Output Transformer Assembly/Construction. Specifically, final bobbin/EI lamination assembly??

[SoulFetish]

I'm looking for some advice and guidance on something. Lets say I have an Output transformer which otherwise worked flawlessly, but there was and accident and the EI lamination was damaged and broken. The amp was not on, and the bobbin and winding were undamaged and in perfect working order.
How difficult would it be to transfer the bobbin to a new EI core of the same dimensions? I used a caliper to get the dimensions below:


If the EI core is M19/26 gauge steel, what would be the best way to proceed? Are these last few steps something I could do on my own, or would it be better to contact a manufacturer and see if they would reassemble it? I know people (some of the members here) wind their own transformers, so any and all advice is appreciated.
Thanks!

[Dave H]

I've assembled a few transformers. Stacking the laminations was easy enough except for the last few which required a bit of wiggling and some persuasion with a rubber mallet. Make sure you take note of how the lams were stacked on the old transformer so you can duplicate that on the new one.

[SoulFetish]

I've assembled a few transformers. Stacking the laminations was easy enough except for the last few which required a bit of wiggling and some persuasion with a rubber mallet. Make sure you take note of how the lams were stacked on the old transformer so you can duplicate that on the new one.

Stacking, eh? I mean, thats not the worst thing in world. But at that stage, it’s varnishing part I’d be I’d probably be cussing myself out about. Now that I think about it, how would I even avoid it if wanted to?
Wouldn’t it would need to be done after the E/I sections were joined? What is the preferred method of potting or varnishing done by “hand”?

[R.G.]

Stacking the laminations is fairly trivial. Obtaining new laminations in the first place is the trouble. If you can get new lams, you can stack a core in a few minutes. You just start inserting the Es alternately, stack it to nearly full without any Is, then insert the Is. As noted, the last few laminations need (gentle!) persuasion by tapping with a non-metallic mallet in most cases.

When all the Es and Is are in place and the end-lams placed, (gently!) tap the stack to squareness on a flat surface, insert bolts and barely tighten, then tap/square again, tighten a bit more, tap/square a bit more until it's solidly in place and the Es and Is are well seated. The objective of all this tap/square stuff is to get the Es and Is in intimate contact where they butt together. This step can cause as much as a 2:1 increase in primary inductance if properly done. Then loosen bolts and put on belly band, end bells, etc. trying to not disturb the intimate contact you got by jogging.

Once completed, jogged, belled and banded, you can varnish. Varnishing improves heat flow out of the transformer as well as gluing stuff together. Varnishing is messy. Ideally you would submerge the whole assembled transformer in a container of varnish and pull a vacuum on the can. This gets the bubbles out. Then remove the transformer and let it sit on something you will then throw away so it can drip for a while, and put it into an oven to oven dry at low heat for ... a long time. You have to get the solvent out of the varnish, but not heat it so much the solvent outgassing causes more bubbles. A day or two at 200-220F is a reasonable guess. Hot enough to dry, not so hot as to decompose.

In your case, you don't really need to varnish the core nor de-gas/dry the core, baking could be shorter. Maybe.

Getting laminates will be a problem. De-stacking and cleaning another transformer is manual-labor-intensive. Ideally, you would get new laminates with undisturbed oxide insulating layers and stack with new.

The decision about whether to do this yourself or farm it out will depend on how much of your time and trouble you want to put into it. It may be reasonable to get this done by one of the boutique rewinders. You'd need to get quotes.

As to what you'd have when done - good question. Theoretically, just replacing the laminates is straightforward, if labor intensive. But jogging/squaring will have a notable effect on the result even if the coils are completely undisturbed. And keeping them square while assembling and varnishing is a pain. Varnishing has no major effect on tone/frequency response other than to increase the internal self capacitance and cut highs, but that's a minor effect in most transformers.

