PIR AM312/412/612

[Nca78]

Hello, I discovered those PIR sensors on Aliexpress today.
They are pretty cheap (2$ for AM312 module, less than 2$ a piece if you buy a lot of AM612 sensors) and all the electronic is included in the PIR sensor "can", no need to have a complex circuitry.
AM312 is a basic type but AM612 has settings for trigger duration and sensibility. It claims to run at 2.7V and with a maximum working current of 20µA (typical 15µA). Means it could be fine on a button cell.

Has anyone used any of the sensors (would probably be the mini am312 module) ? I'd like to know if the claims about power consumption made in the datasheets are real before I dream too much and order some am612 sensors :)

There was a subject about the am312 here :
https://forum.mysensors.org/topic/3658/pir-am312/1
but it didn't go very far.

[NeverDie]

@Nca78

I finally received my ten am612 sensors. Any progress on the blind-spot issue since your last post?

[Nca78]

@NeverDie sorry not yet.
I will make a case and install one or two in my living room to see if it's a problem to detect presence. But not time yet.

[NeverDie]

From looking at the X-Y sectional view in the datasheet (page 9 of http://akizukidenshi.com/download/ds/senba/Pir-Am612.pdf), I'm guessing that having a second sensor, but rotated 45 degrees relative to the first sensor, will help fill in the detection zone gaps. I'm unsure whether additional skewing of one relative to the other will be needed though, or how I would even execute that.

@Nca78 Is sensitivity at maximum when R6 (in the reference design) is 0 ohm, or when it's 333K ohm?

[NeverDie]

So, I've ordered this to see if it helps at all:

The second sensor is rotated exactly 45 degrees relative to the first. I adjusted the headers so that both sets will fit into a 2.54mm pitch breadboard.

It might (?) be that I should space them further apart so that they don't block each other's detection zones, but hopefully it won't matter much, if at all.

[NeverDie]

I also wonder how well (or not) cramming more than one sensor element (rotated relative to one another) under a single lens would work? It obviously runs counter to what's intended, but it would be nice to have a total footprint that's small in size.

[NeverDie]

So, perhaps this might work:

Maybe even three am612's could be crammed under one lens, each rotated 30 degrees relative to the other two?

[NeverDie]

Here is the am612 installed in my primitive breakout board:

Once I wire it up on a breadboard, I'll be testing the different IR lens I received to see whether maybe it performs any better than the lens typically used on an SR501.

[NeverDie]

Good news! I wired up my primitive breakout board on a breadboard such that it turns on a blue LED if it detects motion. I have (I think) set sensitivity to maximum by using just an 82K pulldown resistor and a 4.7uF capacitor on the SEN pin. I powered it using two AA alkaline batteries.

Bottom line: Even in this half-baked breadboard prototype, it seems to perform much better than the am312 I tried earlier. I'm not even noticing any significant blind spots. :) Go figure.

[Terrence]

@NeverDie said in PIR AM312/412/612:

I'm not even noticing any significant blind spots

That's great news.

[NeverDie]

Attached is my attempt at a conversion of the reference schematic on page 9 of the AM612 datasheet (http://akizukidenshi.com/download/ds/senba/Pir-Am612.pdf) to a setup where the AM612 is at maximum sensitivity and minimum "on time" after being triggered.

Can it be further simplified?

0_1498834397166_Schematic for AM612 Breakout v402.pdf

For instance, do the 10uF capacitors any longer serve a useful purpose? Or would eliminating them be of no consequence?

[Nca78]

I can't open your document.
You confirm that the test is with he double PIR PCB under one lens ?
I'm surprised that this work as the lens is supposed to focus the IR on the two sides of the sensor, and if you move them away of the center it should lower sensitivity.

[NeverDie]

@Nca78 said in PIR AM312/412/612:

I can't open your document.

Strange. It's an ordinary PDF. I'm even able to open it on my wife's computer.

You confirm that the test is with he double PIR PCB under one lens ?

What I tested was just a single am612 PIR under one lens. I re-tested last night to be sure, and, indeed, it seems to have excellent sensitivity and no meaningful blinds spots.

[Terrence]

@NeverDie I can't open it either. Here is the error message.

[NeverDie]

@Terrence
Thanks! Not sure what's going wrong. Let's try it again from a different computer:
0_1498851552407_Schematic for AM612 Breakout v402.pdf
In case that still doesn't work, here it is as a .PNG file:

which is large enough to be legible and will serve the purpose.

[NeverDie]

I guess answering my own question, I removed my load switch as well as all of the capacitors from my breadboard circuit, and the remaining am612 circuit still appears to function just fine. :)

[NeverDie]

I updated the board, and it's now 23mmx24.5mm:

It also has pads for an LED for range testing purposes.

[NeverDie]

I also ordered some of these lenses:
https://www.aliexpress.com/item/FREE-SHIPPING-50PCS-PIR-D203S-LHI874-LHI878-RE200B-P228-Fresnel-lens-PIR-Sensor/1295771560.html?spm=2114.01010208.3.31.Be4wmU&ws_ab_test=searchweb0_0,searchweb201602_5_10152_10065_10151_10068_10130_10084_10083_10119_10080_10082_10081_10110_10178_10136_10137_519_10111_10060_10112_10113_10155_10114_437_10154_10056_10055_10054_10182_10059_303_100031_10099_10078_10079_10103_10073_10102_10120_10052_10053_10142_10107_142_10050_10051-10120,searchweb201603_49,ppcSwitch_4&btsid=f043368c-9fc4-4f1a-ae33-850a00860ef5&algo_expid=2cec72d6-8e82-4481-92b3-9a4b26711616-4&algo_pvid=2cec72d6-8e82-4481-92b3-9a4b26711616
They look a little different, so I thought them worth a try. While not plug compatible with the exist SR501-type layout, maybe they can somehow (?) grip the outside of the PCB instead, allowing an even slightly smaller PCB. If only they had manufactured the barbs facing inward rather than outward, they might have been perfect!

[NeverDie]

@NeverDie said in PIR AM312/412/612:

I guess answering my own question, I removed my load switch as well as all of the capacitors from my breadboard circuit, and the remaining am612 circuit still appears to function just fine. :)

It's a somewhat surprising result, because even the cheapskate am312 sensors appear to use caps:
https://www.aliexpress.com/item/Mini-IR-Pyroelectric-Infrared-PIR-Motion-Human-Sensor-Automatic-Detector-Module-high-reliability-12mm-x-25mm/32749737125.html?spm=2114.01010208.3.32.JSXzcv&ws_ab_test=searchweb0_0,searchweb201602_5_10152_10065_10151_10068_10130_10084_10083_10119_10080_10082_10081_10110_10178_10136_10137_519_10111_10060_10112_10113_10155_10114_437_10154_10056_10055_10054_10182_10059_303_100031_10099_10078_10079_10103_10073_10102_10120_10052_10053_10142_10107_142_10050_10051-10120,searchweb201603_49,ppcSwitch_4&btsid=b87e2064-287c-4fe9-ba44-53ca575f5350&algo_expid=3e550c0f-0a35-465b-8691-27d092d21bbd-4&algo_pvid=3e550c0f-0a35-465b-8691-27d092d21bbd
I coun't three caps there. Surely not all of them are there because of the LDO?

[Nca78]

@NeverDie you would have to check the LDO datasheet but it's possible.
Sorry on holidays without computer so I cannot check.