VL53L0X Sensor Distance Reading Decreases as Physical Distance Increases

[FPayn.1]

https://www.fpaynter.com/2020/10/left-side-wall-tracking-success-with-vl53l0x-array/I have three VL53L0X sensors mounted in a linear array on a wheeled robot, pointing broadside to the direction of travel. In a recent test where the robot moved slowly away from a wall, two of the three (front and rear) sensors report reasonable increasing values, but the center distance measurement increased very slowly and then started to decrease as shown in the following Excel plot

Here's a link to a short video showing the physical track.

So, why is the reported distance from the center sensor going the wrong way?

All three sensors are set up for a 20 mSec timing budget and continuous readings. I have a Teensy 3.5 MCU reading all six sensors (3 on left, 3 on right) as fast as possible, and reporting results over I2C to another MCU (Arduino MEGA) that prints the reported values to the serial monitor.

If you look at this post, you'll see that it appeared to work fine two days ago. The Excel plot in the post is from that same center sensor.

Frank

[FPayn.1]

I fiddled with the connections some and the left-side array seemed to behave properly. So, I made another wall-tracking run and this time it intercepted and tracked the desired (300 mm) target offset, with believable results from all three left-side VL53L0X sensors, as shown in the following Excel plot

Still, this whole episode with the center sensor reporting non-physical distances has me more than a little buffaloed. Any thoughts would be appreciated.

Frank