Author Options:

Rewinding a MOT to make a high current supply. Answered

New Question:
Ok, I've done a little more research on this, and it seems that MOTs are just pieces of junk basically. Manufacturers design them to go far into saturation because they don't care how much electrical power is wasted, as long as the transformer is cheap to make. The air gap is probably there just so it's easier to cool with a fan. Now I need to bring this thing out of saturation. There are about 90 turns on this primary, which I want to be the secondary at 30 V. So I would have to wind a new primary with 360 turns of thin wire. That's a lot of work, and a lot of wire. Is it worth it? The other option would be to add turns to the existing primary. This seems more reasonable, but just how much would I need to use to bring it out of saturation?

Original Question:
So, I'm a physicist (in training) and I know the theory behind transformers, but actual transformers are still pretty mysterious to me. I've found quite a few Instructables and other sites related to MOTs, specifically related to making high current power supplies with them. The problem I'm having is that even when the secondary is open, and indeed when the secondary has been completely removed (and the shunts removed), the primary draws upwards of 10A when connected to mains. To me, this is unacceptable. Transformers should draw significant power only when current is drawn from the secondary. I'm trying to figure out why (specifically) it does this. I've got a variac and the transformer only begins drawing more than about 2 amps once the variac is set to about 90V. This seems like non-ideal behavior to me, but what I know about saturation tells me that the amount of steel the core has is more than enough to stave off saturation at open secondary, but I could be wrong. The only thing I can think of is that the primary coil has about a 1cm air gap on either side of the core. Since the efficiency is related to the magnetic flux, and I'm guessing the area of the air gap is roughly 1/5 of the area of the core, this could cause significant inefficiency.

Here's what I want:
I want to make a ~30V (possibly with multiple taps) high current supply out of this MOT that draws less than 1 amp when the secondary is open. I tried connecting the mains to the secondary before I removed that and I got about 5 volts on the primary, but I need more than that, and the windings were so saturated with resin I had to just remove the entire secondary. I'm thinking about winding a new primary (using the old primary as the new secondary) with something like 18 gauge wire that would give me 30V. Would this work? What I really need to know is why the existing primary is drawing 10 amps. That's just ridiculous. If I can solve that then the problem is easy.

Update: There is a difference between resistive and reactive current, and I know that reactive current actually draws net zero power because it's 90 degrees out of phase with the source voltage. Capacitive and inductive loads do this apparently. Is this the reason for the massive current draw? If so, I might be able to minimize the current draw from the mains line by putting a matched capacitor in parallel with the primary. But this wouldn't stop the transformer from heating up.


I have made some similar steps towards a constructing a high current supply from an old MOT.  I cut out the high-voltage secondary winding, and replaced it with about two or three turns of the thickest wire I could find.  I am trying to remember what I wanted to do with this setup.  I think I wanted to make spot-welds, or weld thermocouple junctions, or something.  Of course I got distracted by something else, and this modified transformer wound up in a corner collecting dust.

What I wanted to say was that I also observed large amounts of current through the primary winding, when this winding was connected to the mains supply (approximately 110 VAC, in my home country), and that was with no load, i.e with the secondary open.  I think the magnitude of the current was about 8 A, RMS, and I should note the tool I was using to measure this current indicated that it was largely a reactive current.    The tool, BTW, was one of little Kil-a-Watt(r) power measuring gizmos.  I mean there was likely some significant real power being wasted too. I would have to measure it again to tell you how much.

I think I was getting results similar to yours, but I wasn't too worried because, for what I was planning on doing,  I was not going to just leave it turned on for a long time, but rather just turn it on for durations of just a few seconds at a time.

Anyway, I could set that thing back up on the bench, and take some pictures, and/or measurements, if you think that might be helpful to you.

Oh, don't go to the trouble, that's pretty much what I expected. That's nice that it can tell you if it's reactive or resistive loading. The reactive loading is actually not a problem at all in an ideal world with no resistance, but 10A wreaks havoc on the non-ideal components.

I need to significantly increase the inductance. The simulation I'm running says I need about 200mH of inductance to reduce the open current to 1.5A, which is still pretty high. This would mean more than quadrupling the primary turns. With the tools I have, this is pretty much impossible. Plus, the primary wire is too thick to just add on to. I'd need to completely rewind the coil myself.

I have no need for a spot welder, so I guess this lump of steel is going in the corner to collect dust for a while ;)

Have you considered using a large ZVS royer oscillator configuration to achieve high currents in a high turn-ratio transformer?

What if you were to increase the driving frequency to about 400 Hz? like what US military used to use. Hopefully eddy currents in the E core won't be as big a problem.

What if you added a capacitor in series to the primary? Then maybe it will lower the resonant frequency to something closer to 60 Hz.

I personally, know nothing much about transformers or inductors either, just some thoughts

By the way to remove the wire coils safely put the MOT inan old pot add water and boil for 20 minutes. Then when hot using gloves remove the windings.

Well you could take 2 cores from 2 MOT transformers. Take the two Ш cores. Put then onto each other. You will get one long core.

Remove the primary cores from same type size etc. MOT transformers and put them on your new long core transformer. Wire in series. This will bring the transformer out of saturation for free. If you use 4 cores you will have same amount of space left to wind secondary as you do on singe MOT.

You can also use this trick to have better spot welder as more thick wire will fit on to the long transformer then short one.

Out of interest, does the transformer get hot when its drawing 10A for nothng ?

The primary has 0.68 ohms of resistance, and I measured the inductance to be 41mH. Drawing 10 amps would then be (roughly) P=I^2R= 68W of resistive losses, plus whatever losses occur in the core during hysteresis, and that's with the secondary completely removed. When I actually hook it up for about a minute, the temperature of the primary is probably about 20 degrees warmer.

Hmm, I don't think so, it was working fine before I removed the secondary. I drew normal MOT arcs with it and everything. It was drawing 10A before I cut the primary off.

I know this post is old as dated so, but still out here for us experimenters! lol

If you havent given up on it- and the thought of the last comment of a shorted turn, A megohm meter will test for shorts in coils such as your txfmr up to around 9kv. depending on model you have.

Ok, after much hmm-ing and haw-ing, I've decided to abandon this line of inquiry. In order to bring the MOT out of saturation and keep the open-secondary current below 1.5A, I would need 440 turns of primary. Then I would need 110 turns of secondary to get 30V. To maintain a 1:4 primary to secondary current loading ratio, as well as only using the area available for winding the core, I would need to use 30 meters 12AWG and 120 meters 18AWG enameled wire. At roughly $50 per 30 meters, the project would cost at least $250.

In short, f*** that.