The problem is design error in the Moisture Analyzer.
Solution: Add an amplitude sensor and or calibrate the
MA reading or control the magnetrons amplitude
in closed circuit.
More ?

<nbucska@pcperipherals DOT com> subj: eng-tips

The most basic set-up would consist of, magnitron--> directional coupler, primary path into expected load, or load represenative of actual, or load of variable vswr (load pull).

On the forward path coupled port, attenuators (as required) and transition to spectum analyzer. The SA, will tell you every thing importantant, ie center frequency, spectral bandwidth, and power output.

I was hoping to stay away from a spectrum analyzer, I saw one once and they said it was a $100,000 set up so unless I could get one from a salvager for $1000 I would be out of luck. I am wondering if the above mentioned amplitude sensor could be made to work as a sensor in the chamber. The chamber used in this analyzer is cylindrical it is an offshoot of some kind of dryer system. Rubbright group markets the technology. Are there any books I could look for about Spectrum analyzers and the above mentioned setup? Thank you kindly for taking time to answer-David

A used general-purpose spectrum analyzer with guarentee will run you about $5000 up. But, as GOTWW points out, you can look into issues other than just power output.

Otherwise, you can just substitute a power meter (and appropiate sensor, attenuator). If you keep this to just a simple detector, It'll cost under $500 with extras. A used power meter/sensor will run you about $1000 to $2500 but will be more accurate over a larger range.

Comcokid-Could you steer me in the direction of those sensors, meters and attenuator. What would I be able to do with just a simple detector. As you can tell I know little more than basic repair on microwaves, my understadning of them is very basic at this time but am learning more. I could post a picture of the chamber on Monday on Walmart photo page because it is unique. I have searched on the web but am unsure of the correct terms so I am not finding what I am looking for. I can't believe all the good comments on this it is fascinating.-David

I think the other respondents are going a bit over the top for this application, in fact way over the top. Why is anything more than a loop of wire, a diode, a capacitor, and a hand held DMM needed? We have a massive field in the waveguide. This needs to be dumped into a suitable load like a microwave oven with a bowl of cold water in it.

I believe people (radio amateurs) make simple diode detectors like this for picking up field strength outside microwave ovens, so having direct access to the field should be no trouble. I am only talking about a loop 1mm across.

I can’t suggest what  type of diode, what size of capacitor, etc as I have never tried anything like this. However I believe this approach is far more sensible than using directional couplers and spectrum analysers!

I seem to recall people testing for fields in radar waveguides by sticking ordinary miniature light bulbs into the field on the end of a plastic rod. Perhaps somebody with experience of this could comment.

This is a good approach, but why waste your time building a detector when they a readibly available on the open market. Plus a certain amount of science experimentation is required, over many samples, to get a feel for the correct answer. In the microwave production enviroment, standard procedures are followed using calibrated, traceable test equipment. Otherwise, try defending your unknown quantity diode detector, when your boss is sued or a whole production run is returned for recall. The proper test equipment is a pitance in comparisom.  Sad, but this is typical nowdays, get a mechanical to test the tubes, he's had all the normal kids he needs anyway.

SvcTech - You originally indicated you wanted a simple detector. Logbook's idea is probably closer to what you want - at least you could experiment with something like that before investing more.

disclamer - I do not work with RF power levels above 1/2 watt, and consider myself waveguide challenged. However, I will take a stab at it, and hope that you others just add in where I'm incorrect or may overlook something. SvcTech - you are dealing with hazardous microwave power levels.

Take logbook's idea of a 1mm loop pickup. Solder this on the end of a SMA bulkhead-to-solder cup or similar small connector. Drill a hole in a junk microwave oven, clean off the paint around the hole and install the pickup so the SMA connector makes good contact. Use an attenuator on the SMA to a detector, and run the output of the detector into a meter or oscilloscope. Use a bowl of water in the microwave as a load.

I know when heating in a microwave and I forget and leave a twist-tie on the package, it quickly goes up in smoke. It's possible your pickup loop will arc to the oven wall. The first attenuator you go into needs to both reduce the signal level and be able to dissipate the power of the signal it attenuates.

Now, I don't know how much attenuation you will need. This is affected by the standing-wave pattern in the oven, and the effectiveness of the loop. The idea of the attenuator is to keep RF levels down and to protect the detector from excess voltage.

You could build your own attenuator and detector. An attenuator is only three resistors in (usually) a PI arrangement. A high frequency schottky detector diode along with a capacitor will make a detector. The parts, and even attenuator chips can be obtained from places like Digikey. You will need to use surface mount components soldered next to each other, and your final circuit will need to be very well shielded. Good RF layout experience helps in even breadboarding this type of circuit.

Regular SMA type attenuators and detectors can be bought at places like Pasternack. Typical SMA-type attenuators are rated for 2 Watts. Attenuators/detectors can even be found at on-line auctions places (i.e. eBay) or at hamfest fleamarkets for a lot less.

Why mess around with a loop?, when a straight e-field probe would do, and even a mechanical could tune it with a pair of dikes. I just evision a mechanical engineer in a court of law discribing loops, or even dipole antennas, shielding, EMI effects of human organs, against the expert PE witness in a criminal product liability suit. I want to be out of town, not my house, not my car, and the dog too GEEZ!

I can't belive the response this is getting! Thanks Actually the machine uses a loop detector in it with that diode capacitor setup mentioned above. I am going to post some pictures of the chamber and sensor so you can see it if you are interested.-David

As an electronic engineering type (and HAM) myself, I'm naturally tempted to propose a technical  solution, but I think think that in view of the posters requirements, the KISS principle might be applicable here – IF the number of magnetrons to be tested is small, and high accuracy isn't required - or you just need a quick test.
He says:
“I need to mainly compare one magnetron to another to weed out the weaker ones”

 One very easy, low-tech way is to :
a.  measure the starting temperature of a given amount of water and heat it for X seconds, using a known good magnetron. Record the temperature gain.
b. do the same with a magnetron that you wish to test and compare the temperature rise, for the same amount of time.
 Of course it's only a relative measurement, but it may be enough for certain applications where great accuracy is not important. Again, it depends on factors regarding the need for accuracy or legal concerns, etc.
 Anyway, at worst it could be useful as a quick interim test, until a better system can be established..... and it's as low-tech as you can get.
Requirements:
1 microwave safe container
H2O
Thermometer
Pencil & paper
ability to do simple math  ;O)
Cheers,
tenna


 

I can go one even simpler.  Take an ice cube, put it in the oven, turn on the magnatron, and record the time to melt it.  Power is inversely proportional to time.

If you want something more accurate, come up with some sort of flange arangement, some rectangular waveguide, and a high power waveguide load from ebay or some other surplus source.  Hook a thermocouple temperature meter to the load, and measure temperature at some given time after turn on.  The only drawback is that you have to let the load completely cool down between tests.

If you want a high volume very accurate test, get a rectangular waveguide, a high power load to dissipate the power, a 30 or 40 dB directional coupler, a 10 or 20 dB attenuator for the coupled arm, and a standard waveguide power meter.  You can make repeatable tests and there is no cool down time.  You could buy all this stuff on ebay for $1500 max.

If you do not understand what I am talking about, hire a qualified microwave consultant to help you get the test setup running, and to write some safety rules for you.  High power microwaves are dangerous.  There could be a lot of leakage where the magnetron meets the waveguide if you do not design the fixture properly.