Video: build a simple RF demodulator / detector probe for your DMM/scope

[w2aew]

This video describes a simple RF demodulator / detector probe that you can use with your DMM or oscilloscope to measure the relative amplitude of an RF signal in your circuit.  It is very useful for tuning or adjusting IF stages during alignment, filter responses, peaking RF circuits, and examining amplitude modulation characteristics.  The video describes the circuit and its operation, and demonstrates how to use it and the typical results you'll achieve.  It uses a pair of 1N34A Germanium diodes, but could also be built with other low capacitance, low bias diodes.

[Alexei.Polkhanov]

Great video.

I have attempted to build RF power detector using schottky diodes, but it did not work very well for frequencies that I need - 2.2 to 2.5 GHz. I used HSMS-286 diodes from Avago. I kept the connections very short (under 1 mm) so not sure what was wrong.

At the end I made this little rig with AD8313 - works very well. I would even say too well - I cannot measure anything under -40dBm because it pick up everything, my Wifi, cell phones, somebody using garage door opener downstairs - it shows up on output  Obviously I need good RF enclosure. Here is OOK demodulation from my own garage remote on a scope. I kept the remote about 2 meters from small antenna.

[nctnico]

You need a proper PCB with SMT components. The way this circuit is built is asking for problems.

[Radio Tech]

Nice video Alan. Thanks
Been using one like this for a few years mounted in an ink pen. Really helps for trouble shooting  receivers and transmit circuits

[xwarp]

Alan,

Here is a schematic for a probe that is used with the Sencore SG-165 Stereo Analyzer.

Thoughts?

[Mr. Coffee]

Great vid!

You need a proper PCB with SMT components. The way this circuit is built is asking for problems.

You do not need SMT components. RF and Demodulator probes with essentially flat response exceeding 250MHz have been built with crystal rectifiers, such as the 1N34A or similar, for generations. The leads do need to be short, and point to point wiring actually adds less reactance than most PCBs.


There are two reasons I see for this demodulator probe's response to read low and fall off further as it approaches 50 MHz:

First of all, peak and peak to peak indicating, topologies never reach full charge on the capacitor(s) in the probe due to diode reverse leakage, variable forward resistance, and other reasons, and so a peak reading probe with and output voltage divider to cut the output to an RMS indication will give a much truer 1:1 reading from about 200 mV RMS on up, typically +/- 5% to 10%. Below these levels they can still be accurate, (even down to a few mV) but a correction chart or graph must be made due to non-linearity.

The flatness of response and bandwidth will likely be improved dramatically with an isolation resistor (which is automatically taken care of by the addition of the above divider). This probe's design allows the capacitance of the leads or cable to be seen by the RF circuit to the extent that the capacitors are not charged fully to the full peak voltages. Since there is a difference in rectified voltage and true RF peak, RF current will flow through this capacitance and partially shunt the source. The higher the frequency, the more pronounced it becomes.   

This is all from personal experience building and using many RF voltage and/or demodulator probes.

HTH
Rob

[Howardlong]

You need a proper PCB with SMT components. The way this circuit is built is asking for problems.

For 2.2-2.5GHz I agree. There are an awful lot of antennas in that picture.

I would also recommend a bandpass filter on the front end too to band limit what's going into the log amp. If you can narrow it down a bit further, to 100MHz or so, you could use an off the shelf saw filter.

Also forget BNC at 2.4GHz, use SMA. I use edge mounted SMAs very extensively for these sorts of prototypes, worth having a bag of them handy, you can get them on eBay reasonably cheaply.

Edit: Here's a circuit I made up about 18 months ago with an AD8313, for 1.09GHz, with two cascaded SAW filters and a preamp (this is not my circuit design, but the board and schematic capture are mine, attached). The second saw filter and preamp may be unnecessary for your application, this design is for over-the-air secondary radar reception. As you can see, the intra-component spacing for the RF path is about 1/10th", sometimes smaller, basically as short as I can get it with a reasonable DRC check. Bottom of the board is solid copper ground plane, and it's 1.6mm FR4. To be picky, I'd use a thinner dielectric to make an effort to maintain impedance, but it works well enough as it is. The diode on the picture but not on the schematic is to save me from myself, a rev polarity diode. For 2.4GHz, I'd be going to 0402 parts too, these were 0603.

[G0HZU]

A basic diode based RF detector probe like this is an essential thing to have for anyone doing RF work on a budget. I've built plenty of them over the years, usually from parts salvaged from old CB radios but they usually fall apart over time and become unreliable.

The most rugged and reliable one I ever built is seen below. This uses a dual Schottky diode in SMD format and the probe is built on a thin strip of PCB material and is built as a hand held 'pen'. Note the way the wires to the DMM travel up the body of the pen via the rows of holes. This keeps the DMM wires tidy and out of the way.

The probe has short signal and ground wires at the tip and you can see it also has a black scope probe ground wire and croc clip as an option. If it is used with the short wires at the tip then it is extremely flat across 100kHz to 250MHz and is flat to about 1dB up to 500MHz.

With a slightly better layout and shorter tip wires it could probably work well up to 1GHz but in reality this probe is at its most useful when probing RF oscillators and LO signal paths from LF through to about 100MHz. However, it spends nearly all its life packed away in a drawer because I usually use something else like a scope or a spectrum analyser these days.

The pen style construction makes it really easy and comfy to use.

edit:
In case anybody wanted to copy my pen style probe I've now shown it with the isolated and unused copper areas 'removed' from the end of the handle area. This means there are now no metal areas for the user's fingers to touch and the 'pen' part of the probe is just bare fibreglass.

[Mr. Coffee]

For 2.2-2.5GHz I agree.

I agree, too.
I was mistakenly of the belief that the SMD recommendation was intended for the original poster using it at 50MHz, not for the later poster trying for 2.5GHz. My apologies.

Rob

[Howardlong]

And Alan, before I forget, thanks again for another excellent video. Educational, to the point, and practical.

G6LVB

[mazurov]

Edit: Here's a circuit I made up about 18 months ago with an AD8313, for 1.09GHz, with two cascaded SAW filters and a preamp (this is not my circuit design, but the board and schematic capture

Where did you get the 1090 MHz filters? I need one for ADS-B but can't find them for sale in usual places, like DigiKey/Mouser.

[G0HZU]

You could try Golledge in the UK for a 1.09GHz SAW filter

http://www.golledge.co.uk/pdf/products/specs/ma04282.pdf

[xwarp]

Here is my home made probe using scrap parts. Works pretty well. Need to clean it up though. Built from an Eico design.

[Howardlong]

You could try Golledge in the UK for a 1.09GHz SAW filter

http://www.golledge.co.uk/pdf/products/specs/ma04282.pdf

That was where I got them, ISTR they sent me them FoC, but I do spend quite a bit with Golledge on production quantities of other parts so YMMV.