DIY thermal/calorimeter RF power sensor?

[rf-fil]

I'm attempting to source / build my own RF power meter based on heating of an RF load. I want this as a sort of a transfer standard so I can use it to calibrate other gear like log detectors, spectrum analysers, etc. The idea is that, if done properly, this type of sensor can be calibrated with DC voltage, and will be very broadband, well into GHz range. This type of gear from Keysight / R&S is orders of magnitude above my budget, LOL. And I wonder if anyone has built something like that, so that I don't end up reinventing the wheel? I don't need a huge dynamic range or a short time constant. I just need RF power measurement that is traceable to DC voltage. If there's not much out there, I might try to build something and document it here.

[Conrad Hoffman]

It's long out of production but I wonder if there are any LT1088 chips still around? https://www.analog.com/media/en/technical-documentation/application-notes/an22.pdf

[rhb]

I've only skimmed the app note, but Jim makes clear that  thermal resistance is a significant issue.  The LT1088 also doesn't work at GHz frequencies.

Happily this can be measured very easily by recording temperature vs time both heating and cooling.  It's not quick, but it is accurate.  The key to making it work is to choose a geometry for the thermal mass for which an analytic solution of the heat equation exists.  A sphere is the most tractable as other shapes have complex edge effects at the ends which is difficult to quantify.

A good starting point might be a pin pong ball full of distilled water with an RTD and heater inside embedded in insulation.

Another option is a pair of matching thermal masses and a bridge circuit to compare the RF and DC heating and a circuit to adjust the DC to maintain the same temperature.

While not an easy problem, I do think that good results can be achieved and there is considerable interest here and on amateurradiobuilders@groups.io.

I was doing a lot of numerical solutions of the 1D heat equation 5-6 years ago to determine the correct coefficients with very good results.

Have Fun!
Reg

[maxwell3e10]

If you don't need operation at low power, you could start with a bidirectional 40 dB attenuator. Open it up and attach a tiny thermistor or thermocouple to the ceramic substrate of the resistive elements. Then you can feed DC from one end and RF from the other end (through a DC-blocking capacitor) and monitor the temperature to see the effects.

[daisizhou]

Still sold in China, approximately 140 RMB (including domestic transportation)

[daisizhou]

You can choose either AD834 or AD834 module, which is a multiplier circuit