UPDATE: Replaced my PIR-Sensor with a Doppler Microwave Sensor.

uros76

@mickyvi that is sad to hear. The glass you are using must have some special coating that prevents signal go through.

Did you try to rotate the sensor? I know these sensors are sensing around 360degree but I think the actual antenna is stronger on one side.
Also, what’s the distance from mirrir you are trying to get the sensor to pick up movement? Does it work fine when using without glass?

For me PIR was not an option. I tried cheap and expensive PIR sensors and all of them had too many false positives. Plus it needs to be ouside visible.

MickyVi

@uros76

Hello,

Yes, in my opinion it’s the mirror, damn it, if you only knew how much I paid for it …

Yes, I tried in every way, even putting the sensor behind the monitor (… when you are desperate you try everything …: -) …), turning it, putting it sideways but nothing, it does not detect.
If I move it outside, it detects everything, but even with the “cheap” sensors I had noticed the same functionality.

I don’t know, maybe I will also give up motion detection: maybe a button that takes it out of stand by and then stays on for 20 minutes … I know … it’s shit … but I don’t want to continue to have the mirror resting on the floor without being able to install it and continue testing.

I have to think about it for a while.
Hello and thanks for the support everyone!

uros76

@mickyvi as the last option maybe install the sensor on top or bottom of the mirror frame. Hide it in a small enclosure (wooden or hard paper or 3D printed extension to the frame) and paint the same color as the frame. If done nicely it won’t even look that bad. And it will for sure work :)

Fozi

@mickyvi That’s indeed really strange, and I can only guess that the mirror has a reflecting coating which prevents the signal from passing through the mirror. The thickness of the glass itself is not the limiting factor, as the signal usually passes even stone walls and can detect motion in adjacent rooms. I operated my sensor first behind the mirror, too, and it worked. It is a 4mm Pilkington. Later I moved it under the frame, because it was easier to trim the sensor and hid it behind a 3D-printed cover. Thus, I’d second @uros76 proposal to mount the sensor on top or on the bottom of the frame where it is not visible. Don’t give up!

MickyVi

Hello everybody,

I’m taking the path of the classic infrared PIR sensor.

I made two shells that can be attached under the two arms of the lamps above the mirrors.

Here still in the assembly / painting phase:

I’m gonna give a nice coat of chrome today and we should be there. :-)

Fozi

@mickyvi That’s a totally awsome solution! Really cool!

MickyVi

@fozi
Thanks, the compliments from such experienced people are always very welcome !!

jordan1200

Thank you for this awesome solution! I just got mine up and running with MMM-PIR-Sensor. I was wondering if any of you have got this microwave sensor to also work with MMM-NearCompliments? I was thinking since it was also inspired by PIR sensor, I could get it to work, but no luck yet.

Lipax

hello,
i tried this with the microwave sensor, but im out of ideas. i get wrong positives every minute. (i check every 5 sec).
i added also the ferrite, tried another raspberry, different of these sensors (bought a 5 pack) and placed it in a different room. anyone got any ideas what else i could try?
or maybe similar microwave sensors i could try?

any help would be great, tyvm

uros76

@Lipax said in UPDATE: Replaced my PIR-Sensor with a Doppler Microwave Sensor.:

hello,
i tried this with the microwave sensor, but im out of ideas. i get wrong positives every minute. (i check every 5 sec).
i added also the ferrite, tried another raspberry, different of these sensors (bought a 5 pack) and placed it in a different room. anyone got any ideas what else i could try?
or maybe similar microwave sensors i could try?

any help would be great, tyvm

Did you try and limit the range as described in beginning of this thread? For me option 2 helped to limit oversensitivity.

Lipax

@uros76
the range is not the problem. i get positives when nothing is moving near it

uros76

@Lipax this sensor is very sensitive and catches any movement 360 around from several meters away. I had same problem… If you don’t dial down the sensitivity-range you will keep getting false positives.

kayakbabe

@Lipax try dialing down the sensitivity, it’s probably picking up dust or something small you can’t see.

Lipax

thanks for the help, but i doubt its some kind of movement since its exactly every minute. i tried a different room and its the same.

sdetweil

@Lipax please don’t submit duplicate posts…

all users with reputation below 2 have their posts reviewed… this eliminates all the porn, advertising and trash posts from the forum.

it just takes time for (me) to get to review the posts…

uros76

Can you share your python code? Perhaps there’s something there

Lipax

@sdetweil sry about that

@uros76 i hope thats not the problem

import RPi.GPIO as GPIO
import time
import datetime

switch=11


GPIO.setmode(GPIO.BOARD)
GPIO.setup(switch, GPIO.IN)




try:
    while True:
        if GPIO.input(switch) == 1:
            now = datetime.datetime.now()
            print(now.strftime("%Y-%m-%d %H:%M:%S"), "something")
            
        time.sleep(5)
        
        
except KeyboardInterrupt:
    print("Bye Bye")
finally:
    GPIO.cleanup()

thanks for the help

uros76

@Lipax try different gpio pin, like a pin 18 which is a GPIO 5. It’s default set to input function. Don’t forget to adjust the python code, change switch to 18.

You might check the GPIO 0 (your pin 11) if it is set to input. Use command: raspi-gpio get 0 or raspi-gpio get to check all GPIO’s.

Also, use a multimeter to measure the voltage coming out from sensor. It should be around 3.3V for few seconds when sensing detected. This way you can monitor if actual detection / voltage is coming from sensor.

Lipax

@uros76
i forget to mention i already changed the pin, i first tried pin 7 before

according to pinout.xyz my pin 11 is gpio 17

GPIO 17: level=0 fsel=0 func=INPUT

thank you very much for helping me, i dont have a multi meter but ill ask my father to do that. u think its not the sensor sending the signal? maybe it picks up something with the wire along the way? i dont have a engineering background.

uros76

@Lipax you need to be sure that sensor is not the cause.
Measuring it’s sensing, output voltage and duration is a good way of confirming if sensor works ok.