Hi all,
These days we live in a wireless world buried in a tall stack of protocols
, when everything works - all good, videos stream, MQTT messages are delivered, etc. - we are happy then. When something is not working - it becomes a daunting task of finding at which protocol level things went awry. Personally, I find myself missing an easy and repeatable method on my bench to detect whether a DUT is actually transmitting anything - in absense of such capability I might be chasing interprotocol issue that doesn't exist because it's the TX PHY side that is not working. So, back to basics I went and decided to make a handheld tool that would allow me to detect RF transmissions in a broad spectrum (below 6 GHz) in a near field, i.e. when close to the DUT. I used BAT15-04W diode stack (0.3pF, up to 12GHz) as a front end, configured as a voltage doubler. It then feeds the MCP6001 opamp via couple of 100K resistors (to at least protect the opamp - diodes are potential victims of high RF field or ESD with their Vrbr=4V); the opamp drives an LED and 1230 piezo for audible response (suprisingly useful!). It's powered with 3V CR2032 battery. The tool is mounted in a small 3D printed case, with SMA (not the consumer RP-SMA) at the front end to which I have a "dual band 2.4 & 5GHz" antenna attached. Other antennas can be used as well, like 433MHz if you are optimizing it for lower-end UHF transmissions, like from keyfobs etc. I'm including schematics and a few photos here.
If repeating this construction - note that front end must be air point-to-point mounted, including the IN+ of the opamp being bent up and soldered directly to, not via PCB adapter. See the close up photos.
Also, choice of 2.2pF is not random - it is primarily driven by self-resonance (due to parasitic inductance) frequency of the capacitor, which must be higher than a max frequency we're trying to rectify. So a 2.2pF C0G in 0603 format fits that bill (Kyocera 600S series 0603 capacitors datasheet has useful chart on page 4). I didn't use these specific Kyocera caps, mine are just a run of the mill 0603 C0G 2.2pF, but their inductance is primarily driven by their size, so while Kyocera's 600S Q factor will be better (irrelevant in this use) - their inductance and therefore self-resonance are about the same for all C0G 0603's.
Also, cannot go too low with these nominals - the BAT15-04W own capacitance is 0.3pF, with small signals Cdiode and Cexternal form a voltage divider (reactance of Cdiode : reactance of Cexternal). When Cexternal is too small - its reactance is high and most of the signal from antenna will be dropping across Cexternal, leaving very little voltage for the diode to become forward biased, so no rectification then (until signal becomes large that is = lower sensitivity). And finally too large of a capacitor (aside from the dropping self-resonance frequency) will increase the decay time from the current 30μS (assuming 2 × 2.2pF in series = 1.1pF, plus parasitics = 3pF, τ = 10 MΩ × 3pF = 30μS). The 30μS is comparable with gaps between WiFi packets, which allows the audible signal to have "texture" - you can hear scratchy sound of train of packets being transmitted vs. just a macro level of "packet bursts" you would be hearing with higher capacitance.
It works well, detecting a WiFi AP from ~1m first with sound and from ~30cm with LED flashes. A cellphone on 5GHz WiFi - within 10-20cm (30-40cm when on LTE/5G), same for Zigbee devices. Note that it's broadband, so it "hears everything" (that antenna can receive), so (1) be aware of it and don't confuse legitimate RF TX with an SMPS scream
and (2) be careful with devices like card scanners or especially wireless chargers - they can kill the front end diodes, even if frequency is way off what antenna is for - capacitive coupling at short distances is significant. If you are HAM radio operator - don't bring this tool close to your TX output/antenna
, if a neon bulb can light up there - the 4V max diodes will also... light up... once
.
Happy RF hunting!
Home
Help
Search
Login
Register
