ADD INTELLIGENCE TO RADAR SENSING

Based on ubitquous HW-MS03 microwave doppler modules, with most of what's here reusable in general for other similar devices. Tackles head-on their quirkyness and sometimes erratic operation.

This is a numbered list of Hints and Tips to get the best out of them, avoid pitfalls and improve performance.

1. INTRO

The manufacturers don't know completely everything about them as they tend to copy a standard circuit in the public domain. A manufacturer datasheet is often the public domain info issued in a PDF along with their logo. If you ask the manufacturer anything technical they tend to ask how many do you want to buy and send you the same generic pdf datasheet with their logo.

Sellers usually just know how to apply power and connect to a Arduino or what other product they sell.

Others like Bigclive.com had taken the trouble to see how they work and try some things out. 

This project digs yet deeper and gets underneath the design and the rhetoric going round and ever-cloning itself on the 'Net.

2. OVERVIEW

The idea of simple proximity sensing with RF is donkey-years old, famously used WW2 inside Allied ground-to-air 'flak' proximity shells to trigger at precise moment just passing the aircraft targetted. 

This cheap microwave sensor module we're talking about is based on a single transistor microwave oscillator/ 'S' track antenna coupled to the standard IC BISS0001 used in about every PIR sensor light, you know the one with the turny-dials which set the light shine time and how dark outside it must be. The chip is produced in millions and costs probably less than one cent. 

3. OUTPUT NOT "LOGIC LEVEL"

Contrary to what the datasheet suggests and what we're continually told including by manufacturers:  Output pin 2 does not 'output 0 V' when sensor inactive. When the output isn't H the pin is more like an OPEN CIRCUIT than a L. This can cause phantom 'stuck H' behavior or 'phantom detection' while the sensor is un-triggered. 

The only reason the output drops towards 0 V is because of what's connected is actually doing this pulling down, whether that's your multitester, scope probe you're measuring it with, or whatever.  This is because the BISS0001 IC is for sourcing a current to drive a NPN transistor base. It's an open drain driver not a true logic output and would fail to meet specification if it was. The output does not sink any current. It only takes a few microamps of current to spoil its 'L logic level' / '0V'.

What this means:  If you feed the output into a logic input, do not have any pull-up on the input !. So forget trying to connect to a TTL levels input !! Even inadvertantly configuring a microcontroller's weak pull up enabled would stop it working.

Since it's effectively a floating logic input pin while waiting for trigger you might want to add a pull-down R/capacitor, especially if there are high level of RFI around, or line voltage wires nearby. 

4. PROXIMITY, DOPPLER, VOLUMETRIC, INDUCTION?

One half of the sensor is a single transistor oscillator / snake antenna at microwave frequency, radiating continuously. The oscillator is not fixed to a specific rigid frequency.

Some of the radiated RF energy comes back to its 'S' antenna.  The oscillator becomes influenced by incoming external RF energy, especially so when very near its own oscillation frequency.

That incoming external RF usually comes from its own reflected emissions, (but could come from anywhwere including from another device of that same frequency).

Notes about oscillators and resonance:

A) There is an experiment with metronomes set to very slightly different beat rates, they are coupled with eachother by sharing a single common platform they stand on. The platform underneath wobbles slightly reacting to the movements above. The metronomes passively align their beats to coincide in sync with eachother,...