VL53L1X FOV not operating as per specification

[ALaga.1]

Hi, We have been testing the VL53L1X for a while. The primary intend is to measure the occupation a container (30cmx30cmx80cm) filled by small objects (10cmx10cm). We would like to limit the number of sensors to have a low energy consumption (system is on battery). We are seeing some issues that we cannot seem to figure out with FOV as per spec (FOV 27°). Hoping for any insights.

Premise.

1. We are using Arduino platform for testing.
2. We have tried multiple modules & libraries (sparkfun, polulu, etc)
3. Test involves a container filled with 2 objects (say 10cm size), and sensors are placed On top of the container at 20cm from the surface Of the object (30cm from the bottom of the container). Readings are taken every few seconds.

Concerns.

The object is only visible in the line of sight of site of the sensor. The FOV is not operating as expected at 27° as mentioned in the specification. The FOV (8cm circle at 30cm distance) is confirmed by an infrared camera not being a 27° FOV. 

Différent S/W configuration has been applied not able to change the FOV 

Thanks for your support

[John E KVAM]

If you have a infrared camera, try shining the sensor at a very traslucent bit of flim at some distance - say 30cm. Then take a picture of the back of the film. In our experiments we get the 27 degree cone.

But let us try to make this work...

We have a circular cone of light. But we have a square photon detector (SPAD) array.

So it might be more realistic to consider the FoV as a 20degree by 20degree square FoV.

Next, consider the boundry conditions. One cannot realisticly expect our sensor to trigger if, say, 1 square micron of the FoV were occupied. So your object has to be at least a bit inside FoV before we can get enough photons to detect it.

Then consider the reflectivity of the object. The sensor needs a number of photons. A very reflective surface will generate these more effectively than a black sheet of rubber say. So the distance into the FoV we need to 'see' an object does depend on the reflectivity.

Next, consider the floor of the area you are looking down on. It returns photons too. If it's fairly reflective, you target box is going to need more photons to overcome the fact that we average all the photons we get back. A dull black finish on the floor would help a lot here.

lastly if you still cannot get the coverage you want, read the data brief on the VL53L5. It's due out July 1, 2021, and it has a MUCH wider Field of View and can generate distances in either a 8x8 grid (at 15 frames per second) or 4x4 at 60 frames a second.

(it's big downside is that it's a lot more $$ and it uses a lot more power.)

Find it at ST.com/TimeOfFlight