Iris V2 motion sensor ... stuck temperature and accuracy

I picked up a few of the lots of 10 Iris V2 motion sensors of ebay. While I primarily got them for motion sensing, I also hoped to use them to keep track of temperature changes in different rooms throughout the house, outside, and even in the attic.

As advertised, some look used (dirt on one, superficial cracked plastic on another), but all ten came with batteries (two different brands), all reading 100%, with a plastic tab in place keeping them powered off.

One, unfortunately, had a very loose battery (too loose to make contact; rattled inside before I even opened it) and when I removed the battery, the smaller contact/tab that holds it fell out. When I tried to bend the piece that remained, that broke, too! Warning to others: don't try bending these tabs, they are apparently very brittle and will NOT bend!

The remaining nine paired fine (although I've probably exceeded how many I can add per wired Zigbee device; I'll know what to look for if they start going offline). I have them lined up on my desk for now. All nine are tracking motion (and lack of motion) wonderfully; reporting motion almost immediately and lack of motion in about 30 seconds.

However, temperatures are all over the place!

I plan to get some graphs setup, but in the meantime:

After being connected for 15 minutes, they read:

76.03
84.12*
71.82
74.25
81.38*
75.94
74.88
74.52
74.56

*That 81.38 and 84.12 haven't budged, so they're apparently faulty. Any ideas to them back to life?

An old Go Control motion sensor says this room was 71. An old analog HVAC tech temperature probe I have read about 72. I'd argue 72 is probably a good reference temperature, so I set them to offset to 72 from their current temps and moved them to the garage to see how the reported temps align in that area.

After about ten minutes, the Go Control was at 94. The analog temperature probe, 91.

The Iris, after the same amount of time (and with the offset applied earlier), were lagging far behind.

77.12
84.12*
81.22
80.66
81.38*
82.17
82.22
81.63
82.8

After an hour(!), Go Control 96, probe still at 91, most of the Iris were catching (or caught) up:

88.33
84.12*
85.3
89.37
81.38*
90.69
89.96
89.39
90.14

Temperature accuracy is a little disappointing ... and slow. How are they so slow at coming to temp? Looking at the logs, they're updating with temp changes a little as 1.5 degree (and less in some instances), so it seems like the sensor itself it that slow:

2019-08-18 17:54:03.865 Motion Sensor 03 temperature is 75.78°F
2019-08-18 17:49:03.140 Motion Sensor 03 temperature is 74.24°F

2019-08-18 18:30:29.773 Motion Sensor 06 temperature is 88.71°F
2019-08-18 18:25:28.704 Motion Sensor 06 temperature is 87.79°F

I may message the seller about the three that are currently aren't working properly (unless his ad stated as is, I need to double-check), but I have ten more to unpack plus a bundle of ten contact sensors also on the way.

For the devices that are “stuck”, I would Remove them from Hubitat via their device pages. Then factory reset the devices via the Iris v2 Motion sensor documentation. Then pair them with Hubitat again and see if they start reporting like the others.

You may also need to "calibrate" the sensors via the Edit Device page. You can enter a reference temperature so the sensor can adjust itself to the value you input, adjusting it's calibration accordingly. Also, you might try doing a factory reset on the sensor.

I don't bother with offsets on temperature sensors until they have "settled in" for a day or two. Any Iris V2 motion sensors, Xiaomi Temp/Humidity sensors, even Samsung ST buttons, have been pretty inaccurate in the beginning but eventually come around.

I didn't expect I would need to do that given motion sensing was working so well, but I gave it a go.

I removed it from Hubitat, reset it (removed battery for ten seconds, held down reset button, kept it held down for two seconds after re-inserting battery), and rediscovered it, at which point the temp was 85.03 ... where it has remained stuck, just at a different value now. This is the one that was previously stuck at 84.12.

I've brought them all back inside and I'm going to reset it again after they've all cooled down. I'm wondering at this point if it will show something in the ballpark of the rest of them, but continue to never report temp updates.

Thanks for the suggestions.

I did that. I gave them an hour to all come to the same temp then set them all to reference temp 72.

I'm surprised they detect dramatic temperature changes so slowly. Maybe I should've waited longer.

As for one of the two that never updated its temp, I removed it, factory reset it, an re-paired it. So far, it's doing the same thing (but now stuck on a different temp).

I'll let them sit overnight and see what happens.

When you say "stuck", what do you mean? The sensor is not going to update within seconds of a temp change. For example, one of my iris v2's took 2 hours to change from 75.44 to 74.54. In general, it will not report a change until it changes by almost a full degree.

Hmm .. okay. I'll give them a bit more time before passing judgement. Thanks.

Right. I anticipated that.

Actually, I anticipated no update until a two degree change occurred (that's what the Go Control does), but the Iris logs showed some updated with less than a one degree change.

I have ten sitting in a row (well, 9 Iris plus one Go Control, given one Iris was somewhat DOA) ... and they went from a room that was 72 degrees to a garage that was 91 degrees. I know they can be slow detecting slight changes, but was surprised how randomly (and slowly) they increased from 72 (after setting reference temp) to 91 degrees, which none of them made it to, but I was going to accept a 2 degree margin of error (two did made it as high as 90.69 and 90.14).

One thing I tried to do before Iris shut down was to update the firmware on as many devices as possible using an Iris Hub that I bought at the beginning of the year for just this purpose. I am wondering if some of your Iris v2 motion sensors might have an older firmware version on them? This is just a hypothesis.

I believe SmartThings will attempt to update the firmware on some Iris v2 devices. If you have an old ST hub laying around, it might be worth a try.

They're not the best temp sensors out there. The drastic a temp change might be tough for them to deal with because that isn't really a normal shift for that quick a period of time. They may have a minimum reporting period. It won't update until it reaches this time minimum, since temps shouldn't change that frequently. See how the devices work in real world changes.

Good question. It doesn't look like it reports version info to Hubitat.

I wonder if I pair one up with my old Wink hub what I'd see ...

I have nine paired with Hubitat and lined up in a row.

While I don't expect all ten to report the same temperature (just be close to one another), I would expect them all to follow the same rules regarding any threshold that needs to be met before reporting a change in temperature, be that lapsed time or temperature change.

However, as @ogiewon suggested, maybe they're each on slightly different firmware. Another possibility, maybe they have other settings Hubitat doesn't access regarding update frequency. I'll perform a factory reset on all ten of them and see if that changes anything.

I am convinced, however, that two are just dead no updating temps (they update motion fine). It's coming up on three hours since I factory reset the one and over eight hours since I paired the othe; temp hasn't updated once (beyond initial pairing) on either of them.

The rest, those that are updating temps, are currently reporting between 67.23 and 73.01 (actual sensor temp, not offset adjusted temp).

This topic was automatically closed 365 days after the last reply. New replies are no longer allowed.