Wireless PIR / Microwave Sensor recommendations? (want wifi + batt powered + Tasmota + cheap)

I want to set up motion detectors around my house for triggering of automation and better camera motion detection. Are there any detectors that match the following:

1) WiFi-enabled
2) Battery Powered (at least 6 months/battery)
3) Tasmota-flashable
4) Long range detection
5) Cheap (under $15)

I originally planned to hardwire the sensors to a centrally located Rpi with my stock I have below, but I really do not want to run/fish wire throughout my house if there’s a more efficient wireless equivalent.

PIR Sensor - $8.99 / 5

Microwave Sensor - $8.99 / 5

Any ideas on devices that match/are close to matching the 5 criteria outlined above? Or smart alternatives?

Edit: it seems like Wifi is definitely a no-go for battery life. So now the question shifts to - has anyone built a cheap solution with a low-power transmitter + sensor (receiver can be a usb plugged into the hass server), and if so, could they share the components?

Wifi and battery operated with a battery life of 6months would be difficult.

I have used those Microwave sensors with an esp8266 (Plugged in) it was way too sensitive for me, in an apartment it was detecting movement though walls and though the floor (The neighbors) it might be ok if your in a house.

I use the battery powered xiaomi PIR sensors with Zigbee2mqtt

I ran a nodemcu on wifi off a 10,000 mah battery pack as test and it only lasted about 4 days.

it was powering two small (20ma) LEDS continuously and measuring two thermocouple probes every 10 seconds.

Well, it sounds that Wifi is out of the question for battery powered operation. I now wonder if a low-powered transmitter would work better. I’d connect the sensor, transmitter to a battery and then have a usb receiver on my hass server to receive the signal. I don’t know what cheap solutions people have come up with, but hopefully someone will chime in.

Anyone have experience using xiaomi / aqara PIR sensors with zigbee2mqtt? The aqara listed below is $11.59:

And I assume I’d buy the tools and stick for zigbee2mqtt ~$20:

Wifi+battery+pir is doable using the “ch_pd wake” method, as long as the location doesn’t see high traffic… I did this PIR esp8266 device that lasts about 2 years on a single 18650 charge, but it only gets triggered briefly ~5 times per day:


It requires soldering a bare esp module, an smd regulator (not mentioned in the link, but I used an ADP-160au-3v3 regulator), and an smd buffer. That means it’s not for everyone, but if you have the skills and time IMHO the results are well worth the effort. You can find easier to solder stuff, but you probably will get much higher idle currents, and it won’t last more than 1-6months in the same situation.

Also, don’t forget about esp-32 lora… those can be setup for very long battery life as well… but the setup is quite a bit more involved (you need 2 esp32-lora modules, one in the device, another as a ‘hub’). They’re meant more so for infrequent bursts of data, but I image they would be well suited to motion use.

Regarding the rcwl-0516, I use some of those as well, but IIRC they have a pretty high idle current (mA range). So I don’t think those would be well suited to battery devices in general; an 18650 is lucky to last a few weeks at over 1mA, so…

Did you check mysensors?

1 Like

Thank you for the link, the one below gives me some good information:

If anyone wants to share their experiences building battery sensors or using zb2mqtt/xiaomi PIR sensors, it would be appreciated.

The main thing I can tell you from my experiences building battery powered IOT stuff, is to pay attention to current specs everywhere you turn, and be ready for some detailed solder work (the best parts are usually smd). It is tempting to grab some spare regulator breakout board from of your spares box and use it in a build, but when it comes to battery devices that’s likely not going to work out well. You need to search for components based on specs you may not usually pay attention to, like Iq for regulators, and idle current for anything else. That article hints to this… they hack off the built in LDO form the nano boards because those LDO’s suck for battery use. Also, some things that you may not normally think about can kill batteries fast, like a power status LED… or say you want to measure battery voltage with a divider and analog pin, well better think about input impedance because you’ll want to use as large of resistances as you can get away with. That 4k7 pulldown… maybe 47k would work? What about those stiffening caps… do they have low leakage? Believe it or not, most electrolytic we are familiar with are trash for battery use because of leakage, and unfortunately you usually need a pretty big cap to stabilize most low quiescent LDO’s.

Bottom line, literally everything on a schematic is fair game to modify when it comes to saving battery power. So keep this in mind… otherwise you might be settling for less than you are capable of. Like, you could otherwise have a 3year circuit, but your battery needs charging every 3months just because of capacitor leakage. It’s daunting to think about at first if you haven’t done battery designs, but once you get familiar it is way more fun that the usual 'duino breakout board level projects. The pace of tech means we see lots of neat new efficient stuff these days. Regardless, keep a list of those diamond in the rough part #'s handy once you find them. :wink:

1 Like

A lot of great information here, thank you for sharing :slightly_smiling_face: