VL53L0X - What is the realtime accuracy ?

[KVenk.1]

I recently bought X-NUCLEO-53L0A1 to evaluate the distance sensor. My application requires accuracy of 1mm at the least.

PFA video file showing my evaluation. Pls open the file with appropriate media player.

As you can see, there is 2 to 3mm fluctuation in the distance measurement in HIGH ACCURACY mode. The fluctuation is on both -ve and +ve sides. It is not predictable.

I wanted to know whether it is possible to achieve 1mm accuracy using this sensor in realtime scenario (not in perfect lab environment) ?

[John E KVAM · Accepted solution]

The speed of light makes a 1mm round trip in 6.7 picoseconds. To resolve this one would need to measure at 300 GHz, or about 100 times faster than your PC. That's not happening in a few dollar sensor. The sensor gets it's range by using statistics and lots of measurements. If you increase the timing budget you will get a more accurate number.

In general, 4X timing budget gives 2X more accuracy.

But there are limits.

If you have the time, and the power, you can increase the timing budget to a half second, or maybe even a whole second.

If you change the reflectivity of your target you will never get the accuracy you desire, but if you are always looking at the same target, you will do better.

That being said, the sensor is rated at +/- 3% which is quite a bit more than the 1mm you wanted.

Another thing you can do is make absolutely sure the target object covers the entire field of view. the light goes out at 27 degrees, so you need a target with a diameter of 1/2 the distance to the sensor. (So if at 14cm, the diameter has to be >7cm.) And make sure the sensor is not seeing the table.

• john

[KVenk.1]

Makes sense John, Thanks for your thoughts.

[Eleon BORLINI]

Hi @Community member​ ,

can you please mark John's answer as Best answer, if it was useful for your needs?

Just for helping future users' search in case they would have the same concern.

Tks,

-Eleon