Finally figured out the operational specs on that old metal box

I pulled something like this out of a dead wire welder I bought at an estate sale a few years back. No markings anywhere, same story with the porcelain bushings and heavy studs. What finally worked for me was doing a ratio test with a small wall wart transformer I had in the junk drawer. Fed a low AC voltage into what I figured was the primary, then measured what came off the other winding with my meter. The ratio between the two readings tells you pretty quickly whether you're dealing with something modest or something that bites. That's how I found out my mystery unit was nowhere near what I guessed from its size alone. It looked beefy but was really just a chunky low voltage unit with oversized cores.

One other trick that saved me some guessing, I photographed every angle and posted it on a welding forum, and an old timer recognized the lamination pattern from a specific line of machines. Sometimes the shape of the core stack narrows it down better than any stamp would. I'd be careful trusting a voltage estimate just from wire gauge and physical size. I've been fooled twice doing that, once with a choke I assumed was rated way higher than it acutally was.

I had a similar unmarked unit years ago, pulled from an old sign ballast setup, and I went down the same road of guessing from the size and the windings. My eyeball estimate was off by a wide margin. What looked like a low voltage part turned out to push out several times what I figured once I actually tested it, so I stopped trusting visual guesses on these. What worked for me was bringing it to a local motor rewind shop. The guy there put it on his bench and ran a proper test for about the cost of a sandwich. Turns out those shops see mystery transformers all the time and can pin down the rating in minutes. Worth a phone call if you've got one nearby.

If you'd rather test it yourself, do it gradually. Feed the primary from a variable supply and creep the voltage up while watching what the secondary does with a meter. That way you learn the real ratio instead of betting your estimate on wire gauge, which can fool you since diferent eras used very different winding practices. The old ignition units especally can look deceptively modest on the outside but step way up on the output side. Porcelain bushings alone don't tell you much, plenty of humble parts wore fancy insulators.

I once found a similar part on the back of an old radio, and instead of checking the wires, I just looked closely at the thick plastic coating on the leads. If the plastic looked brittle or had any tiny little cracks, I figured it was probly not worth testing at all. It felt safer to just skip it then. Do people ever look at the thickness of the metal casing itself to try and get a hint about how much power it handles?

I totally agree with the ratio test method since those big old parts are often misleading. It's always safer to verify the output than just eyeball the core size. When I had a similar transformer piece, I didn't use a wall wart, but I did peel back the outer layer and look at the winding insulation itself. The gauge and quality of the enamel on the coils gave me a much better sense of its actual current carrying capacity than the heavy porcelain did.

Update: tried the ratio test with a small plug-in transformer like the first guy suggested, and it worked. Finally know the turns ratio, so the output side makes sense now. My eyeball guess was way off, just like the sign ballast guy warned. Still stuck on how much current I can safely pull from it though. Any way to estimate that without a label?