My project is to build my own swr and power meter, with an ESP32. uC has 12-bits ADC under 3.3V (sensitivity should be about 800uV)

A friend of mine (Pierre, F1FDD) had the idea to make the HF coupler with some copper sma rigid coaxial.

Before that, I want to characterize the HF detector. My device under test is like this photo :

hf detector rigid copper coax


simulate this circuit – Schematic created using CircuitLab

I should use 1nF and 1kohms, but I don't think it could explain the issue I have

Then, I made measuremnt of output level at 150MHz (from -10dBm to +10dBm) because I have no way to measure less that 1mV. My results are :

out voltage versus input power

I feel disappointed as the output level is quite low. I hope to be able to go to -30dBm, that would lead to several mV, but I'm in the noise of my oscilloscope.

I expected to be able to measure -30dBm, as it seems possible in usual power meter.

Is my measurment correct ? Are my output voltage result typical value of HF detector with schottky diode ?

Is there a way to increase the output voltage ?

Thanks in advance for your help, 73's Fabrice, F4BJH

  • $\begingroup$ (sensitivity should be about 800uV): I think you come to that conclusion because 3.3 V / 2¹² = 805 µV, right? But you can't calculate that way; 12 bits are not the ENOB of that ADC, and you'd need to incorporate the noise figure of all circuitry between the source and the ADC as well. $\endgroup$ Commented May 9 at 10:56
  • $\begingroup$ ah, and, definitely not an answer to your question, but might be worth considering: ham.stackexchange.com/a/16316/4076 $\endgroup$ Commented May 9 at 12:02

1 Answer 1


Have you tried to bias the diode with a small current? This should bring it into the region of conduction where it can better rectify the RF.

Something like this, where the current is determined by experiment:


simulate this circuit – Schematic created using CircuitLab

For better stability with temperature changes etc, you should have a second diode next to it (in the same package?) with the same bias and load but no RF, connected to another ADC pin, and subtract the two.

Also consider an RF detector IC like the AD8302 which is almost built for RF directional power meters.
It has a clean log response, input range -60 dBm to 0 dBm, two channels, and also a phase detector between them so with just a bit more software you can find the reactance of the load, not just the swr. Demo boards available online for just $10.
The only problem, if it's connected to an antenna, will be keeping stray RF out of the detector. Your LO / transmitter needs to be the strongest signal in the directional coupler, and put a lowpass filter before the detector to remove FM and TV signals.


You must log in to answer this question.

Not the answer you're looking for? Browse other questions tagged .