Microwave oven repair

[jrmymllr]

Ok, not the usual "interesting" repair item but it's an important part of my kitchen! I hate throwing things away so I keep fixing them as long as it makes financial sense.

I have a Goldstar MV1610WW (if that helps at all) basic microwave from 2007. Appears to operate normally but not heating and not drawing nearly enough watts to do much. Last year the HV capacitor shorted out. I replaced that and it's worked until now.

I've already performed lots of troubleshooting and read through the old but excellent repairfaq.org.

-capacitor tests normal as near 1uF. on resistance, starts as low and approaches the 10M internal bleeder resistance
-diode tests normal by applying ~15V through a resistor. 8.8V drop across diode one way, full voltage the other way
-mag terminals test as a fraction of an ohm across both. infinite resistance from chassis to either mag terminal
-transformer mag connector tests as sub-1 ohm across both terminals
-transformer HV wire to chassis tests as 86 ohms
-microwave uses ~100W with transformer primary disconnected and microwave "on", ~185W with primary connected and HV wire and mag connector disconnected, ~215W by adding mag connector, ~230W by adding HV wire to capacitor (so clearly each of these things are drawing power)
-no strange smells or sounds
-the transformer, diode, cap and mag all directly connect to each other. no fuses or thermal cutouts involved

I was hoping for something obvious but it's just not there. Any thoughts? Is it likely that the mag is bad but doesn't test bad? And before anyone says it, yes I understand these things generate very high voltages. I discharge the capacitor each time and am not making internal measurements while it's running...although I really wish I had a HV probe to check that.....

[Alex_Baker]

can you measure the resistance of the primary? The MOT that I have measures about 1.2-1.5 ohms. Check that there are not any thermal fuses anywhere on or in the transformer as well.

[jrmymllr]

can you measure the resistance of the primary? The MOT that I have measures about 1.2-1.5 ohms. Check that there are not any thermal fuses anywhere on or in the transformer as well.

Primary resistance is 0.19 ohms. But there's a big difference in overall power consumption between having the primary connected and disconnected, so it's doing something.

Also interesting is the power consumption goes up by 30W between having the mag filament connected vs. not connected (with the transformer HV disconnected), and as I understand the filament is typically 3.3V @ 10A...so that makes sense.

[BrokenYugo]

Sudden failure or gradual?

[jrmymllr]

Sudden failure or gradual?

Good question. It was fairly sudden, but didn't seem instant. This morning I successfully heated something up that usually takes several minutes, although towards the end it wasn't boiling and bubbling like usual. Then my wife used it and noticed the item was not as hot as it should have been. Also during this time I noticed that the usual constant transformer hum noise instead sort of pulsed every 0.5 to 1 seconds. After that it essentially didn't heat at all.

I've noticed for months that when the microwave hasn't been used for awhile, when cold, it cooked with quite a bit more power than when it got warmed up. I confirmed this with my Kill A Watt. Cold it was using over 1600W, then dropped a few hundred watts once the (magnetron maybe?) warmed up. I don't know if this is normal or if it always did that.

Edit: you got me curious. I don't recall exactly how much the power consumption would drop from cold to warmed up, but it was easily enough to notice in cooking times and I want to say was maybe a couple hundred watts. So I put the power meter on my back-up microwave. Ran it for 3:30 heating water. Started at about 1240W. Near the end, 1190 or 1200W. So not as much.

Bad mag? I'm wondering.

[bsfeechannel]

According to the schematic, there are two points of attention you must check out. The relay contacts and the secondary switch contacts. When everything is working, it is a good idea to check the voltage at the primary of the MOT.

[jrmymllr]

According to the schematic, there are two points of attention you must check out. The relay contacts and the secondary switch contacts. When everything is working, it is a good idea to check the voltage at the primary of the MOT.

That checks out. 118VAC on the primary of the MOT when microwave is supposed to be on.

[themadhippy]

Also during this time I noticed that the usual constant transformer hum noise instead sort of pulsed every 0.5 to 1 seconds

has it got a low power or defrost mode?

[bsfeechannel]

That checks out. 118VAC on the primary of the MOT when microwave is supposed to be on.

In that case, if you're not having an intermittent problem, and if you have tested everything else and it is OK, then perhaps it's time to investigate the magnetron.

[amyk]

Gassy magnetron?

[jrmymllr]

Also during this time I noticed that the usual constant transformer hum noise instead sort of pulsed every 0.5 to 1 seconds

has it got a low power or defrost mode?

It doesn't. Just a simple microwave generator circuit with a single relay to turn it on.

[jrmymllr]

More information.

I really wanted to know if the capacitor is building voltage but without a HV probe I wasn't going to touch that. It has a built-in 10M bleeder, so I took advantage of that...I tested it by charging to 5V and observed the discharge rate, did some RC calculations, then ran the microwave, unplugged it, waited 20s then put my meter on it. Back calculating, the capacitor must have roughly 1500-2000V the instant I unplugged the power cord.

I also put an inductive pickup on the filament wire and measured almost 10A. It has to be the magnetron! It bothers me however that it doesn't test bad, but I suppose there's multiple failure modes with these things.

[BrokenYugo]

Magnetrons are magic as far as I'm concerned, but I know it is a thermionic vacuum tube of sorts, so the emissive coating on the cathode must wear out eventually.

Now I wonder if one can get a little more out of a weak one with the usual tube trick of getting the heater/cathode hotter. Temporarily (with a bench supply and no HV to just cook the cathode for a while), or permanently (add a turn to the filament winding), likely an ill advised experiment.

