Full Wave Microwave Transformer Supply

[WU2D]

I have been collecting a few of the the various junked microwave transformers from the serious beasts of the 80's - 90's that are filling up the metal piles.

There are plenty of reasons to ignore these in favor of real HV iron - But...

I noticed that they all attach one side of the HV to the frame of the transformer. I would assume that they are using half wave self rectification on the oscillator.

Simply separating this winding from ground to float the secondary is not a good idea because even if you could, the dielectric gap is just not set up to permit this to be safe. This generally limits the usefulness of these beasts to standard half wave rectifier and half wave doubler circuits. Actually this is not a bad approach because they have a very high current capability.

The output voltage seems to be between 1200 and 2000 VAC on most of these.
If one was to procure a matching pair, I would think that you could get them to work in a conventional full wave, grounded CT circuit in a safe manner. See diagrams. In A we have a half wave doubler. We should be able to get some fairly high voltage out of this! In B we have a really slick well regulated supply with all kinds of current capability. In C we have a high voltage bipolar supply for a PDM project.

I also have heard that they benefit from some capacitance on the primary since they are a saturating type and the voltage and current get out of phase, causing current surges. They also tend to draw current even without a load and the cap helps this as well...

Ideas/ Experiences?   

Mike WU2D   

[W4EWH]

It's a great idea: most of them can be had for the asking.

Should the schematics show the primary feeds inverted on one of the transformers for the full and bridge circuits?

Bill, W1AC

[WU2D]

Right Bill - Good point electrically - but I guess you can assume that the second transformer is upside down - dunno.

Schematically this is seldom shown to that detail and even if is were (with phasing dots), Murphy says most of the time I would get it hooked up wrong anyway!  I know that when I am ganging transformers, even those that supposedly come from an identical lot, sometimes they mess up the phasing on the primaries, so it is best to test with a variac.

MIke WU2D

[w4bfs]

what is the approx value of the primary cap or do you just choose one that gives minimum current in the primary with no load on the secondary ?

[Tom WA3KLR]

I have one old microwave transformer here that I pulled from a failed oven, never have incorporated it into anything.  I was quite surprised how high the primary magnetizing current was with no load, sign of an inefficient/cheap design.

I wouldn't run it all the time like in linear amp service, but o.k for PTT with an AM rig - a high load that makes the magnetizing current reasonable for the load.

I didn't check it for my rule of thumb at the time, but for conventional 60 Hertz transformers - 15 to 30 watts per pound.  I think we had a thread on transformers a while back - perhaps up to 60 watts per pound for Hypersil type of laminations, if the consumer engineers would have dared to specify this.

Also - Not good to run half-wave loads on transformers.

[Opcom]

I know of several used for half wave and there was not a problem, at least for the short time used with a tesla coil. The shunts were removed to get rid of the unwanted magnetizing current. I think he set up a pair for FWCT with choke input of 10H and 20uF output cap to run 811's.

I am not familiar with ones using a primary cap. Maybe some old ones for ferro-resonant operation? I'd avoid those.

[Jeff W9GY]

The few I have connected to 120 V primary (no sec. load or rectifier) had, at least to me, unacceptle loud mechanical buzz.

[k4kyv]

One of my first rigs used a pair of TV power transformers with the secondaries connected in series to make a full wave rectifier power supply, using the common connection point as midtap.  I used it for several years.  In fact, that's the power supply that I got the arm-to-arm 1000 volt jolt from on election night, 1960.

But that circuit is still a half-wave rectifier for each transformer that includes all the inherent inefficiencies.  I would have done better to have built two regular full wave  circuits and wired the DC outputs in series.

[AB3FL]

Why can't you run a full wave bridge with a single transformer, if you make sure the transformer is isolated (ie up on ceramic stands, plastic, etc)?  I do happen to have one out of a 1200W microwave oven that I want to make something out of....

Tom - AB3FL

[Ed/KB1HYS]

I've built a full wave powersupply using these beasts.

2200-0-2200 VCT. I'm getting 2kv at 500mA no problem (using 866A rectifiers).

I ran it at that load using a single section PI output (7H & 8uF/2.2kv) and nothing got warm or blew up. (yea I was a little worried).   I ran it into a saltwater load at 4000ohms for ~15minutes (old buzzard simulation). No problems other than the bleeder got pretty warm. Muffin fan fixed that.

Get two that have the same part number/manufacturer. I got two that looked the same, but one had aluminum wire on the secondary and is a little worrie some.

Take out the magnetic shunts- else the Voltage regulation will be lousy.

