This is a slightly off topic question, but how did you know the scale was -/+ 1024?
The configuration parameters and supported command classes of z-wave devices are usually listed in the manual or you can get them from z-wave alliance, but I haven’t seen anything similar for Zigbee devices.
Besides doing a lot of trial and error, how can you determine how a device like the Iris Keypad works? (I’m aware of the existing DTH, I’m just using it as an example because it’s a complex device)
I’ve worked directly with 3-axis accelerometer chip hardware (for an Arduino based project: These chips literally have microscopic “pitchforks” in them that bend super tiny amounts by G-forces !), and the one I used output a value in that degree of precision with that range. It can be configured to be low accuracy but high magnitude (e.g., ±8g) or high accuracy smaller magnitude (±2g).
So I just assumed that the ZigBee cluster allows passed along the raw output (±1024 = 11 bits; but fits in 2 bytes), and let’s the product designer decide what to convert it to.
So if you’re seeing 1024 as straight up or straight down, and never a higher value, then the range is obviously ±1g…
The raw cluster sent during join attempt should specify which ZigBee spec it conforms to (ZigBee HA, or LL, or etc.) along with its list of endpoints, and the standard clusters (ie, “Capabilities”!) it offers. But the manufacturer might take a lot of liberties after that… Just like SmartThings DTH developers do.
There are some folks here who believe that SmartThings “Capabilities” are a risky redundant abstraction, because both ZigBee and Z-Wave already have their own.
But, of course, SmartThings combines ZigBee and Z-Wave (and cloud devices and WiFi, …) into one platform and thus invented a 3rd vocabulary.