[floobydust]

OP I think you've done all the tests, it looks like the magnetron has very low (thermionic) emission.
Usually failures involve short-circuits and something burning up, which makes it easier to figure out.
You really need a HV probe to fully test the power, but you've checked all the components- the diode, transformer and capacitor.
I would pull the magnetron and see if there is anything visual and to get the full part number.

If it wasn't making microwaves, due to say broken magnet or melted antenna, I would expect the tube to still consume full power. After all, it is a big diode. If air got in, filament current goes high and I'm not sure if they arc inside. http://www.magsells.co.uk/magnetron-can-fail-number-ways

Goldstar 56324A magnetron is made by LG 6324W1A001L is a 2M246-050GF. Amazon guy said they are shipping 2M246-01TAG. I could not find datasheets explaining the many many suffixes for the mounting plate, air flow and connectors etc. Toshiba 2M248J is listed as an alternate part.

[james_s]

Well you've got voltage across the primary of the HV transformer so you know all of the control circuitry and interlocks are fine, all that's left is the transformer (unlikely), capacitor, diode, wiring and magnetron. Personally this sounds to me like a faulty magnetron, they are vacuum tubes and they do wear out. They're usually not particularly expensive and it's even possible in many cases to sub one that is similar so at this point I'd try replacing it. Look on ebay and see if you can find a suitable replacement.

[jrmymllr]

I'm buying a new magnetron and crossing my fingers. I'll report back once I install it.

[jrmymllr]

New magnetron fixed it. Power consumption is also much more consistent now compared with how it used to cook more vigorously until the parts warmed up.

For what it's worth, I found Martin Microwave (in US) had one of the better shipped prices and guaranteed the exact replacement part.

[james_s]

Maybe worthwhile to conduct an autopsy on the dead one to see if you can see anything obviously wrong inside.

[Gyro]

Good idea, they're good fun to pull apart and the magnets are useful... If intact!

https://www.eevblog.com/forum/projects/will-a-magnetron-draw-480-watts-without-radiating-energy/msg3210292/#msg3210292

[jrmymllr]

Maybe worthwhile to conduct an autopsy on the dead one to see if you can see anything obviously wrong inside.

Oh I plan to do that. It would be a shame to toss it without getting something out of it. And the magnets as Gyro mentioned.

[james_s]

The magnets they use are pretty nice, I have 3 or 4 of them on my refrigerator from some magnetrons I tore apart years ago. I painted them with several coats of epoxy enamel which makes them less prone to chipping, otherwise they are ceramic magnets and are quite brittle.

[eblc1388]

Take great care not to cut or smash the ceramic parts on a magnetron as they contain beryllium oxide, which is highly toxic.

[Gyro]

Take great care not to cut or smash the ceramic parts on a magnetron as they contain beryllium oxide, which is highly toxic.

This comes up every time magnetrons are mentioned. The ceramic parts are Alumina, not Beryllium oxide. Any device containing Beryllium would have safety labels, but it was never used in consumer magnetrons.


EDIT: Whilst on the subject, it's probably worth noting the best way of opening up a magnetron...

- First pull off the sheet metal antenna cap from the top of the magnetron, this is easily done with a pair of pliers.

- Underneath you will find the copper pinch vacuum seal. Saw through the base of this. This both releases the vacuum (very small hiss) and also detaches the thick copper anode connection that comes up from the main body.

- Now the hard part. saw through the Copper anode cylinder itself. There is normally a small groove where the copper anode meets the steel end piece which holds the top insulator. You can use this as a saw guide to minimise the amount of anode cylinder that you have to cut through. Hold the anode cylinder (not the bottom insulator) gently in a vise (it distorts fairly easily) rotating slowly as the saw just breaks through. Don't try to cut straight across the whole thing as you will damage the internal structure.

- Once free, lift to top portion slowly, unthreading the (aforementioned) anode connection that goes up to the pinch seal.

- Marvel at the beautiful un-oxidised copper internals while they are still fresh and look for the failure.

[chris_leyson]

Thanks Gyro for your comments on magnetrons and  Beryllium oxide. I think most people who say the anode insulator for the coupling electrode is Beryllium oxide are just repeating what people have posted on the internet and haven't bothered to read the data sheet. Some magnetrons have a pink insulator for some reason and maybe the pink ceramic has better RF properties.

Also, if the insulator was Beryllium oxide there would be a warning on the data sheet.

[SeanB]

Pink is merely they added impure aluminia to the oxide sintered to create the seal, as it will turn to a glass internally, making for a better seal.

Common for a magnetron in use for 15 years to go low emission off the filament, and to evaporate off tungsten as well, coating the internal copper anode with it, making a higher resistance thin film, and thus make it not want to start oscillating reliably, along with the magnets becoming weaker with hundreds of thermal cycles on them. So it draws current, has heater current about right, but it just is a very high power dissipation diode, as the electrons do not swirl around and excite the cavities well, or one or more of the internal straps, there to join the alternate cavities to reinforce the correct oscillation mode, have cracked loose, and thus the magnetron oscillated way off frequency, and thus the cavities do not resonate to build up the power to couple to the output. I tend to keep old good ones, as they often are useful as donors, seeing as typical failure on the cheap ones is the actual metalwork rusting through, or the CCA wire wound transformer going short between turns. Last few magnetrons that failed, blowing fuses, the RF feedthrough to the filament and negative supply had broken down internally.