Author Topic: MOT (microwave oven transformer) cooling.  (Read 11014 times)

0 Members and 1 Guest are viewing this topic.

Offline davelectronicTopic starter

  • Frequent Contributor
  • **
  • Posts: 772
  • Country: gb
  • Life's too short.
Re: MOT (microwave oven transformer) cooling.
« Reply #25 on: January 05, 2019, 03:17:09 am »
I am limited as the property is a 16 floor tower block, I doubt, or through limited ventilation couldn't use a lead acid battery. And it's really just for the fun and something for nothing. Although all told it still cost something. Would a high rated triac type dimmer work ? I know they only convert the wave form. But unsure if that would be suited to an inductive load. I know 1000 watt dimmer type switches are readily available. But will that create heat as it's a modified wave form.
 

Offline Circlotron

  • Super Contributor
  • ***
  • Posts: 3362
  • Country: au
Re: MOT (microwave oven transformer) cooling.
« Reply #26 on: January 05, 2019, 03:45:16 am »
If it’s just for fun then it’s perfectly valid in my books...
For a power supply of that current I’d consider using a choke input filter. It will make life easier for the transformer and filter caps, and the ripple will be more or less sinusoidal, not sawtooth, so easier to clean up. Caps and bridge rectifier present a power factor of about 0.6 to the transformer with very high current peaks, but add a choke and those peaks (and the consequent transformer I2R copper losses) go way down. The transformer secondary voltage will have to be higher though. Instead of the dc being 1.414 x the rms volts it will be about 0.9 x the rms. Plus the rectifier voltage drop of course. Back in olde tymes with really serious power supplies the choke input filter was king.
« Last Edit: January 05, 2019, 03:47:15 am by Circlotron »
 

Offline bsfeechannel

  • Super Contributor
  • ***
  • Posts: 1668
  • Country: 00
Re: MOT (microwave oven transformer) cooling.
« Reply #27 on: January 05, 2019, 04:03:07 am »
MOTs are designed to work under heavy saturation at their rated mains voltage. That's why they overheat without forced air cooling (a.k.a a fan).

To "correct" that you have three options: reduce the primary voltage , enlarge the cross sectional area of the center leg of the EI core or add turns to the primary.

Since the first option is a bit difficult for you and the second impossible, the third, although a bit crazy, is probably the most viable.

By my calculations, to tame that beast you'd need some additional 200 turns on the primary.
 

Offline davelectronicTopic starter

  • Frequent Contributor
  • **
  • Posts: 772
  • Country: gb
  • Life's too short.
Re: MOT (microwave oven transformer) cooling.
« Reply #28 on: January 05, 2019, 12:16:42 pm »
I couldn't fit another 200 turns on the primary, just not enough area to add these turns. I'm still wandering about a triac type circuit, I know strictly speaking it doesn't lower the input voltage. But rather cuts off part of the time a certain area of the sin waves on period. Modify the wave form, but unsure if it's a viable option due to heat build up from the modified wave form the triac type unit would supply to the primary.
 

Offline joseph nicholas

  • Frequent Contributor
  • **
  • Posts: 408
  • Country: mx
Re: MOT (microwave oven transformer) cooling.
« Reply #29 on: January 05, 2019, 12:29:33 pm »
If you use a triac rated for higher power, it should work.  Just remember ohms law and you should be ok.  I built a soldering iron using a triac and and a MOT and it works well, just get a high power triac, if there is one rated for the job the MOT's will be doing.
 

Offline Ian.M

  • Super Contributor
  • ***
  • Posts: 13216
Re: MOT (microwave oven transformer) cooling.
« Reply #30 on: January 05, 2019, 01:11:26 pm »
I couldn't fit another 200 turns on the primary, just not enough area to add these turns.
You'd need to build a former in-situ (assuming you haven't cut the welds holding the core together) next to the primary out of presspahn, fish paper, and linen tape, using phenolic adhesive to hold the extra 200 turns  and provide reinforced insulation between it and the secondary and core, then wire it in series being careful of the phasing. That would reduce the available area for the secondary, so unless you start with a MOT from a 1200W oven you'll not get the power you need out out of a single MOT.   2x 800W oven MOTs, primaries in parallel, secondaries in series should be viable. N.B. as the primary current rating doesn't increase, you need to derate the total permissable VA by the ratio your extra winding bucks the line voltage by.

It would be worth looking at Circlotron's suggestion of a choke input filter after the bridge rectifier, possibly using the core from the second one with the center leg cut to air gap it o t doesn't saturate when used as a high current choke.   That arrangement has the potential to make about 40% better use of the VA rating of the main transformer.

