Built-in Zooz 4-in-1 Sensor driver reporting incorrect luminance values

Of interest to @mike.maxwell and @agnes.zooz I think...

The current built-in Zooz 4-in-1 Sensor driver is not decoding/reporting luminance correctly.

The Zooz 4-in-1 is documented as reporting luminance as a percentage with a range of 0-100. The built-in driver is reporting luminance as lux values of 0-50. It's unclear how the lux values are being determined, but it doesn't appear to be simply dividing by two. Regardless, full daylight is not 50lux.

Following are some examples with the value in the Multilevel Sensor Report, the % value intended by the sensor, and the value being reported by the driver:

0x2710 100.00%  50
0x25F9  97.21%  47
0x18B4  63.24%  33
0x0CCB  32.75%  18
0x0654  16.20%   9
0x03AC   9.40%   4
0x015C   3.48%   2

The driver does report report temperature and humidity correctly, which are encoded in the same manner.

If I remember correctly there is an old post that discusses this and the fact that reporting lux as a percentage has no meaning because it's reporting a percentage of what maximum? I don't remember if there was any resolution but I know @mike.maxwell got involved.

EDIT; found it!

I believe it is based on the old ST DTH by @krlaframboise, which uses a ranged multiplier convert the percentage to a lx between 0 and 50
SmartThings/zooz-4-in-1-sensor.groovy at master ยท krlaframboise/SmartThings ยท GitHub

1 Like

I spent a lot of time comparing the % being reported by the zooz 4-in-1 against the lux values being reported by a Dome sensor and Fibaro sensor sitting next to it and it would always max out at 100% way before the other sensors reached 100 lux.

It really doesn't matter how the driver handles the conversion because maxing out below 100 lux renders it practically useless as a light sensor...

3 Likes

Interesting. I would love to know where getLxConversionData() data came from. I reached out to the author to ask.

There's a great quote from the author saying "One of the Zooz engineers came up with a conversion chart that went up to 100lux, but based on my testing itโ€™s way off." I found that table, which which indicates that the sensor maxes at 80 lux.

Sigh. Further searching brought me to the discussion below. It appears that I'm not the first person to identify this issue, and @mike.maxwell has already indicated that the driver will not be changed. Sorry to have wasted everyone's time.

And special apologies for not recognizing you as the developer... :slight_smile:

These sensors work OK for light vs dark detection, even if it's not on a regular scale. They're not good for detecting small differences in light level.

The bigger issue now is that they pair in S0 mode on a C7. The batteries will die quickly and negatively affect the Z-Wave mesh. You can pair them unsecured via a USB secondary controller and PC Controller software.

Yes, I have both of mine paired insecurely. One on a C7 and one on a C5.