good luck! :-)

JF

Thanks to (almost) all for helping here. I'm pretty confident, now, that we can fix this thing in short order.

"ODS Practicum: Overview The Office of Dietary Supplements (ODS) at the National Institutes of Health (NIH) is offering..." "Welcome to the Oregon Dressage Society An affiliate of the United States Dressage Federation that seeks to educate, promote and inspire.." "Official Document System of the United Nations (ODS documents.un.org/..."

Seems the baby gull has no clue about your reference...

There are hundreds of reliable transformer shops in the US which implies similar competition in the EU. And making them is no big deal. What the heck is "special" about them anyway?

** Got no idea who the "baby gull" is but my comment is very simple -

try saying it aloud.....

.... Phil

--
"Odious"???

That's why I'm considering joining the Organization for the Elimination
of Acronyms or, as it's known around here, the OEA.

JF

They work.

When I was a kid, I played around with transformers a lot. I disassembled quite a few and rewound many of those for my own purposes.

So, I have a bit of experience as to what happens when part of one winding shorts. Open circuit inductance by many measurements falls greatly, but the transformer can largely continue to function as a transformer at audio frequencies, sometimes even at 60 Hz.

If only a smaller percentage of the primary is shorted, changes in open circuit primary inductance are large, with the normal inductance of the unshorted portion being paralleled by leakage inductance resulting from coupling between the shorted and unshorted portion. With resistance of the shorted portion being in series with the leakage inductance, I can imagine inductance effectively varying with frequency (for equating impedance to one resistor and one inductor in series). I wonder if this can account for your non-constant LR time constant.

As for how much your DC resistance dropped: Did you measure the DC resistance of an undamaged transformer? If that is smaller than nominal, then the change is less and the change in turns ratio is less. Meanwhile, if part of the primary is shorted, increasing losses of the transformer could cancel an increase in voltage gain from increase in turns ratio.

- Don Klipstein ( snipped-for-privacy@misty.com)

I would dissect one or two of them myself - unwind them. I have been there, and found that turns tend to stick to each other (obvious when unwinding) where that happens. If only some of the damaged turns stick, any sticking indicates that the insulation is being softened to melted by heat.

- Don Klipstein ( snipped-for-privacy@misty.com)

That would require outright carbonizing something or shorting at least two turns to the outer legs of the core, or adding conductive material into the transformer from outside.

- Don Klipstein ( snipped-for-privacy@misty.com)

Will they sell you 1,000 of them for ~$1/ea, though? -- That sounds about like the price and quantity that Keith is shooting for, and AFAIK U.S. shops can't really compete with that.

Goo - gull...

Cheap (in all senses of that term) but defective VS moderate cost and NO problems (for you as well as the customer). But then again if you have money you can afford to piss off the customer (who will spread the word).

at

HE IS THE SPITTING IMAGE OF YOURSELF TROLL WEIRD JUSTICE BUT JUST, ALL THE SAME

ENJOY THE FRUITS OF YOUR LABOR

I AM PROTEUS

Fair point, but even with all costs considered, for the same quality level there are very few products that can be manufactured as inexpensively within the U.S. as off-shores these days. Now, I'd agree there is a certain marketing value to being able to proclaim "made in the USA!" and I'm all for keeping jobs in the USA in general, but realistically there's no piece of electronics made today that doesn't consist of significant portions manufactured off-shores... after all, it's not like TI or Analog Devices or Linear Tech has had the bulk of their IC manufacturing within the U.S. for years now, you know?

It's also not obvious that in Keith's specific scenario here -- a problem brought about due to the relatively recent RoHS nonsense -- that a U.S.-based manufacturer would be any further "up the curve" in learning how to "RoHS-proof" their transformers than the current vendor is. Indeed, many U.S. manufacturers might still be beyond their, e.g., Asian counterparts due to their quantities being so much lower!

Agreed, pissing off your customers is very, very costly.

---Joel

I've been thinking about that. I'm off this week so won't get a chance to try it. I'd like to take a bunch and test them before and after a run through the oven, too.

They function, but very poorly. We primarily see the difference in the losses, though sometimes impedance matching (two transformers in a bridge sort of topology). If the damage is small, it'll pass functional test. We now have that taken care of.

Perhaps. It really looks like two distinct time constants, one being the original and one 2000x less. Where the TC change occurs seems to vary with the number of shorted turns.

I'm not sure what you're asking. We were measuring the resistance of

*all* transformers and failing those with a low resistance. The worst being about 2/3s of nominal. The turns ratio (measured with a 1kHz sine) seems to be constant in almost all samples. It seems strange that open-circuit losses exactly cancel the number of shorted turns, but it would explain the measurements.

...and they work in the circuit. That's pretty much it in a nutshell.

Do you really believe that US manufacturers have no manufacturing issues?

Certainly, but you've offered nothing that shows that US manufacturers are perfect. BTW, this isn't the only place that RoHS has bitten us in the ass.