DIY Isolation transformer question

[Cliff Matthews]

I have 3 pairs of old Hammond transformers (~80va). To create isolation transformers, I'll wire them back-to-back and in-phase. I also propose to mount them side by side to increase the field coupling with only a thin piece of plastic between them (edit: see 2nd pic with half-cycle arrow overlay). Will this create any caveat's, adverse effects, or overheating?

[mij59]

Keep the transformers apart, there is hardly any coupling between the transformers.

[Shock]

If you cleaned those up for sale you could probably buy a nice secondhand isolation transformer with the proceeds. Isolation transformers aren't commonly sold anymore (unless you pay serious cash for a new one) you have to keep an eye out locally, I managed to pick up an 240V 8A to replace my 3A.

1000

[Karel]

Problem with ordinary isolation transformers is that they still have high parasitic capacitance between between primary and secondary side.
Hence, the leakage current is rather high. In case you want to do sensitive measurements, it's better to use medical grade iso transformers.
They usually let not overlap the primary and secondary windings which garantuees a much lower parasitic capacitance between both sides.

https://toroid.com/Transformer-Products/Isobox-Transformers

Leakage current is less then 100 uA.
Ofcourse, they are more expensive...

[dom0]

They usually let not overlap the primary and secondary windings which garantuees a much lower parasitic capacitance between both sides.

This is only done for low power transformers with EI or UI cores ; toroidal medical grade transformers have one or more shield windings between primary and secondary.

[Cliff Matthews]

Thanks for all the replies! I'll be getting a real isolation transformer next year I hope, since 80-90VA is a bit small. FWIW, I put the 10v/8amp units on a 200 ohm load for about 2 hours at various proximities with almost no difference in output. I don't know exactly why, but the 1st transformer (line side) was significantly warmer than the second, but it never exceeded 35deg-C.

Tests with input @ 122v and no load output @ 122v

450 ohm - 116.5v @ 0.25a (secondary 11.15v @ 3.6a)
200 ohm - 110.7v @ 0.55a (secondary 10.75v @ 6.0a)
150 ohm - 107.4v @ 0.71a (secondary 10.64v @ 7.7a) <-- Edit: I think this a useful limit..?
100 ohm - 100.8v @ 1.01a (brief test only)
50 ohm -  86.0v  @ 1.72a (brief test only)

[T3sl4co1l]

Parallel transformers may not share evenly due to mismatched turns or coupling.  Series one side, parallel other side is ideal.

Common transformers aren't good for back-to-back use because their regulation is nominal to begin with; two in cascade means the regulation is twice as bad, so you get half the VAs for the same regulation.  And for an isolation transformer, you probably want unusually good regulation to start with: you're emulating the AC line, and extra sag can actually increase fire hazard -- fuses don't blow as quickly!

(Anecdote on that one: I was working on a single switch flyback SMPS.  The transistor overheated and shorted.  The circuit was fused, but the fuse took several seconds to blow, due to the limited short-circuit current available from the isolation transformer.  The transistor erupted into sooty flames until the fuse finally went "splat".   That was using a 500 VA "Control Transformer": an industrial, 120+120:240+240 isolation/step-up/down transformer.  I had both sides wired in series, so it's actually 240:240 500VA.  Using it at 120V actually gives me 1/4th the VA capacity for the same regulation.)

Tim

[Cliff Matthews]

Parallel transformers may not share evenly due to mismatched turns or coupling.  Series one side, parallel other side is ideal.

Ok, I can live with that. As it turns out, I can get 120v (210VA) by coupling two 38v (61VA) and one 44v (88VA). I'll see what the no-load voltage is going to be. Barring that, I could try two 44v's with three 10v (80VA) to get 118v and 416VA (quite a few failure points though).

edit: A chain is only as strong as its weakest link. In option 1, the 38v secondaries max out at 1.6a. Taking option 2, the weakest link is the 2 amp secondaries on the 44v transformers. So a decent combination still needs to be found and tested.

[BradC]

If you cleaned those up for sale you could probably buy a nice secondhand isolation transformer with the proceeds. Isolation transformers aren't commonly sold anymore (unless you pay serious cash for a new one) you have to keep an eye out locally, I managed to pick up an 240V 8A to replace my 3A.

I picked up a 10A ex-hospital unit for less than scrap metal value a while back (speaking of backs, I nearly put mine out lifting it). Apparently they are still commonly used in some medical facilities, so it's worth keeping an eye out.
I found this one locally on e-bay.

I've done quite a bit of work in private hospitals lately and noticed that often surplus equipment gets stacked out around the engineering departments. When I asked I was told that most of it was sitting there waiting for someone to exchange it for a donation to their respective charities. I scored a beaut ex-operating theater ducted heat pump which now cools my whole house for a couple of hundred bucks donation and a bag of tinned food.

Keep your eyes open, and never be afraid to ask. The worst that can happen is you get told "no" (or prosecuted for trespassing). Also some of the scrap metal places have bins for mixed metal (copper/steel primarily). Those often contain motors, hermetic compressors and transformers. The lazy scrap metal guys leave them in-tact. The sharp ones slice the core open with an angle grinder and separate out the copper. You can generally buy stuff out of there for scrap value plus 50% ( a couple of bucks a kilo at most ).