Microwave oven - bizarre circuit??

[GMB]

Our Thorn 2003 ancient microwave oven just developed a fault indicated only by the fact that the normal buzzing made by the transformer increased in volume very considerably. Luckily it has only been run with the fault for about 3 bursts of 1 second - so no smoke yet.

Given the small number of parts I was hoping for a quick fix, and indeed there is some evidence that the SRK-11Z diode is where the excess current may be going.

*BUT* I am gob-smacked by the circuit!
My problem is that the diode is connected between the 0.88uF 2100VAC capacitor and - - - mains earth!!!

There is no doubt. I noticed this when I saw that the diode connection tag had only got a 0.5mm clearance with a bolt clamping the capacitor to the box. I was just thinking that maybe this is what caused a short when I saw that the green wire from the tag went to the primary mains earth to case connection point anyway so obviously not!

What is going on

[G6Tanuki]

The circuit is a half wave voltage doubler.

The second diode being the magnetron itself.

Which is also the 'load' for the doubler!!

Not intuitive at first sight but once you have thought about it a bit it makes perfect sense.

[GMB]

Yes I understand that, but the power is flowing through the case!!

That is not my understanding of a safety earth!!

[Cobaltblue]

Maybe this generic circuit will help.

It's on the secondary side, there's no real difference to an amplifier chassis with the electronics being earthed on the secondary side of the transformer.

Cheers

Mike T

[G6Tanuki]

The Magnetron anode and one end of the HT transformer are also connected to chassis, it makes sense as it reduces the number of bits that are at high voltage above ground.

[G.Castle]

You wouldn't want it to float.

[McMurdo]

they use an isolating transformer so the cooking energy doesn't flow down through the mains earth

[GMB]

Ah yes, it does seem to be a safety isolated transformer.

I still think it might be better if the power went through actual wires rather that taking its chances with a steel case.

So what do I do about the diode which I am sure will turn out to be the problem? I presume these diodes are all much the same. I can't find an actual spec for it but I am guessing that different ones will be about the termianls rather than the spec.

SRK-11Z is the part.

Any ideas for a modern obtainable one?

[FIXITNOW]

i would just search "microwave high voltage diode rectifier" eBay and see what there was with the correct ends on i wold have through the spec would be much the same depending on oven power rating , The one i did i salvaged a diode from a old rusty but working unit

only data sheet i can find are for 12Z versions
https://www.datasheet-pdf.info/entry/SRK12ZB

[GMB]

I just found that the actual fault is that the capacitor is a dead short. It must have happened quite suddenly. I have to say I am a bit surprised.

The diode looks like it works (only tested to 1000V) but has been stressed I would think.

[GMB]

I can't help but notice that microwave oven capacitors seem to have rather precise values. Any idea what tolerance is acceptable? Clearly this will set the magnetron current but it isn't exactly a stabilised supply!

I am not finding 0.88uF advertised - but close values seem to exist.

[Radio Wrangler]

The output of the magnetron is several hundred watts at 2.5GHz. ordinary coax would not be a good choice! so it's a metal wave guide. Leaks at this power level would be dangerous and invisible, so the outlet port has to go directly to the wall of the cooking chamber. No mucking about with trying to include high voltage insulation for DC or mains frequencies. The clearances would be quite large and a lot would leak. There are ways, but costly.

So the magnetron anode, cooling fins and magnet all have to be connected to the metal structure of the cooker. The high voltage needed to power it has to be negative with respect to the earthed cooker case, and connects to the filament (directly heated for efficiency) So the transformer has to have a well insulated heater winding. The EHT winding gets connected to the earthy world at one end, so that can be at the bottom of the bobbin, so stress is low on the bobbin near the core.

Even the diode gets earthed at one end.

The HV capacitor has to float at HT. If metal clad, the case gets earthed for safety, leaving two high voltage terminals which are dangerous.

As already pointed out, one subtlety is that the magnetron itself acts as a power rectifier diode in a half-wave doubler, so the microwave output is pulsed at 50Hz.

