PIR sensor got very sensitive

[gregebert]

Here's a strange story....a few weeks ago, I noticed my 7971 clock was running a lot more than it was a few years ago. After some experimenting, I found out the PIR sensor is getting a lot of false triggers (no movement in room, even covering-up the sensor didn't keep it off). Turning-back the sensitivity 1/4 turn did nothing, so I put it at minimum sensitivity and it seems to stop the false triggering. It still detects motion.

Anyone else see this happen ? The sensor has been in-use for about 6 years; it's one of those cheap 5V units that cost about 1 USD and worked fine for many years.

[newxito]

Maybe you have to clean the trimmer resistors or there is a little bug inside, I had that problem but it was with an outside PIR :-)

[Mac Doktor]

On Nov 21, 2024, at 1:50 AM, gregebert <greg...@hotmail.com> wrote:

the PIR sensor is getting a lot of false triggers (no movement in room, even covering-up the sensor didn't keep it off). 

Power supply problem?

Terry Bowman, KA4HJH
"The Mac Doctor"

https://www.astarcloseup.com

“...the book said something astonishing, a very big thought. The stars, it said, were suns but very far away. The Sun was a star but close up.”—Carl Sagan, "The Backbone Of Night", Cosmos, 1980

[Max Di Noi]

I had that problem once and it was due to a faulty capacitor on the high voltage power supply creating instability on the HV circuitry and affecting the PIR circuitry, which was odd as that itself had its own capacitors located close to the chip and the main power supply had plenty of stabilised power. 

Now I'm trying to stabilise a microwave motion sensor that I designed based on one of those commercially available. It's embedded in the pcb, works OK but I need to figure out how to calibrate it more easily. It depends on temperature and material of the pcb. I purchased a frequency receiver up to 6gHz and I can see the signal on the screen. So that's a good start! Sorry, got sidetracked, slightly off topic :-) 

--
You received this message because you are subscribed to the Google Groups "neonixie-l" group.
To unsubscribe from this group and stop receiving emails from it, send an email to neonixie-l+...@googlegroups.com.
To view this discussion, visit https://groups.google.com/d/msgid/neonixie-l/f054d906-771b-4e73-9593-53c3e95e56f5n%40googlegroups.com.

[gregebert]

So I checked my supply voltages, and that opened an unexpected can of worms. The PIR sensor has it's own onboard linear regulator, and it's pumping-out the expected 3.3VDC. The weird thing is the main 5V supply was only putting out 3.8V. Despite that, the Raspberry Pi also had the correct internal 3.3V supply (it also has it's own onboard supply).

So, the first mystery is why a DCDC converter rated for 1amp, and typically supplying 400mA conked-out when it's not getting warm or overloaded. I confirmed the RasPi typically uses around 200mA, and occasionally peaks around 280mA.

I have to dig further into the regulators on the RasPi and PIR sensor to understand why they still produce 3.3V when their input is way-below 5.0VDC. I'm glad that they do, but I want to confirm why.

I'll see what I have laying around for replacing the 5V DCDC converter, and if they beefier ones I have will fit on the PCB.

Once I get this thing back together, I'll see if the PIR sensor is back to normal. I have some replacement PIR sensors on the way.

Why the DCDC converter failed is another mystery; I use these on all my projects, and seeing one fail is disturbing, especially because many of my gizmos dont have onboard A/D converters to monitor the power supplies.

[Instrument Resources of America]

Having only a 1/2V of headroom is not good. I'm surprised that you're getting your 3.3V. Anything else operating from that 5V supply could be real iffy.

Ira

To view this discussion, visit https://groups.google.com/d/msgid/neonixie-l/340ffb9a-b775-4a49-b1ae-0fde15e3ac6bn%40googlegroups.com.

[gregebert]

For the PIR sensor, the LDO regulator datasheet doesn't spec the minimum input voltage (Terrible, I say). I measured +3.3 out with my DMM, but that doesn't tell the whole story, so I would need to put a scope on it to be sure it's stable.

For the RasPi, the datasheet says we can run down to 3.5V and still get 3.3V reliably. So that explains why I never had any crashes, etc.

Later today I will get the supply fixed back to +5V and see what happens.

[Leroy Jones]

You have a messed up power supply.   It is putting tons of noise on the power bus.   Take a sensitive high impedance audio

amplifier and connect it across Vcc and ground.   Listen.   It should be very quiet if it's correct.   I bet yours has all sorts of hash in it.

That hash will cause all sorts of weird circuit behavior.

[gregebert]

Audio-amp idea is a very good suggestion; I have to save that for next time as the DCDC converter has been replaced and the old one was damaged during removal.

Problem has been fixed, and no false-triggering on the PIR sensor, even at higher sensitivity. I suspect an LC filter the input to the DCDC converter may have caused overvoltage over time. The filter was recommended by the manufacturer of the DCDC converter, but at the last minute I swapped to a different DCDC converter without changing/removing the LC filter. It worked fine for over 7 years, but damage was likely being done since Day 1.