temp sensors in microwave ovens

[ProfK]

How do the microwave oven manufacturers allow a metal (stainless?) probe to function properly as a temperature sensor? When I experimented by putting thermocouples or metal tubing inside the cabinet through the vent holes, they get quite hot even though I ground them well to the cabinet. There was a time you could buy microwaves with probes covered in a flexible armoured conduit (~1/8 inch OD)to verify meat was cooked. Regardless of if they were thermistors or thermocouples, they should get hot and give false readings or even arc or burn. Or is there a trick to that??
What is the trick?

 

[logbook]

My "academic" answer (having never experimented with this)is that a piece of highly conductive metal inside a microwave cavity should not in principle get hot. The walls don't get particularly hot do they? Metal will distort the field but dissipating power requires a finite resistive component in the impedance.

I guess you could consider this in terms of the impedance of the object compared to the wave impedance of the field; or you might think in terms of the coupling from the field to the object. Metal is reflective rather than absorbing. More conductive metal will therefore give less loss. You see this with screening cans over tuning coils. I believe Copper and Aluminium cans are better than Iron cans for loss.

 

[IRstuff]

But there was arcing in older microwaves, no?

Perhaps it has more to do with the uniformity of the field inside the oven. Current flow requires a gradient, if there's no gradient, there's no current.

TTFN

 

[logbook]

Well since I have an interest in all things electromagnetic I thought it was about time I put my theory to the test. I was a bit worried because it is not my microwave oven. On the other hand it has been playing silly beggars recently so if it blew up it wouldn’t be too great a loss.

Equipment:
Sanyo EMS 153 800W "showerwave" domestic microwave oven
Cup of tap water
stainless steel teaspoon

Apply power for 15 seconds.

On the first test I had the spoon in the corner/side of the oven to minimise the electric field.

The first time I tried this I thought I had warm water and a cool spoon. I just repeated it and I was incautious in picking up the spoon. BIG MISTAKE. The water was pretty cool but the spoon was hot enough to burn (ouch!). If I had touched it sensibly then it might not have been a problem but as I had done it before I "knew" it wasn’t going to be hot.

I just tried it again. The part of the spoon right in the corner seems to get less hot. Also the heat dissipates quite quickly so if you wait a few seconds it isn’t that hot.

This is all straight out of the text books in terms of skin effect. There was a famous incident reported by Lord Kelvin (I think) where somebody picked up an iron rod that had been subjected to an RF burst. They got burned too, but the rod was found to be cool just a few seconds later.

Of course the magnetic field near the walls is at its strongest and that is where a low impedance metal object will get the most heating. I think stainless steel is a relatively poor conductor, hence the heat. There was no arcing.

The microwave has gone back into its "sulking mode". It won’t accept front panel inputs any more. Gives me a chance to soak my fingers under the cold tap some more.

 

[ProfK]

Excellent work. It supports my observations too. But it doesn't answer how one can make a metal sheathed probe to measure temperature inside the cavity in many of the older high-end microwave ovens and still have it register a proper temeprature? Could it be the probe tip was at meat temperature but the sheath could heat up significantly higher in temeprature? If so, when cooking a big hunk of meat this way, couldn't the sheath get hot enough to melt or arc after 30 minutes of heating? Inquiring minds want to know.
BTW, my materials were copper and inconnel. One higher and one low conductivity than stainless. Same result. When I put the TC into the water load, it registered the water temp up to 100 C. The exposed lead/sheath got hot enough to smoke and when removed from the water got hotter faster and in seconds got to 260 C.

 

[Higgler]

A spoon added to a cup of coffee probably makes it heat quicker and will not arc. The metal spoon becomes an antenna with a coffee load resistor.
Temp sensors usually will behave similarly, put it in the meat or water and the current induced on the sensor can't resonate (bounc back and forth) and build up a high field because it has a load on one end and the other end is short circuited to the side wall. You will induce current on it however, especially if it's co-polarized with the power source.
Most Temp sensors probably come out of the walls crosspole to the fields too. The sensor metal may slightly increase the reflected power back to the source location because even a loaded monopole has a radar cross-section whereby energy is reflected.

kch