1
\$\begingroup\$

I have taken a microwave oven transformer and removed the 2nd winding which originally put out 2000 volts. I have rewound the 2nd coil to produce 21 VAC which I then rectify and smooth with 2 large capacitors. I am going to use this transformer to make a variable voltage power supply, but noticed that it gets pretty hot. I have added a large heat sink with multiple cooling fans, and it still climbs in temperature a little more than I like. I decided to check how many amps it was pulling from the mains outlet (120vAC) with no load on the secondary. I was surprised to see that it was pulling just over 10 amps. I know transformers have a no load current on the primary but I did not imagine it would be anywhere near 10 amps. So my questions are... Is this normal for a transformer?

Why is it pulling so many amps with no load?

Is there a way of lowering the no load current without sacrificing useable amperage from the finished power supply?

P.S. I am just a hobbyist who is doing this project mostly to learn things in the process but does plan to use the power supply for simple things around the shop. Please keep in mind when giving very technical answers, that I have no formal education for electronics, so explanations are appreciated.

\$\endgroup\$
4
  • \$\begingroup\$ qsl.net/kh6grt/page4/xfmr/xfmr.htm \$\endgroup\$ Commented Oct 23, 2019 at 3:34
  • \$\begingroup\$ @Bruce Abbott - That is a very interesting article. I wonder why MOTs act that way, but regardless, I guess adding more windings on the primary might be the way to go \$\endgroup\$
    – ICEMAN
    Commented Oct 23, 2019 at 4:04
  • \$\begingroup\$ Yes, more turns will help. Microwave oven transformers are optimized for high output in a small size. Since they normally always run at maximum power a high magnetizing current is OK - in an oven. BTW here's another example that drew 10A no load electro-tech-online.com/threads/… \$\endgroup\$ Commented Oct 23, 2019 at 4:30
  • \$\begingroup\$ The next MOT you re-purpose (1) leave the core assembled intact, cut off the secondary windings, remove shunts and filament winding. (2) Thread in some more primary turns until the no-load current gets down to your requirement (3) Thread in your secondary. \$\endgroup\$
    – Neil_UK
    Commented Oct 23, 2019 at 5:03

4 Answers 4

1
\$\begingroup\$

A paper written by D. Ludois, J. Lee, P. Mendoza, G. Venkataramanan suggests that adapting a MOT transformer for other use could benefit from removing the magnetic shunts and the filament winding and adding about 10 turns to the primary. There may be information on the internet that shows how to do that. Look at:

Reuse-of-Post-Consumer-E-Waste-for-Low-Cost-Micropower-Distribution.pdf

  • Document can be expanded to full page or downloaded.

Adding turns will reduce the magnetizing current, the cause of the problem. The primary current is high because the MOT design causes a high degree of iron saturation. Adding turns also reduces the secondary voltage slightly and increases the secondary current for a given primary current. Reducing the magnetizing current reduces the losses making the transformer run cooler for a given load current or allowing a higher load current without overheating.

\$\endgroup\$
6
  • \$\begingroup\$ I have actually already removed the magnetic shunts and the low voltage filament winding. However adding about 10 more turns to the primary would be difficult considering the already added modifications. Would adding more turns lower the amperage and is the a way to know how much it may lower? \$\endgroup\$
    – ICEMAN
    Commented Oct 22, 2019 at 23:11
  • \$\begingroup\$ See addition to my answer. \$\endgroup\$
    – user80875
    Commented Oct 22, 2019 at 23:29
  • \$\begingroup\$ The Ludois et al paper indicates 10 amps peak as the "normal" primary current for an unmodified transformer. The RMS current is a lot less than that. If your ammeter is giving an accurate reading of RMS amps, the transformer is drawing a lot more current than it should be. Removing the magnetic shunts may increase the magnetizing current slightly, but it still seems like something is not right. \$\endgroup\$
    – user80875
    Commented Oct 23, 2019 at 0:08
  • \$\begingroup\$ Well I picked up a cheap clamp meter today for the purpose of reading amps on this transformer and I know clamp meters in general aren't the Most accurate tools. Costing less than $15 probably doesnt help its accuracy either. But being that the power cord gets pretty warm as well, (the original power cord from the microwave) I feel 10 amps is accurate \$\endgroup\$
    – ICEMAN
    Commented Oct 23, 2019 at 0:36
  • \$\begingroup\$ Do you have a shorted-turn in your new secondary? \$\endgroup\$ Commented Oct 23, 2019 at 4:12
2
\$\begingroup\$

Since microwave transformers (MOTs) are made to be as cheap as possible, they are pushed heavily into saturation. A typical MOT with the secondary open pulls 9.5-10 Amp. In my application (USA), I added half of the primary turns, so instead of 100 turns, my primary now has 150 turns. The primary current is now 1Amp with the secondary open. The secondary voltage/turn ratio has dropped by the same ratio, but now the transformer runs cooler and more efficiently.

\$\endgroup\$
1
\$\begingroup\$

Microwave transformers are cost optimised.
They run at high magnetisation current so that the core material is "very well used". Removing the shunts, as you have done, is a good start.

Adding more primary turns may be difficult with the windings that you have added, but, given that you have the now removed shunt volume to work with, this suggests that you MAY have more copper and/or winding area than is strictly needed for your task.

Presumably it's not a 60 Hz transformer being run on 50 Hz - which would add handsomely to your woes.

Ensure that you have minimised core air gaps.
Presumably the laminations were priginally welded and you ground off the welds?
Ensuring that the two parts of the laminations are in intimate contact is "a good idea".
A method of ensuring good clamping force is in order.


Added 2022:

uWave transformers are driven VERY hard. Heating results. Adding 10 turns to primary reduces core flux and saturation. Removing magnetic shunt (funny metal bridge in winding space) helps. 10 turns added to primary can be anywhere in winding space and added in series to current primary winding.

\$\endgroup\$
2
  • \$\begingroup\$ Yes it is 120v 60hz transformer which is being used in America (120v 60hz outlet). And yes you are spot on, I ground off the welds to separate the top and bottom. The laminations were pretty well held together and I never separated them. To rejoin the top and bottom however I placed them in a bench vise after applying some superglue. Not sure if that was the best idea or not. I was just trying to keep as many pieces isolated as possible because of the little bit I have learned about Eddy currents. \$\endgroup\$
    – ICEMAN
    Commented Oct 23, 2019 at 0:38
  • \$\begingroup\$ @ICEMAN The eddy currents are those which follow magnetic paths which do not pass through turns. The main magnetic flux that passes through the windings also flows across the interlamination sections gap. While airgaps are sometimes added to tailor inductance and performance, in this case the gaps need to be as small as reasonably possible. \$\endgroup\$
    – Russell McMahon
    Commented Oct 23, 2019 at 6:12
0
\$\begingroup\$

You should check if your insulated wire has a short. This can happen easily when wrapping around the sharp corners of a laminated core. A scratch on the insulation.

Alternatively, the insulation on the laminated steel plates could be damaged. The plates are oxidized by a very carful process to keep the oxidation very thin to maximize the volume concentration of metallic/magnetic iron. Areas of thick, un-insulated iron will cause huge eddie-current losses which heat the core.

\$\endgroup\$

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service and acknowledge you have read our privacy policy.

Not the answer you're looking for? Browse other questions tagged or ask your own question.