0
\$\begingroup\$

I have a microwave motion sensor that can be triggered in almost every angle.

enter image description here

enter image description here

I want to limit its angle of view to only one side of the board but so far I had no success.

What I have done:

  1. Designed my board so that the surface that module mounts on is a ground plate. It is still triggered by motion that happens behind it.

enter image description here

  1. Put an extra copper plate almost three time size of the whole device behind it and connected its copper layer to the ground of the board.

enter image description here

The only thing that comes to my mind and I haven't tried is to reduce the distance between the module and the board that it is mounted on to few milliliters. (Now it is about 1cm.)

\$\endgroup\$
2
  • \$\begingroup\$ Have you tried not connecting the copper of the board to ground? Some devices also block signals which are very near to it because of antenna and housing reflections and those are then ignored in the signal processing afterwards, so the distance might need to be increased. \$\endgroup\$
    – Arsenal
    Commented Aug 25, 2023 at 13:06
  • \$\begingroup\$ @Arsenal Thanks, I've just tested without ground connection. No difference... \$\endgroup\$
    – AKTanara
    Commented Aug 25, 2023 at 13:11

1 Answer 1

2
\$\begingroup\$

Oh my! What an interesting design that PCB is!

It looks to me as if the copper ring on the bottom of the board is actually used as a resonator, which, together with the FET on top of the middle of it forms an active oscillator. The circular patch (with the company logo) in the middle on the bottom would then form the actual antenna, and looking at the two feed vias, it's circularly polarized.

So,

The only thing that comes to my mind and I haven't tried is to reduce the distance between the module and the board

that would most likely only succeed in detuning that resonator.

Your radar (and that's what this is, at least if we're a bit lax on the "ranging" part) is plain quite sensitive; it probably works happily through the multipath reflections that occur (nearly losslessly) on your copper plane and then your environment.

You have to face the fact that this device was meant to be sensitive to movement in the general environment – the patch antenna that's smaller in diameter than the resonator is long is a testament to "we don't want this to be very directive".

In all honesty, a different sensor board might be the cheapest and easiest solution; everything else will, in the end, involve microwave engineering, expensive (think: house-it-stands-in-priced) measurement equipment, or much trial and error.

You can put your whole sensor system in a closed copper-clad box, and by the design of slits in that box couple out the energy into a waveguide and finally an antenna of the directivity you want (or go for slit antennas). Or, you could try to design a near-field "antenna" that you put on the circular patch that's very good at coupling the RF out of that near field, and use that to feed an antenna of your choice, but:

  • that antenna will necessarily be much larger than the on-board patch to achieve that directivity,
  • all of this will require quite extensive design, and FEM RF modelling,
  • both approaches will in first prototypes most probably work suboptimally and reflect too much energy back into the system, deafening it, and possibly even destroying the receiver.
\$\endgroup\$
1
  • \$\begingroup\$ Thanks for the comprehensive analytical answer. I'll go for testing all other modules of the same kind to find the one with desired characteristics. \$\endgroup\$
    – AKTanara
    Commented Aug 28, 2023 at 18:58

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service and acknowledge you have read our privacy policy.

Not the answer you're looking for? Browse other questions tagged or ask your own question.