Another subtlety is that the running magnetron is an oscillator, with a lot of positive feedback going on inside it. One side effect is that the impedance it presents to its power supply looks like a negative resistance. This makes control of the power level rather unstable. The quick and dirty solution would be to add enough series resistance to out do the negative resistance. A better solution would be to use some inductance instead. This is easier done than explained! The mains transformer is designed to INCREASE its leakage inductance. The leakage inductance makes the transformer output look more like a constant current supply than a constant voltage one. So that leakage inductance that audio output transformer designers work like the devil to reduce is actually wanted.

The big iron transformers in old-style welders also uses this trick to control the welding current. The big wheel of the current control turns a screw and slowly moves a laminated magnetic shunt which short-circuits some of the magnetic flux in the overall core, letting it bypass the secondary.

The microwave cooker just wants one fixed value, so it's designed into the core/bobbin structure.

Radio amateurs keep looking at microwave transformers as cheap high voltage generators for big valved power amps, but they then find their 'voltage regulation' is poor.

David

[Radio Wrangler]

I took a while over that reply, so now have just seen the last few replies.

The capacitor plays a resonance trick with the transformer flux leakage as seen through the secondary inductance (when the input is connected to low-Z mains). In this way it's a bit like a shunt-L into series C antenna tuning unit matching into a high-Z load and giving some more voltage boost. It" best to stay close to the intended value. The resonance Q is very low, but the leakage inductance of the transformer is wide-ish tolerance and then there is mains frequency where people in power cuts run the things off of generators. You can probably change the C value but you'll offset some careful tolerance calculations.

Those capacitors work hard!

It's remarkable how many subtle tricks there are which don't look obvious on that deceptively simple circuit diagram.

David

[G6Tanuki]

The capacitor and the inductance of the transformer secondary also - by resonating at mains frequency, and the flywheel effect, help reduce the amount of harmonic energy reflected back into the mains, which could otherwise be problematic.

It's principally the third harmonic energy that is the problem, just as with switched mode power supplies.

[Richard]

Hi

A search on ebay for 0.88uF finds a few, both new and used.

Richard

[JonSnell]

If you fit the wrong value capacitor, too high a value that will result in to high a current flow through the magnetron. Too low a value means too little power.
ALWAYS fit the correct part.
The diode will be fine.

Have a look around domestic appliance repair workshops, they may have one.
Unfortunately a replacement oven may be required.

[GMB]

The ones on eBay that exactly match the spec. all seem to be salvaged parts. I missed one due to their typo which turns out to be the exact replacement. But do I really want one 35 years old to replace my dead 39 year old one?
(I skip stupidly expensive parts costing more than a new oven).

So I have ordered one salvaged from a damaged brand-new oven which is 2% off the value (of a 3% part). That has to be close enough I think. There is no way the Q of the transformer is all that high given it also powers the heater.

It may yet come to getting a new one as maybe this has been a warning. The snag here is that our kitchen is rather well adapted to its exact shape and size so it will take a lot of sifting through showrooms to find a replacement where the installation will not involve collateral damage.

I am also aware of the "they don't make them like they used to" effect. Our direct experience of new ones falls into two camps.
1) A cheap simple one a bit like ours (i.e. with simple rotary timer and one button operation) that we bought for our other location developed a dangerous fault one day outside the warranty period.
2) Modern flashy ones we have encountered in rented holiday cottages have the annoying problem that you almost have to write a program to work them. The great feature of the microwave oven is that it does not need any "features" for its primary role. The last modern one we saw wanted you to tell it what exactly it was cooking. Clearly it had lost the plot.

[kalee20]

Originally Posted by G.Castle

You wouldn't want it to float.

Definitely not!

Sod's law would have it floating to something you really didn't want...

[line sync]

Thank youi Radio rangler ( david ) for explaining how the simple transformer type microwave oven circuit works , i didn`t realise the mains transformer was designed like that.

Robin

[Glowing Bits!]

I found an interesting video on YouTube of a guy who pulled apart a magnetron, he explained in moderate detail how it works but I got lost in technical translation by then.
The most fascinating bit was the spiral filament, does anyone know how bright it glows?
Apparently it takes 3 seconds to warm up the heater, another 10 mins to cook the food😁