If you can get the phasing to work out, use the two secondary lines that go to the core as the center tap. If not lift those leads from the core  isolate and insulate.

A good idea would be to mount both of them on a non-conductive base.

Take a third one and chisel off the secondary. Now you have a great core for winding filament and operating voltages on, plenty of room for Big wire. (works FB)

They do have a bit of hum to them when they are running. Though if you are running QRO and using a blower you probably won't notice much. Mount them on 3/4 inch plywood and Isolate the base from the sheetmetal using rubber mounts. (probalby a good idea anyway)

Keep your hand in your pocket, and the Jesus stick with in reach. 

[w1vtp]

Mike

I have one of those beasts -- I think it is actually just fine except that the contactor won't pull in any more.  I talked with Timtron some time ago about it.  Seems there is something a bit different about about MW power transformers -- perhaps someone can verify if I'm correct about that (see above post).

But anyway, if you're interested you can have it.  It's a beast of a microwave.

Al

[WU2D]

I have heard of 20 - 30 uF yes uF right across the primary to get them in phase!

Hi Al, I'm pretty much off the air during trout season but look for me soon.

73's Mike WU2D

[Opcom]

My friend Jacob has done alot of work with those in all kinds of lashups including oil baths in his Tesla work. Maybe he can comment on the capacitors, the insulation, and the magnetic shunts.

[w4bfs]

ok Mike ...finally got around to experimenting with the microwave oven power xfmr or mot .... took an average 1kw dead microwave apart .... mot was as described previously with primary of 14 ga and secondary of roughly 26 ga with one end tied to frame ... removed 3 turn magnetron fil winding and drove out 2 sets of magnetic shunts which opened up enough space to add 15 or so turns in series with the primary winding ... used a fluke clamp on ammeter for primary current tests which went as follows: (no load on secondary other than dvm with hv probe)

1. just original primary draws about 3.5 A just to magnetize and gets hot quick (5 min or so)   
2. primary and additional 15 turns in series drops to about 1.8 amps and still gets too hot
3. just the original primary in parallel with 20 uFd oil cap drops the current to .6 A - running much cooler
4. primary and additional 15 turns in series with 15 uFd cap drops current to .57 A and after running 15 min could not detect but the smallest amount of warming. The output voltage was 1740 vac.

additional capacities made very little difference in tests 3 & 4.  The transformer remained silent (no buzz).

I believe it is worth more testing to see how a pair of very similar transformers would perform with power factor correcting primary capacitors and a full wave ct secondary as a hv supply ...73...John

[Opcom]

I'd wonder why the MOT works better with a capacitor across the primary, and how it is different from a "normal" plate transformer that does not require one.

[w4bfs]

Patrick ... in order to save $ on the mot cost, the mfgrs use the least amount of copper possible, imho .... since the application has no standby power requirement as a microwave oven, the primary powers only at full power ... therefore the design concern is for full power transfer .... may be the reason they are so compact for the power level of operation ... any xfmr designers out there ? ...73 ...John

[WA1GFZ]

All the wire can be smaller since there is no worry about temperature rise running a low duty cycle. It may take almost an hour for a transformer to heat soak to full operating temperature so running one for a couple minutes an hour allows you a lot freedom.

[Ed/KB1HYS]

I've got the 2-mot full wave setup running.  I think I could put a cap across the primary and see what kind of difference it makes.  Right now running under a simulated load i can pop the breaker when keying the HV.
So I think the current it's pullling is high.

When I get a chance I'll run the experiments. Anyone recommend a size of cap that would work?
teh two primarys are in parallel.

[w4bfs]

hi Ed ....just a suggestion ... look at parallel thread on BW 5100 plate xfmr testing ... unhook secondary, ballast primary, parallel primary winding with ac rated capacitors and keep adding value until magnetization current will not reduce ... i'm guessing 30 to 40 uFd may be required for 2 xfmrs with primary windings in parallel supplied from 120 V ... if the windings are in series for 240 V op then each may have to be determined one at a time from 120 V supply ... consider adding windings to the primary, if possible, from space vacated by magnetic shunts and magnetron fil windings ... I used some teflon coated 13 ga wire ,,,73 ...John

[Opcom]

thought this would be a decent place to throw in a typical? microwave oven circuit.

[kf6pqt]

Unfortunately for us scroungers, there's typically no 500pf bypass caps in the filament feed. Instead, all the ones I've torn open have filament chokes inside the magnetron can.

73,
Jason kf6pqt

[Opcom]

If the bypass caps are there, they are usually in he wall of the filament "can" in which you are finding chokes. They look like feedthrough terminals. SS of NE sells these 500pF 15KV ones.