Jeff Liebermann wrote:
On Mon, 10 Aug 2020 01:44:30 -0000 (UTC), bob prohaska
wrote:

This may be bending the OP's intent a little, but (many) years ago I
was led to believe that it was possible to detect semiconductor devices
by "illuminating" them with RF and looking for second harmonic scatter.
It's merely a "there or not there" test, supposedly used to test secure
spaces for for eavesdropping devices. No firsthand experience, alas.

Have I been misled?

It's quite real. The principle is still being used in some retail
theft prevention tags. The tag has a 900 MHz antenna, diode, and 1800
MHz antenna. When the customer leaves the store, the tag is
illuminated by 900 MHz. If the tag is intact, the diode produces a
2nd harmonic, which is re-radiated by the 1800 MHz antenna. A
receiver near the 900 MHz transmitter listens for 1800 MHz and sounds
an alarm. When the customer pays for merchandise, the checker waves
the tag over some kind of induction device, that fries the diode so
that it doesn't trigger the alarm as the customer leaves the store.

You could do something like that to detect cell phones. The problem
is that if the sensitivity of the receiver was sufficient for such a
system to work with most any semiconductor, it would also false
trigger anything with electronics inside. It has no way to
distinguish between a smartphone and car alarm key fob. It might also
trigger on harmonic generated by diodes created by dissimilar metals
and corrosion. Such false triggering would probably make it useless.

You mentioned 13.56 MHz being used by cellphones. Combined with the
2.4GHz (or maybe 5, on a modern cellphone) would it not be possible
to illuminate with both frequencies and then look for harmonics? That
still isn't perfect, but it'd help exclude some false positives.


It wouldn't take much to shield a wireless bug from being detected by
such a system. The best place to hid a bug is inside something
stuffed full of electronics, such as a phone, TV, monitor, keyboard,
mouse, etc. If you "sweep" a room looking for diodes, it would be
triggered by all these devices, which would then need to be
individually disassembled and inspected for tampering.

Drivel: I wonder how such a system would respond to an OLED display
in a smartphone, which has at least one diode per pixel.



Wouldn't that sort of setup have a relatively low Q with fairly uniform
response over a wide frequency range? Perhaps I'm suggesting not looking
for junctions specifically, but for resonant circuits connected to antennas
that must be exposed for the device to function. Obviously no help if the
phone is under a tinfoil hat, 8-).

One question is power levels; if the interrogation signal starts melting
chocolate bars to get a recognizable return it's likely a bad idea....

Thanks for replying!

bob prohaska