If you use a triac rated for higher power, it should work.  Just remember ohms law and you should be ok.  I built a soldering iron using a triac and and a MOT and it works well, just get a high power triac, if there is one rated for the job the MOT's will be doing.
The TRIAC firing angle must be very close to exactly the same on both half cycles.   Any significant assymetry will allow DC current buildup leading to saturation and a 'Weller style'  :-DD primary meltdown.
 

Offline T3sl4co1l

  • Super Contributor
  • ***
  • Posts: 22436
  • Country: us
  • Expert, Analog Electronics, PCB Layout, EMC
    • Seven Transistor Labs
Re: MOT (microwave oven transformer) cooling.
« Reply #31 on: January 05, 2019, 01:16:00 pm »
How would a net field "build up" when the winding is open-circuit for part of the cycle?

200 turns sounds like overkill.  Again, probably 10 or 20% is all that's needed.  I'm guessing the original primary is only about 300 to begin with.

Tim
Seven Transistor Labs, LLC
Electronic design, from concept to prototype.
Bringing a project to life?  Send me a message!
 

Offline davelectronicTopic starter

  • Frequent Contributor
  • **
  • Posts: 772
  • Country: gb
  • Life's too short.
Re: MOT (microwave oven transformer) cooling.
« Reply #32 on: January 06, 2019, 12:54:39 am »
Lot of options available really. I kind of thought a triac type unit wouldn't work, I remember reading about heating being a problem to do with altered phase angle of the AC. The triac tricks the transformer into thinking it's less voltage by changing each half cycles of the wave form. I know it doesn't lower voltage AC that is. Only a variac will do that. I only ever split the welds on one previous MOT, on reasembly it was far worse magnetic circuit after that. I'd sooner take longer winding but maintain the iron integrity. I've no idea why the magnetic circuit was worse, but it definitely was. I remember putting a 300 watt halogen lamp load on the uncut weld MOT, the voltage drop was quite small, even with that load.
 

Offline Ian.M

  • Super Contributor
  • ***
  • Posts: 13216
Re: MOT (microwave oven transformer) cooling.
« Reply #33 on: January 06, 2019, 02:15:14 am »
How would a net field "build up" when the winding is open-circuit for part of the cycle?

200 turns sounds like overkill.  Again, probably 10 or 20% is all that's needed.  I'm guessing the original primary is only about 300 to begin with.

Tim
Yes. thinking about it, the flux shoud reset during the off time. However about forty years back I saw a TRIAC dimmer circuit smoke a transformer.   It was almost certainly a DIAC based firing circuit, which would only have generated one firing pulse per half cycle in normal operation, and I can only suppose that residual flux caused enough asymmetry during the first cycle that commutation was delayed till after the next half cycle's firing pulse, which as a result led to a rectifying action, and the TRIAC must have been really beefy compared to the primary's current rating, so it didn't simply fuse the TRIAC, restoring a normal full-wave supply to  the transformer before it cooked off.

On the 200 turns isssue - yes that does sound excessive. Bsfeechannel dropped in that number as a result of some 'calculation'.   It would make more sense to ramp up the supply with a Variac, carefully monitoring the primary current, preferably with an isolated current probe to let you see the waveform, to determine the voltage at onset of saturation, then check the voltage on a 10 turn temporary secondary winding to let you calculate the volts/turn and thus determine how many additional turns would be required to shift the onset of saturation to 10% above the nominal supply voltage.

@Davelectronic,
Yes splitting the core is a last resort, as it almost invariably introduces a slight air gap unless you can clamp it with enough pressure and re-weld it afterwards, or totally grind out the welds and rebuild it with interleaved E laminations from each end with the I laminations in the gaps between them either end, but that always seems to have issues getting the last few laminations back into the primary's coil former, so you end up leaving a couple out making the saturation problem even worse.

If you've still got the one you've split, it would be a good candidate for a core for the DC choke - ideally grind down the center leg to gap it, but if you don't mind a high external field, you could gap it with layers of varnished paper or thin card right across between the E and I.

It may be worth considering a separate buck transformer.  e.g. to buck the 240V supply to a 1 KVA MOT by 24V, the buck transformer only has to handle the MOT primary current of a bit over 4A, so a relatively small 120VA transformer with its secondary in series with the supply to the MOT will do that with plenty of margin.
« Last Edit: January 06, 2019, 07:56:58 am by Ian.M »
 

Offline bsfeechannel

  • Super Contributor
  • ***
  • Posts: 1668
  • Country: 00
Re: MOT (microwave oven transformer) cooling.
« Reply #34 on: January 06, 2019, 04:26:30 am »
On the 200 turns isssue - yes that does sound excessive. Bsfeechannel dropped in that number as a result of some 'calculation'.   It would make more sense to ramp up the supply with a Variac, carefully monitoring the primary current, preferably with an isolated current probe to let you see the waveform, to determine the voltage at onset of saturation, then check the voltage on a 10 turn temporary secondary winding to let you calculate the volts/turn and thus determine how many additional turns would be required to shift the onset of saturation to 10% above the nominal supply voltage.

