Gadgetoid
commented
3 years ago Accuracy/repeatability varies by distance, ambient light and a bunch of other factors, depends upon long/short distance mode, etc.
I'd recommend looking over sections 3.4, 3.5 and 3.6 of the datasheet to get an idea of what to expect - https://www.st.com/resource/en/datasheet/vl53l1x.pdf
There should probably be a way to get the measurement error code, since many of them shed light on configuration/environmental errors. This guide mentions some range statuses which I think are converted to error codes by this library: https://www.st.com/resource/en/user_manual/um2510-a-guide-to-using-the-vl53l1x-ultra-lite-driver-stmicroelectronics.pdf
There are five range statuses: 0, 1, 2, 4, and 7. When the range status is 0, there is no error. Range status 1 and 2 are error warnings while range status 4 and 7 are errors. When the range status is 1, there is a sigma failure. This means that the repeatability or standard deviation of the measurement is bad due to a decreasing signal noise ratio. Increasing the timing budget can improve the standard deviation and avoid a range status 1. When the range status is 2, there is a signal failure. This means that the return signal is too week to return a good answer. The reason is because the target is too far, or the target is not reflective enough, or the target is too small. Increasing the timing buget might help, but there may simply be no target available. When the range status is 4, the sensor is "out of bounds". This means that the sensor is ranging in a “non- appropriated” zone and the measured result may be inconsistent. This status is considered as a warning but, in general, it happens when a target is at the maximum distance possible from the sensor, i.e. around 5 m. However, this is only for very bright targets. Range status 7 is called "wraparound". This situation may occur when the target is very reflective and the distance to the target/sensor is longer than the physical limited distance measurable by the sensor. Such distances include approximately 5 m when the senor is in Long distance mode and approximately 1.3 m when the sensor is in Short distance mode. Example: a traffic sign located at 6 m can be seen by the sensor and returns a range of 1 m. This is due to “radar aliasing”: if only an approximate distance is required, we may add 6 m to the distance returned. However, that is a very approximate estimation.
But ultimately it's hard to pin down why this might be happening
Does the signal return to the previous value if you leave it long enough. IE: does it look like sample averaging? (from your paste, it doesn't seem so)
Hi all,
I just try to hook up a VL53L1X to a Raspberry Pi and it initially worked very well.
However I encountered a problem (see images):
Once I start a measurement (e.g. every second), I seem to get reasonable readings.
When I then put my hand into the beam, I also seem to measure the correct distance.
However, when I pull my hand back out, the reading doesn't go back to the intial value.
--> Could somebody help me out here? Any hints on why this is happening? Could this e.g. be caused by a missing pull-up resistor?
When I close() and open() the sensor again, everything is "back to normal"...