[Tony Bones]

Edcor sells M6 lams to the common man in reasonable quantities. https://edcorusa.com/m6-steel

[Helmholtz]

Why worry about varnishing or even vacuum impregnation? If the coil was impregnated it still is. If the core disassembles easily it wasn't vacuum potted/impregnated.

[SoulFetish]

I did a bit of digging and put together some info on materials and suppliers if I (or anyone else) decide to take on a DIY transformer project.
The lamination spec for this particular project is:
M19 26 gauge EI-125 (STD) 1-1/4" CENTER LEG
about 1.5" stack, but I'll get an accurate count later on.
The best option for secondary insulation looks like Dolph's AC-43 GL Clear Air Dry Polyester Varnish. see datasheet here:
AC43.pdf
I was also able to find some suppliers who seem to specialize, or at least offer supply for small production/prototyping. I will email them and find out exactly what kind of minimums and cost I'll be looking at probably tomorrow or monday. But here are some of the US suppliers:
http://www.cwsbytemark.com
http://www.protolam.com
http://www.polarislaserlaminations.com
http://laminationspecialties.com
http://www.totalwinding.com/ (supplies the varnish, magnet wire, and other supplies)

A couple of questions on construction details:
I've been hearing conflicting opinions on what grade of electrical steel is best for guitar amps. What are some of the benefits/drawbacks to using grain oriented M6 vs NGO silicon M19?
Edcor pushes M6, stacked for the least amount of air gap for all audio applications, including guitar amps.
But I would think that stacking (butt stacking?) to allow some air gap would be beneficial so the OT is a little more tolerant to imbalance and won't easily saturate with some DC offset.
What say you folks?

[Helmholtz]

But I would think that stacking (butt stacking?) to allow some air gap would be beneficial so the OT is a little more tolerant to imbalance and won't easily saturate with some DC offset.

It is essential to reproduce the original stacking method. Typically (and for good reason) alternating Es and Is for lowest air gap is used for PP and butt stacking (+ spacer) for single ended amps.
Butt stacking increases net air gap and DC tolerance but reduces primary inductance. To compensate (especially regarding bass response) such designs use higher turns numbers.

[R.G.]

Yes, what he said on stacking. Plus - there are a number of different stacking patterns that were (and maybe still are) used in transformers, for different effects. It's not just a binary decision to either fully interleave or butt stack. You can lay two, three, or more laminations the same way, then reverse with another two/three/four. So you can get partial butt stacking and partial interleaved.

Why you'd do this was always a little abstract to me. I looked it up a few times back in the 70s, had my eyes glaze over, and dropped the issue. There was some issue of flux crowding at the end of one lamination when you alternate one and one that was supposed to be a little better if you stacked two and two. As I (dimly, dimly) remember it had to do with heat losses in the core in power transformers. I'd have to go do more research - which is a breeze today on the net compared to ordering technical stuff from libraries in the 70s - to get back some of the arguments. I would follow the advice above - when restacking, go with what you had. When designing, getting enough primary inductance is always hard, and core losses are not so bad in OTs compared to power transformers, so I'd go with 1:1 interleave, as other linterleaves would have lesser primary inductance IIRC. DC tolerance is a real issue, but it ought to be not a big issue except in SE designs.

Grain oriented versus non oriented is an interesting consideration. Grain orientation increases permeability (and hence inductance per turn) for flux along the direction of rolling/orientation, and decreases it for flux across the grain direction. The idea is to use grain oriented iron in a way that makes the flux preferentially use the along-the-grain direction. E-I laminations do this pretty well, as the flux only has to turn the corner across the orientation direction at the ends of the transformer. The question is whether this is an advantage in OTs.

I'm just guessing here, but I can't see much advantage or disadvantage. If an OT is butt stacked, the loss in permeability of the air gap is vastly larger than any change caused by grain orientation. If it's 1:1 interleaved, the grain orientation is a bit of an advantage, but probably not huge. This is a guess, based only on old memories.