Magnetron waveguide orifice

[BurningTantalum]

My domestic microwave oven suffered a failure last week. It is a Panasonic inverter type without motor driven platter or stirrer; 4.35kV (900W)
I do not like meddling with microwaves, but the electronics seemed OK and I suspected the magnetron despite never having found a dead one before.
I managed to find a donor unit at the local dump and installed it which returned the unit to service.
The question is- The hole in the end of the waveguide on the original magnetron was square (see pic) The replacement hole is triangular. Is there a reason for this ?
The magnetron in my oven 'fires' horizontally into the underneath of the cavity 'floor'. The donor magnetron fired downwards into a pressed metal box on the side of the cavity, which had a motorised turntable.
It all seems to work OK but I am just curious.
Regards,  BT

[TerraHertz]

The microwave field doesn't come out of the hole. The entire stub is an antenna, including the ceramic part lower down. The shiny metal cap is just a cover over the inner copper tube, which was used to evacuate the magnetron during manufacture, and has a sharp squished and welded shut end.  The hole on the end has no functional purpose I'm aware of. Perhaps it's for holding the cap while applying it originally?

Also, your oven *does* have a field circulator. It's under the glass floor plate. To access the small synchronous motor you'd have to cut some breakaway tabs on a small panel underneath the oven.

I recently found one like this as a street toss. A Panasonic NN-SF550W. It's interesting because the power supply control circuit *isn't* on a small rider board with a custom propriety control IC.  It's all on the main PCB with a more standard looking IC, so hopefully there's a better chance of being able to hack it for purposes like a regulated HV power supply. Unlike all the previous Panasonic inverter microwave ovens I've found.
(This paragraph edited to correct a memory glitch.)

Is yours like this:

[German_EE]

OK, I have learned something today. I have never investigated microwave ovens in depth but I was always told that the microwave radiation comes out of the hole in the end.

[BurningTantalum]

Thanks for that, TerraHertz.
I was thinking "end of waveguide..." Now I see.
The unit in question is indeed the very same model as yours.
We haven't had it long, and Mrs BT and I were evaluating it still when it died.
I must admit that I hadn't noticed that there was a rotating 'stirrer' or 'director', or whatever it is. The cooling fan is extremely noisy- masking the syncro motor etc.
The unit seemed handy for large square baking dishes, which is why we said 'yes' to it when it was being tossed out by a friend.
I'm still not sure of the failure mode of the magnetron- The filament seems to be a dead short, and I gather that it takes about 10A when operating. I was only able to pass 2A through it (bench PSU limit) and got about 6mA across it, but I realise that it would be cold at that. I didn't attempt to measure the voltage (~3VAC) across the filament on the dead unit or the new one- it was safer to fetch a 'new' mag and PCB from the tip !
Apart from loss of vacuum I cannot see how else the component can 'fail'.
Regards,  BT.
PS German_EE - glad that we both learned something today !

[BurningTantalum]

Incidentally-
The local 'Mung-Bean' greenies at the dump, who are contracted to manage the recycling, try to rip every item to bits to 'recycle' parts. They always smash up the microwaves and rip out only the magnetrons. These have the magnets removed and the remains are then tossed in a heap- for what purpose I cannot determine. I steer well clear of the Beryllium Oxide dust but last week I grabbed a handful of the brass RF gaskets. They seem to work well as tip cleaners for soldering iron tips (stapled in a heap on a bit of wood).
BT

[ChunkyPastaSauce]

I think the hole shape on the antenna cap might be used to identify the cap or assembly during manufacturing. There are only a few magnetron manufactures producing all the magnetron variants; changing the punch shape, rather than setting up an additional marking step, is probably cheaper.  Small dimensional changes with the magnetron can have impacts on the performance since impedance needs to be matched and needs to properly be located in the waveguide.; the punch shape makes it easy to identify. Imagine manufacturing a bunch of parts that look similar but are not....asking for problems

Also, most of the microwave energy is emitted perpendicular to the axis of the cap, little through the tip. And, the microwave wavelength is much larger than the hole diameter, so I doubt it effects the radiating pattern much. If you looks at the same magnetron from various manufactures, they don't seem to care about the rotational orientation of the antenna cap. And there seems to be nothing in patent literature with the antenna cap hole shape.

All speculation.

[TerraHertz]

Apart from loss of vacuum I cannot see how else the component can 'fail'.

I've cut a few old dead magnetrons open, and there's a common failure mode. Refer to pics of the structure here:
  http://www.dutchforce.com/~eforum/index.php?s=f7e54f6f2f8f7dfb5d9375271a2a73f7&showtopic=38412&st=25
  https://upload.wikimedia.org/wikipedia/commons/thumb/b/b0/Magnetron_section_transverse_to_axis.JPG/800px-Magnetron_section_transverse_to_axis.JPG

Those small rings that are brazed to alternate radial vanes... they and the vane ends develop melted spots. Looks to me like electron beam welding, and it probably is.
The rings' purpose is to enforce proper alternate phasing of the resonances in the vanes. With the vane tips and rings damaged I suppose it all goes to hell. It's apparently a failure mode that snowballs.

