A good source of hackable home automation parts has come for a while in the form of inexpensive modules offered by large retailers such as Lidl, or IKEA. They’re readily available and easy to play with, they work with open source hubs, so what’s not to like! As an example, [Circuit Valley] has an IKEA Vallhorn motion sensor for a teardown, it’s as you might expect, a passive infrared sensor (PIR) sensor coupled with a Zigbee interface.
Inside the ultrasonic welded case is a small PCB and a Fresnel lens on the inside of the top cover, and a small PCB for the electronics. We applaud the use of a Swiss Army knife can opener as a spudger. The interesting part comes in identifying the individual components: the Silicon Labs EFR32MG21 SoC is easy enough, but another mystery 8-pin chip is more elusive. The part number suggests an Analog Devices op-amp for signal conditioning the PIR output, but the pinout seems not to support it and from here we think it’s too expensive a part for a budget item like this.
There’s a handy header for talking to the SoC, which we’d love to report is open and ready to be hacked, but we’re not getting too optimistic. Even if not hackable though, we’re guessing many of you find uses for these things.
Brief note: Simple single-sensor IR motion detectors do not play well with mini split heat pump systems. They trigger on the moving plume of warmed air. A real ir camera and image processing might solve that, but would drive up cost and/or effort and could introduce its own security / privacy issues. So I’m reluctantly starting to think about solutions other than IR.
It’s highly unlikely they “trigger on the moving plume of warmed air.” because the emissivity of air is pretty darned low compared to the objects that said warmed might hit. A sweeping warm air plume will cause rapid temperature changes of things like fabric or lampshades as it hits them.
So, just turn off the swing or sweep function.
This is a fairly well know issue with PIR (Passive Infra Red) security systems. They get false positives due to force air heating systems turning on. The problem is usually solve by combining PIR with some other sensor (audio, ultrasound, light, etc.).
PIR security systems should be fine for radiator heating, underfloor heating, and any heating system that gradually changes temperature (i.e not burst of hot air on start up).
A second cause can be an insect or spider walking across the lens. It will momentarily obscure any localized warm spots in the room, and thus generate an event signal. Because it is adjacent to the lens it effectively appears huge to the sensor.
My outside PIR light is triggered by wafts of warm air form my boiler flue so yeah it’s a problem.
Warm fog has extremely high emissivity in the IR…
A more common problem is the difference between security and lighting sensors. The simpler designs only require a single event to trigger the unit, whereas a well-designed security sensor will require two or more in quick succession. This is to overcome so-called popcorn noise in the sensor itself, where the pyroelectric ceramic will spontaneously generate a large sharp noise impulse. A person moving across the field of view will normally generate many such signals as they pass between the field of view of several lens facets. Obviously the consequences of a false trigger in a lighting circuit are less serious than one in a burglar alarm system.
Or an automated turret!
Load of BS. I’ve had PIR motion sensors in nearly every room of my home for my alarm system for more than a decade and they have never once triggered because the central heat / AC kicked on.
you are so lucky
Though it might superficially resemble one, the lens in front of a PIR sensor is not (necessarily) a Fresnel lens. It’s a set of independent lenslets (each of which could be a Fresnel, but usually is just a simple lens).
The idea is that the array of lenses make a set of narrow(ish) beams separated by blind spots. A warm object passing in front of the detector passes from sensitive areas to blind areas, making an AC signal in the sensor. That’s what is detected, not the absolute (DC) object temperature. Stationary hot objects are not registered.
Interesting teardown and interesting price ( 9,99€) . Really interesting for me would be if the pir sensor and circuit is ultra low current. In the past such sensors for example from murata was really expensive. For me it would therefore interesting to know more about the sensor and current consumption. Thank you really for the interesting article, your work and time.
A couple of weeks ago I literally tore one of these down to fix it’s biggest flaw, the LDR is so badly obstructed from light that it’s nearly useless. Here is how I fixed it https://x.com/geoffrey_mcrae/status/1855802982702715382
Could you expand on that?
De-soldered the LDR, put it on some kynar wire, drilled a tiny hole in the enclosure and glued it in place outside the enclosure. Then simply soldered the leads to the original component location.
The only sesnsor needed for true high millimeter accuracy and rough environmentaly safe radar sensor. It can live on a small battery for years! Check out http://www.acconeer.com
Could that 8 pin device be a ZMOTION processor? It seems a little overkill, given the SOC next to it and the low price point, but it nicely offloads the motion sensing.
https://www.zilog.com/index.php?option=com_product&task=parts_zmotion&BL=147&familyId=151&productId=Z8FS040BS&product=ZMotion%20MCU&Itemid=77
If only there weren’t major negative implications and we could just RFID chip everyone for presence detection.
It’d be so nice to be able to just have things automatically know who is operating them or where someone is.
A simple bracelet or ring would work. So would an UWB beacon in your cell phone. And if you wanted to just put it in a drawer.
I accidentally pressed ‘Report Comment’ instead of ‘Reply’ on aliaalia’s comment, which is actually blameless and has now vanished. Sorry.
I have no technical know how but enjoyed that. How would you go about learning how to hack or tear down electronics?
You could try Big Clive (BigCliveDotCom) on YouTube
https://www.youtube.com/bigclivedotcom
To elaborate more. I just went and looked at all of my sensors and at least 3 of them are directly facing air vents so they should be triggering if this were true. PIR sensors don’t trigger because heat suddenly appears. They trigger when a source of heat moves across it’s field of view. This is why there is a lens on the front of them that creates multiple points of “light” aka heat across the sensor so even a small movement will trigger the sensor.
Well, the funny thing is that the manual of the device clearly states to avoid hanging it above or in Line of Sight of a heating device. Why would that be, one could ask…