Also, before trying the likely difficult process of getting usb power to the sensor, OP can try powering the sensor off an external usb battery and set it to USB power to see if it will work better that way before trying another sensor.
I replaced the rechargeable batteries with the factory-supplied non-rechargeable one. The motion sensor seems to work as expected, as well as tamper detection. However, light detection is unchanged (it went from 10 lux to 11 lux, even in direct sunlight) and UV index remains 0. Something is definitely wrong here.
Even if I were to plug it in to microUSB, that is not realistic as this sensor is supposed to be installed in close enough proximity to a shower that it would not feel safe. There is also tile that should not be disturbed.
How long did you wait? From the manual, you will not get initial readings until the first hour:
If you are having issues with any of the Multisensors sensors not sending data while paired to your current existing network, below are a few types of troubleshooting you can do to bring it back to life. If none of these steps work for you, please get in contact with our support team.
Wait at least 1 hour.
By default, the Multisensor 6 reports all sensor data every 1 hour (which is configurable using Parameter 101-103). Wait 1 hour and see if the values update on your gateway. If you do not see any update to your sensors, please continue onto the next troubleshooting steps.
After that, you will need to change the parameters for more frequent reporting.
Also, even as a lux sensor, it is not intended to be in direct sunlight.
I gave it about 30 hours. The lux readings are creeping up slowly but not anywhere near reality. I am seeing 30-40 lux (currently 43 lux) in a brightly lit room. However, 30-40 lux cannot be accurate, as bright rooms should be a few hundred lux. Outside is anywhere from a few thousand lux to 100,000 lux (if aimed directly at the sun at midday) but the light readings did not go over 40 lux outside. The UV index remained at 0 as well.
The suggestion to change the reporting rate seems to have helped; I set it to report every 8 minutes. However, although it “helped”, the reasons are still way off.
I am pretty much through with this product and just want to try something else. Should I try the Fibraro motion sensor next? Any other suggestions?
I think the Fibaro is a more accurate multisensor than Aeotec, certainly more configurable, so it would be a good next candidate.
Any others? I was reading about the Xiaomi / Aqara devices, specifically:
Although the latter, Mijia Smart Light Sensor, apparently isn’t natively supported with SmartThings, it looks like folks have developed a pretty robust DTH as of now. The former appears only to report lux values when motion is detected; I would be interested in lux values all the time if possible.
Then you will definitely want the Xiaomi Mijia Smart Light Sensor device. Mine has been working flawlessly and has never dropped off my mesh.
Fair warning though, these devices ship from China and take FOREVER to get here. My first one came from Gearbest and that was a 45 day wait. Aliexpress is going on 60 days… I should have bought 2 from Gearbest if I knew Ali took this long, but I wasn’t 100% sure these devices would work.
Alright: Fibaro motion sensor it is, plus two-pack of Xiaomi Mijia Smart Light Sensors in 4-6 weeks. I will post about experiences after more products roll in.
To follow up on the Fibaro Motion Sensor:
- What is the runtime of a non-rechargeable CR123A battery in this device, assuming default parameters and a “normal” room (let’s say a family room)?
- Can it take a rechargeable battery?
- What is the runtime of a rechargeable CR123A battery in this device, same default parameters and normal room type?
Rechargeable batteries should not be used in PIR motion sensors unless the manufacturer recommends a specific one. They tend to lead to a lot of false positives and drive the sensor crazy.
Battery life for the Fibaro model is highly dependent on the parameter settings. The more frequent the reporting interval, the more drain on the batteries.
The Fibaro Motion Sensor’s battery life is approximately 2 years at factory default settings
But as soon as you start changing the settings, you affect the battery life.
I have several of the Zooz 4in1 sensors
ZOOZ Z-Wave Plus 4-in-1 Sensor ZSE40 VER. 2.0 (Motion/Light/Temperature/Humidity) https://www.amazon.com/dp/B01AKSO80O/ref=cm_sw_r_cp_tau_3SL0EbSY0NXXB
I also use rechargeable batteries and have had no false positives. Temperature reporting is disabled to preserve battery life which is approximately 1 month.
My results may differ from others.
I have had the Fibaro Motion Sensor running for the last few hours.
Temperature and lux readings are fine. I put the sensor behind a book in a dark hallway and the readings dropped to 1 lux. In a windowed bathroom the lux readings are consistent with expectations: 65 to 373 lux.
However, the motion sensor is not performing anything like what one would expect. There are delays of up to 6 minutes for motion to be registered. Sometimes it will register motion immediately; other times it takes forever or does not trigger at all. I even picked up the sensor and shook it around about 15 minutes ago: the tamper alert went off but the motion sensor did not! Crazy!
Motion detection sensitivity (parameter 1) is set to “High sensitivity”. Motion detection - operating mode (parameter 8) is set to “Always Active”.
Color me unimpressed…
I know it may seem strange, but picking it up doesn’t have anything to do with motion detection for this device. It’s a PIR sensor for that. So “motion“ is actually a change in temperature passing across the field of the sensor. Once you pick it up, the heat of your hand is going to be essentially equal no matter what you do next. So no change, no report from the sensor.
The following FAQ explains The detection field for this kind of sensor:
As far as the variable Behaviour you’re seeing…What are the reporting interval settings?
Parameters 2, 3, and 4.
The “always active“ parameter just means it will report during both daytime and night time. It doesn’t change the blind interval.
2: Motion detection - blind time: 8 s
3: not shown - so it must be the default
4: not shown - so it must be the default
I am using the SmartThings Classic app with the v2 hub.
