Microwave sensor RCWL

Anyone got one of these RCWL-0516 devices (or similar) to work? I ordered a version with a 3-pin header that produces a 3.3 volt positive step output when motion is sensed, but I recently two 5-pin (identical PCB) that are dead, and another version that is encapsulated in a small metal box that outputs very small signals (30-200 negative millivolts) that are useless as a logic input. The failing 5-pin unit uses the RCWL 9206 chip.

The 3-pin version (vendor long forgotten) works great in the lab, detecting me walking around in a 12' space, but provides totally random outputs when installed outdoors. This version uses the RCWL 9196 chip.

Both the 3-pin and 5-pin chips appear identical in all respects except for the chip number.

provides totally random outputs when installed outdoors.

Those sensors don't work out of doors. The detection principle seems to be interference between various reflected waves (including from the moving object). See https://www.rogerclark.net/investigating-a-rcwl-9196-rcwl-0516-radar-motion-detector-modules/

In the link provided above, I do see the discussion of "reflections of reflections," so the poor performance outside confirms that. However, in the steady state of "outside," it seems to me that a car passing within 8 ft or so of this device would create a strong reflected wave. It may be that the detector doesn't work until the local environment is stable. The sudden big reflection might be interpreted similar to the long turn-on delay, waiting to stabilize.

More experimenting needed.

The thing works great indoors - don't breath too hard or you may set them off, that kind of sensitive.

I haven't tried them outdoors myself but it has been reported here before that they don't work (well) outdoors. Some walls and other objects nearby may be required for them to do their thing. Plants probably do not qualify, foliage is known to absorb radio waves, but for the same reason I don't expect plants to be a problem either.

I need an alarm on the driveway. You'd think these little stinkers could do that. I can't use PIR because they get blinded by the sun in the evening. Magnetic detectors are not as straightforward as you might think, but I guess that's where I'm headed.

I use a magnetometer to detect vehicles in the driveway, and it works fine. Detection range is a few meters, obviously depending on the size of the vehicle.

The program maintains a running vector average of the local magnetic field, and detects large, short term departures from it.

I am giving myself a little dope slap. I never imagined that a magnetometer was affordable. It turns out they are under $5. Maybe one of those 3 axis chips will work.

I was thinking of doing the loop detection method, like the stop light detectors which, if you haven't looked, even the surplus ones are $100-ish and up.

Thanks tons.

The LSM303DLM magnetometer breakout from Pololu works well in my application.

Pretty much all smartphone have magnetometers in them.