X-band Microwave sensor housing issue

[rizwanspirit]

I am using Parallax X-Band (10.5GHz) Motion Detector, with AVR microcontroller, wireless communication is being done through zigbee 900 MHz module. The housing of this circuit is made of acrylic material. The sensor works fine when it is outside the box, when it is mounted inside the box it starts sending false pulses. After a research i came to know that acrylic have high reflection coefficient for X-band. so i replace the acrylic in front area of sensor with glass but the problem is still there.

The radiation pattern have some area in the back side as well, please tell me that if a complete side of acrylic is replaced by glass , will it work?
Secondly can the Zigbee 900 MHz module antenne cause any interference?

 

[FvM]

Every dielectric window in front of the sensor will at least reflect part of the RF. Er and thickness are the main factors. Acrylic is probably better than glass (lower Er), but you should try to reduce the thickness.

Many motions sensors are however operating through windows. You didn't describe the problem clearly, is it modified radiation pattern or signal loss?

 

[rizwanspirit]

The above attached antenna radiation pattern is provided by the company in their product datasheet.
The thickness of acrylic material is 6mm and the glass portion have thickness 4mm.

 

[jiripolivka]

Do not attempt to use glass - a typical window glass attenuates some 4 dB at 10 GHz. This is why acylic is used.
Try to adjust the distance between the window and horn aperture. Check that your test is done in a "free space", your hands and body far away.

- - - Updated - - -

If you acrylic is still a problem, get a thin Teflon foil instead.

 

[rizwanspirit]

@jiripolivka The attanuation caused by glass must decrease signal strength , and hence the detection range, why false alerts are still there?

 

[Terminator3]

Try connect sensor output to computers mic input, record signal with and without enclosure, microcontroller, zigbee, etc and see spectrum or just listen to it

 

[jiripolivka]

I agree with the above advice. Combining microwave sensors with digital circuits generates problems of unknown origin.
I would follow an analog test first. Maybe the sensitivity is set too high.