As far as picking it up: When I picked it up, I approached it from the front. While also in the bathroom I was in there definitely within view of the sensor doing various things for several minutes.
For what it’s worth, the LED blinked various colors (I recall mostly green) occasionally, including when I first entered, which indicated that the sensor was picking up something. However, while the sensor continued to report lux changes and even tamper detections, it did not report motion events. It is unlikely that there was a communication problem between the sensor and the hub.
Approaching from the front may not trigger it. Again, see the discussion of “angle of incidence“ in the FAQ linked to above.
When you say the other parameters are “not shown“, not shown where? Did you use the Z wave tweaker to expose the parameters or something else?
At the default settings, that sensor will only report if it detects two separate movements across the detection field in a period of 12 seconds.
Once it detects motion, it will then go quiet For 8 seconds before starting the next 12 second interval.
Parameter 80 determines what the flashes mean when motion is detected, so it would be good to have that value also.
But the main thing is even though it’s an eye, don’t think of it as a camera that’s looking at you. You should expect it to detect you when you walk crosswise to it, as explained in the FAQ.
Do you have a good understanding of how the Samsung SmartThings Motion Sensor (latest version, 2018) works? Is it similar?
I am thinking I should put up that one right next to the Fibaro Motion Sensor to compare results.
Yes, all PIR sensors work the same way. The other common technology for fairly inexpensive motion sensors is the one in cameras, but those aren’t used in bathrooms, obviously. Good for baby monitors and outdoor locations, though.
Anyway putting the two side by side would definitely be an interesting test.
Indeed, I have been doing this for the past week: put both the Fibaro Motion Sensor and Samsung SmartThings Motion Sensor side-by-side to see the results.
I would rather not post pictures of my bathroom right now but I can report the following observations.
I tried the sensors in the following locations:
- on a mini-shelf on top of a toilet paper roll holder, aimed directly at the bathroom door
- in a corner with the shower (aimed 45 degrees downward)
- on the ceiling, above the glass/metal divider between the shower and the rest of the room
- on the wall, above the glass/metal divider between the shower and the rest of the room
- on the bathroom vanity counter
This bathroom has a bath/shower with a sliding glass door between the shower and the rest of the bathroom. The sliding glass door has a metal frame on top and an additional void space of about 18 inches between the frame and the ceiling. The shower side is tiled while the rest of the bathroom is painted drywall. Opposite the shower is a vanity that occupies the entire east side of the room (the vanity could hold three sinks, but there is only one sink close to the door). Above the vanity is
a wall-to-wall mirror that goes up about 3 feet from the vanity; above the mirror is a window that goes to the ceiling (about 2 and a half feet high, 5 feet wide). The toilet is on the south side, and the bathroom door is on the north side right next to the shower door.
There are two sets of Lutron light switches: one is a dimmer (PD-5NE) and one is a switch (PD-6ANS). The switches are in the same 2-gang bank. The 2-gang bank is right next to the bathroom door, above the sink on the north wall. To test activation, I set SmartThings automations that automatically turned the lights on or off when each sensor was activated.
Overall, I am much more pleased with the Samsung SmartThings Motion Sensor (“SST”) than the Fibaro Motion Sensor (“Fib”). The SST activated very consistently and almost instantly as one would expect. In fact, when pointed directly at the door from the toilet paper holder mini-shelf, the SST activated its programmed lights before I came into view of the sensor. The motion of the door was sufficient to activate the SST (or the SST is so sensitive that it could detect my body through the door, which seems unlikely). In contrast, sometimes the Fibaro would activate only after I opened the door and set foot in the bathroom; other times the Fibaro would not activate at all unless I went right up to it next to the toilet.
Neither sensor performed at all through the sliding glass door between the shower and the rest of the bathroom. When I stuck my hand through the sliding glass door and waved it around, the SST activated with less body parts exposed. The Fibaro required a significant amount of arm. When I was down on the floor, I could see the sensors quite clearly (when mounted on the ceilings and walls) in the east wall mirror, but the sensors did not activate. Once I stuck my hand up beyond the threshold of the metal frame for the sliding glass door, however, the sensors activated. Again, the SST activated almost immediately while the Fibaro required waving around a bit.
In this particular environment, the SmartThings Motion Sensor is the clear winner to me: it just works. The shower door test was disappointing though because one reason why I replaced the Acuity Brands WSD 2P WH switch on the wall with this new setup is to make sure that the lights do not turn off while someone is taking a shower. This test made it pretty clear that PIR sensors do not work through glass, so either the sensor needs to be mounted on the ceiling above the metal frame (so it can capture both sides), or one needs two sensors.
I thought that maybe the SST had an advantage over the Fib because there is a SmartThings Outlet (2018) on the other side of the wall that might repeat the Zigbee signal more consistently and reliably. However, I unplugged the outlet and experienced similar results, so Zigbee vs. Z-Wave probably does not make a difference in this case.
Also: do not use packaging tape for these tests, as it left a colored residue on the drywall and tile that now I have to figure out how to remove. I probably should have used painters tape instead.
To solve your shower issue, you could also put a door contact sensor on the bathroom door and setup an automation with a ‘latch’
Basically, if the bathroom door is closed and motion happens inside the room while the door is closed - it stands that the light should stay on until the door opens again. (this BTW is the exact use case for why I use the rooms manager code linked above). You can do this with webcore, or a careful combination of virtual switches and builtin automation.