BS612 Seems to get stuck detecting motion

[jpmc42]

Hi:
Looking for suggestions. Using a BS612 motion detector with a ESP32-S3 feather to create a network reporting motion detector. It seems that once motion is detected, in many cases, the sensor can continue to report motion for quite a long time (> a minute to several minutes) after motion stops. Note that sometimes works as expected. Its not consistent. Wired according to instructions. Using gpio 6, pinMode(T6, INPUT_PULLDOWN), so should not be a floating issuse. Put a .1uf cap across power and ground thinking would help with noise. Leads are about 4 inches long to mount on simple case. Currently powered by usb battery pack. Suggestions would be appreciated... Perhaps need to try other sensors...
Thanks much.

[bidrohini]

Are the motion sensor and the ESP located close to each other? If so, keep them apart. Because I have read earlier that these sensors interfere with the internal circuit of ESP32 modules.

[jpmc42]

Thanks for the info. That is what I suspect is going on. Is there any info available on recommended distance? I have only tried a couple inches as the data sheets also recommend short leads. This seems to help, but only a little. Will try a longer distance.

[jpmc42]

Moving up to 8 inches away, while minor improvement, does not seem to
make things any more consistent.
Tried the PIR Sensor (product id 189) after modifications to run at 3.3 v and
it works exactly as expected. Even within an inch or two from the micro.
However the larger sensor (#189) also adds packaging/mounting complexities.
Looking at the larger pir sensor, I see that there are a couple 47UF caps on the
board. The ref circuit for the 612 shows a 10uf cap. So, next tries will add more capacitance to
the smaller sensor. Will also try it with a pi pico W to see if I can get my cost and
size back to where I wanted...
Thanks for all the feedback!

[jpmc42]

Just in case anyone is interested, I ported code to a pi pico w
and am now using a BL412 sensor. I added a .1 uf cap and a 54uf cap
across the power and ground pins and the instability is gone.
The project now works as I would like, and the cost is below
where it needed to be. So, a success all the way around.
It seems the smaller (and less expensive) sensors need extra
capacitance when used with a micro with wifi.
I dont have access to an analyzer so cant tell what kind of noise
is getting in the way, but the added 2 caps seem to get rid of it...

[adafruit_support_carter]

Wired according to instructions

Just to check - this included wiring the ONTIME pin to GND?