Motion Sensor II Battery Level

[tomho]

Hello all. Newbie to the forums, but I have been doing basic ISY programming for 3-4 years. 

I just upgraded from 2.7.5 to 5.0.16C, and, in regards to this post, I have a single 2844-222 v.46 Motion Sensor II. 

My battery died a few weeks ago, and I discovered it because I noticed the sensor wasn't detecting motion. So I wanted to find a way to notify me in the event of any future battery failures. 

Just to jump to the point, I found that the "Battery Level" % reading will not update unless the battery is actually disconnected and then reconnected. I discovered this by substituting an adjustable DC power supply for the battery and setting it to various voltages, using a program to run a Query every time  the motion was tripped ON. I adjusted the supply all the way from below 1V up to 3.4V and the Battery Level "xx %" reading would not change UNLESS I disconnected/reconnected the power supply. 

Also, throughout all this testing, the "Low Battery" Node was always blank. 

Just in case anyone is interested, after the power cycle; 3.4V = 90%, 3.0V = 80%, 2.5V = 70%, 2.0V = 60%. The % readings are approximate - I saw 3 to 5% variation over the couple days testing. Of course this is marginally useful information, who wants to pull the battery and re-install it just to test it. 

So, based on this test on this sensor, it appears that any sort of low battery test is just not going to work. Or am I missing something? Any bright ideas? 

 

[ELA]

[larryllix]

Yeah. I don't trust anything that comes off of the MS II units except for motion sensing. They appear to do this quite reliably but the rest is just nonsense, so far.

When queried they can go into oscillation and cripple your whole ISY system. UDI wanted to dump this unit for ISY development, until Insteon produces an API spec or some documentation.. What we have so far is only a hack from UDI efforts.

[tomho]

Thank you ELA for the information. 

On my ISY, the "Battery Level" is part of  node 1 (as is also Temperature and Luminance), and reads out directly in xx %. I had to add the node for Low Battery Status, I think it's node 3. I also tried playing with the 'threshold' but still never got it to report anything (always blank). 

I'll try the "...24 hrs.." idea. I did let it run (without cycling) overnight with no update, but never went  24 hrs without a subsequent power cycle. 

I too found that it would report motion down to as low as 0.76V with a low impedance supply. But by using as little as 5 ohms in series it would not work below somewhere  around 2.5V - it would 'lock up' and not recover unless I raised voltage to about 3V. 

 

 

 

 

 

[tmorse305]

I use this program to keep track of the battery level in the MS.  I can't take credit for it, I found it in a thread here years ago.  It uses the light/dark sensor to confirm that the motion sensor is still alive.  If the light/dark sensor doesn't toggle in less than 26 hours I get a message to check the battery.

 

DW Motion Battery Check Copy - [ID 0076][Parent 0001]

If
        'Outside / Driveway-Sensor / Driveway-Dusk.Dawn' is switched On
    And 'Outside / Driveway-Sensor / Driveway-Low Bat' Status is Off
 
Then
        Wait  26 hours 
        Send Notification to 'Text-Tim' content 'Battery Check'
        Set 'Outside / Driveway-Sensor' Write Changes
        Run Program 'DW Motion Battery Check Copy' (If)
 
Else
        Wait  5 minutes 
        Send Notification to 'Text-Tim' content 'Battery Check'
        Set 'Outside / Driveway-Sensor' Write Changes
 

[ELA]

I thought this thread was about MSII Battery level?

I use this for MSII:

The Low_bat_check  compares the variable against a low battery percentage constant and alerts when exceeded.

 

[tomho]

Thanks again ELA. You are correct, the MSll does update the battery level on about 24 hrs intervals, apparently from the last power cycle. For my original objective, this will work and wrote a program like the one you posted. 

In the interest of obtaining maximum battery life, I would like to share some additional information I have discovered. If this is repeating other data, please forgive me as I have not been aware of it. 

When in the 'sleep' mode (meaning just waiting for a motion event), the battery current is about 70uA at 3.0V supply and increases to about 100uA at 2.4V. 

With no Query, the motion ON event pulls about 35mA for about 0.1 to 0.2 second, and then 20mA for 5 to 6 sec. It then goes back into the sleep mode for the Motion Timeout period (I set it to 40 sec). Then there is another 20mA for 5 to 6 sec activity to send the OFF command. This means a single motion ON & OFF event takes as much power as about 50 min of sleep. 

With the "Query when Motion is Switched ON" is enabled, it almost doubles the ON event power, 20 mA for 9 to 11 sec. Of course no change to the OFF event. My Wait delay is 1 sec rather than the 2 sec in ELA's program above - don't know if it makes any difference. It does suggest however that to extend battery life, some kind of event counter could be set up to run the Query only after some number of motion events, depending on how much activity one is expecting. 

Another life extender is possible by changing the "Motion Condition" to something other than "Always". I tested the "Only At Night" mode. When motion is detected in this mode when it is 'daytime' (probably determined by the the Day or Night Threshold values - didn't try adjusting), there is only the 35mA current lasting for 0.4 to 0.9 sec. It does not use either the ON or OFF 20mA current. Of course it reverts to the above when it is night. 

 

[ELA]

search for his thread:

The Mysterious Motion Sensor II

 

[oper8]

I tried the program presented by ELA and I don't get the value of the battery level loaded in the variable I create for the notification. I don't think the motion sensor is reporting the battery level in any way.... I might just give up on having a program notify low battery and just change it every 2 years max. I hope a higher capacity CR123A can last for 2 years on a normal use.

[ELA]

Have you turned on the battery node in the MSII?

Turn on a level 3 event viewer and confirm that the MSII Query is sending the required data required to load into your variable. 

[oper8]

I left it on and now it does load the value in the variable. It seems it's working but I don't 100% trust the number. It's showing 81%. It was installed 2 years ago...

I'm slowly retiring the INSTEON devices. I see not much future with them. I'll be moving all my devices to Z-Wave. And I still want to continue to use ISY. I love it.

[larryllix]

Two of my three units were battery run and those two batteries both went dead in less than a year each. One went dead with the spare battery in that time so it consumed two batteries in less than a year. All three are on USB adapters now and that hassle is over now.

My other 12 MS1s are still on 9v batteries and I get between 8 to 15 months out of most discount (alkaline Walmart, Panasonic, EverReady) batteries, depending on activities.

Duracell 9v batteries lasted about 4-5 months and I finally gave up on them since they became junk.