Challenge accepted.

Using a variac I plotted the primary idle current (i.e. the magnetizing current) on the primary against its voltage.



If we can agree that after around 0.5 amps (beware that on my graph commas stand for dots) the transformer really gets saturated, its voltage shouldn't be greater than 160V. The volts/turn ratio of this transformer is 1.1. So the primary has 200 turns (hence the magic number I proposed earlier). 160V/200 turns gives us a 0.8 volt per turn ratio. To get back to 220V on the brink of saturation we would need (220 - 160)/0.8 = 75 turns. Much better than 200 turns.
« Last Edit: January 06, 2019, 04:28:59 am by bsfeechannel »
 

Offline Ian.M

  • Super Contributor
  • ***
  • Posts: 13216
Re: MOT (microwave oven transformer) cooling.
« Reply #35 on: January 06, 2019, 05:12:08 am »
I agree that 160V is the beginning of hard saturation.  Assuming your nominal supply voltage is 220V, you still need some margin for supply voltage variation.   10% overvoltage is a reasonable safety margin for most developed countries but if you are in a 3rd world s--thole*, you may need more margin.   Lets run the numbers for 10%.   Max input voltage is 220V * 1.1 = 242V.  242V - 160V = 82V.  82V / 0.8V/turn = 103 turns.

That's about half the turns of your original estimate, and if the primary fills the bobbin and is a close fit in the core window, is unlikely to be practical to add as an overwind on top of the primary, so its going to be a total PITA to have to build a reinforced insulation bobbin for the extra turns  as you cant simply 3D print two half bobbins and glue them as it wont take the heat, hence my comment earlier to use presspahn board, fishpaper and linen tape (which must be bare linen or with a thermosetting adhesive) and phenolic adhesive.

*  If you are in North America in a property with split phase 120V/240V, and old wiring, backstab outlet connections, or aluminum wiring, even 20% margin wont prevent a 'Weller style' meltdown from a bad neutral so for <expletive>'s sake fuse the primary properly!
 

Offline bsfeechannel

  • Super Contributor
  • ***
  • Posts: 1668
  • Country: 00
Re: MOT (microwave oven transformer) cooling.
« Reply #36 on: January 06, 2019, 06:43:20 am »
Around here, 220V installations are allowed to go from 202V to 231V. Up until very recently we were serviced by a subsidiary of AES Corporation. I can only hope they were not based in one of those 3d world sitholes (I learned that spelling with AvE).

Anyway, I think we are running out of options.

I use my MOT together with the variac as a poor man's high voltage source. I never go past 100V on the primary, since the high voltage winding has a 1:10 turns ratio and none of my meters can handle more than 1kV.
 

Offline davelectronicTopic starter

  • Frequent Contributor
  • **
  • Posts: 772
  • Country: gb
  • Life's too short.
Re: MOT (microwave oven transformer) cooling.
« Reply #37 on: January 06, 2019, 07:35:05 pm »
This is very interesting data and ideas there. Unfortunately I don't have the split core transformer anymore. Going back a few years, I measured the primary current with no load on the secondary. That was right on 3 Amps with the 240 ish input. A lot of wasted energy. What I'm unsure about, is that current "wattless" a bit like an AC induction motor spinning with no load ? I kind of think the motor scenario and the transformer are not the same, identically. What seems to be the go to option must be primary ballast in series with the primary, so it's not saturated all the time. And adding extra primary winding is another great idea and option, but gets a bit trying on a closed core.
 

Offline bsfeechannel

  • Super Contributor
  • ***
  • Posts: 1668
  • Country: 00
Re: MOT (microwave oven transformer) cooling.
« Reply #38 on: January 06, 2019, 08:53:47 pm »
The current is not "wattless", otherwise your transformer wouldn't be overheating, but part of it is "wattless". It has to be, because part of that energy is stored as a magnetic field in the core. However,  I'm not sure about the proportion exactly.

One thing crossed my mind while I was taking a nap. Normally we remove the MOT's magnetic shunts. What these shunts do is to provide an embedded ballast by increasing the leakage inductance. It acts as a current limiter and as such ends up reducing the voltage seen by the primary according to the model below.



I'm not sure how effective this embedded ballast is, but perhaps it helps reduce the number of turns to be added to the primary to take the core out of saturation. My graph doesn't show, but at 214V, I got 3.4A. However, my MOT is "shuntless". It would be nice if some MOT still intact could be measured for that current so that we could have an idea.
 