Incidentally, you can cut a magnetron open much more neatly than in that pic, using a lathe.
Oh, and the ceramic parts are not BeO, they are harmless alumina.

[ChunkyPastaSauce]

Oh, and the ceramic parts are not BeO, they are harmless alumina.

How do you tell? Patent search for ["beryllium" AND "magnetron" cook food oven] shows that beryllium had most definitely been used in consumer microwaves.

Also be careful with the central filament as it is probably made of a tungsten-thorium alloy.... don't want to breath/ingest thorium dust.

[tron9000]

Oh, and the ceramic parts are not BeO, they are harmless alumina.

I wouldn't dismiss that! Treat the whole unit like glass.

[T3sl4co1l]

Actually I doubt they use thoriated tungsten filaments.  Probably molybdenum or tungsten coated with oxide, like most radio sized rectifiers.  Hey, it's cheaper (and, they do have to build these things down to a price).

Tim

[coppice]

Actually I doubt they use thoriated tungsten filaments.  Probably molybdenum or tungsten coated with oxide, like most radio sized rectifiers.  Hey, it's cheaper (and, they do have to build these things down to a price).

Tim

Practically all of them are made in China, and China has massive piles of thorium.

[T3sl4co1l]

Actually I doubt they use thoriated tungsten filaments.  Probably molybdenum or tungsten coated with oxide, like most radio sized rectifiers.  Hey, it's cheaper (and, they do have to build these things down to a price).

Tim

Practically all of them are made in China, and China has massive piles of thorium.

Germany has massive piles of coal, therefore everything from Germany is made of coal...

Tim

[coppice]

Actually I doubt they use thoriated tungsten filaments.  Probably molybdenum or tungsten coated with oxide, like most radio sized rectifiers.  Hey, it's cheaper (and, they do have to build these things down to a price).

Tim

Practically all of them are made in China, and China has massive piles of thorium.

Germany has massive piles of coal, therefore everything from Germany is made of coal...

Tim

Germany values its coal as a fuel. Unless China can crack thorium nuclear power, thorium is mostly a hazardous waste nuisance.

[ChunkyPastaSauce]

Actually I doubt they use thoriated tungsten filaments.  Probably molybdenum or tungsten coated with oxide, like most radio sized rectifiers.  Hey, it's cheaper (and, they do have to build these things down to a price).

Did some googling and it looks like thoriated tungsten is still the primary choice for oven magnetrons filaments. I did find some recent Chinese and Russian patents (last 5 years) for rhenium yttrium tungsten cathodes, but almost all other patent literature specified thoriated tungsten

[Gyro]

Those small rings that are brazed to alternate radial vanes... they and the vane ends develop melted spots. Looks to me like electron beam welding, and it probably is.
The rings' purpose is to enforce proper alternate phasing of the resonances in the vanes. With the vane tips and rings damaged I suppose it all goes to hell. It's apparently a failure mode that snowballs.

Presumably not a coincidence that the melted vanes are adjacent to the power takeoff point. I wonder if this is the sort of damage that results from high reflected power - Operating the oven without anything (water containing) inside?

[TerraHertz]

Those small rings that are brazed to alternate radial vanes... they and the vane ends develop melted spots. Looks to me like electron beam welding, and it probably is.
The rings' purpose is to enforce proper alternate phasing of the resonances in the vanes. With the vane tips and rings damaged I suppose it all goes to hell. It's apparently a failure mode that snowballs.

Presumably not a coincidence that the melted vanes are adjacent to the power takeoff point. I wonder if this is the sort of damage that results from high reflected power - Operating the oven without anything (water containing) inside?

That's exactly what I was wondering too.
*Someday*, when I finish setting up my lab, one of the (many) things I want to do is set up a sacrificial microwave oven with field strength instrumentation (main cavity, and waveguide) and try a few experiments. The 'floating arc' being one.

Actually I doubt they use thoriated tungsten filaments.  Probably molybdenum or tungsten coated with oxide, like most radio sized rectifiers.  Hey, it's cheaper (and, they do have to build these things down to a price).

Did some googling and it looks like thoriated tungsten is still the primary choice for oven magnetrons filaments. I did find some recent Chinese and Russian patents (last 5 years) for rhenium yttrium tungsten cathodes, but almost all other patent literature specified thoriated tungsten

Interesting. Note to self: collect some more old magnetrons and rip out the filaments. Because the decay product particles of thoriated tungsten are easily measurable even with my crappy gear. See pics.
Those are 2% Thoriated tungsten TIG welding electrodes. 10 of them, held right up against a scintillator that can detect alpha particles.
Room background is about 30 CPM (Actually I think there is a tiny hole in the foil and photons are getting in, as the count is higher if the face is illuminated.) From the TIG rods in that configuration there's 400 to 500 CPM. With a sheet of plain photocopy paper interposed, it cuts down to around 40 CPM. Thus: Alpha.

That open magnetron doesn't allow me to get the scintillator close enough to the filament to tell if it's emitting Alphas.