Offline T3sl4co1l

  • Super Contributor
  • ***
  • Posts: 22436
  • Country: us
  • Expert, Analog Electronics, PCB Layout, EMC
    • Seven Transistor Labs
Re: MOT (microwave oven transformer) cooling.
« Reply #39 on: January 06, 2019, 10:58:57 pm »
No, the shunt doesn't affect the primary side stats -- it's wound around the same core limb regardless.  The shunts do divert flux from the secondary, which means the secondary side core limb runs less Bpk.

The shunts are usually not the full core cross section, and not gapped much, which should tend to run them in saturation, drawing ~constant flux away from the secondary.  Which means the secondary regulation should still not be terrible, just a lower voltage.  Doesn't seem very useful, I guess it just happens to work out well enough for a magnetron.

The equivalent circuit is two ideal transformers sharing flux; the unlinked flux, is, of course, the leakage. :)

Tim
Seven Transistor Labs, LLC
Electronic design, from concept to prototype.
Bringing a project to life?  Send me a message!
 

Offline davelectronicTopic starter

  • Frequent Contributor
  • **
  • Posts: 772
  • Country: gb
  • Life's too short.
Re: MOT (microwave oven transformer) cooling.
« Reply #40 on: January 09, 2019, 02:21:02 am »
Just a thank you for the help,  ideas and data provided. Generally I've knocked out shunts, having read they help in a crude way to avoid overheating in the transformers saturation. Maybe I should have left them in, and the area that they take up in small. Although a couple more secondary turn can be had with there removal.
It's always going to be a poor man's linear transformer, but it's a means to an end.
 

Offline Wolfgang

  • Super Contributor
  • ***
  • Posts: 1826
  • Country: de
  • Its great if it finally works !
    • Electronic Projects for Fun
Re: MOT (microwave oven transformer) cooling.
« Reply #41 on: January 09, 2019, 01:58:41 pm »
Yeah, right. You cannot expect too much from a junkyard item. If not run in a microwave oven, 300W is realistic with a series inductor and the magnetic shunt out. Still better than nothing if you need some cheap HV.

Happy (and safe) experimenting !
  Wolfgang
 

Offline Berni

  • Super Contributor
  • ***
  • Posts: 5050
  • Country: si
Re: MOT (microwave oven transformer) cooling.
« Reply #42 on: January 09, 2019, 06:13:37 pm »
Id say just wire the primaries of two MOTs in series so they are effectively running on 110V and split your 12V output winding in to two 6V winding on both transformers. Acording to the graph that will bring the magnetizing current down to something normal and save a lot of losses in the primary copper wingdings.

As for the shunt id say just hammer it out. It should make very little if any difference to the magnetizing current on the primary, but will have a big effect on your output load regulation. As you put a load on your secondary coil you don't get all of the flux from it reaching the primary where it accordingly causes the input current to rise and put more power into the transformer. Because of that as the current in the secondary increases and produces an opposing flux it starts to cancel out the original flux and reduce it. Because of that the voltage on the secondary drops.

This makes the shunt effectively a current limiting feature. Without it under a short circuit condition the current is mostly limited by the resistance of the wingdings and this causes a lot of heat. But a magnetic shunt simply hides some of the secondary coils load from the primary causing it to draw less current from the power source. This effect is heavily used in neon sign transformers where they are magnetically shunted so much that you can short the output and leave it running indefinitely without any damage to the transformer.
 

Offline Cliff Matthews

  • Supporter
  • ****
  • Posts: 1911
  • Country: ca
    • General Repair and Support
Re: MOT (microwave oven transformer) cooling.
« Reply #43 on: January 09, 2019, 07:21:36 pm »
Quote from: Wolfgang on Today at 10:58:41 am
Yeah, right. You cannot expect too much from a junkyard item. If not run in a microwave oven, 300W is realistic with a series inductor and the magnetic shunt out. Still better than nothing if you need some cheap HV.


A bit more maybe. I got 330 watts applying 78-vac to 120-vac MOT primaries (testing with lamp cord as secondary winding)
 

Offline Wolfgang

  • Super Contributor
  • ***
  • Posts: 1826
  • Country: de
  • Its great if it finally works !
    • Electronic Projects for Fun
Re: MOT (microwave oven transformer) cooling.
« Reply #44 on: January 09, 2019, 07:28:04 pm »
... some people even use a lamp as a limiting resistor. IMHO the end of the game is reached when the MOT core saturates. From then on it will get really hot. Its also a question of what kind the load is usec (rectifier with cap, resistive, ...). Anyway, as log everything stays cool, why not draw 330W.
 


Share me

Digg  Facebook  SlashDot  Delicious  Technorati  Twitter  Google  Yahoo
Smf