cancel
Showing results for 
Search instead for 
Did you mean: 

How do the TOF sensors distinguish between their own bounce and another sensor's [optical] ping?

edsut
Associate

I saw somewhere in the forum that the TOF sensors see [optical] pings from other sensors as ambient light (meaning they are able to ignore them) but are still able to detect their own bounce. Is that true? If yes, can someone explain how that works?

Note, I'm referring to the situation where you have independent devices in an area moving around, each of which have their own TOF sensor(s) for obstacle avoidance. The point here is that there is no synchronization between each device, so there would surely be overlapping [optical] pings.

(I accidentally posted this to the wrong forum group, so this is a re-post of the same question)

 

1 ACCEPTED SOLUTION

Accepted Solutions
John E KVAM
ST Employee

I tell people that the sensor emits the light, and stops the stopwatch when the light returns. While true, this is an over-simplification of what is truely going on. The light pulses - maybe a million times or so - and the sensor is almost continously collecting photons. So just prior to the light turning on it's known how much ambient light there is. So that amount of light can be subtracted and what is left is signal over ambient. 

Another thing going on is the clocks aren't running at precicely the same rate. This is accounted for inside the sensor, but externally that phase shift averages out, and you are left with the right answer.

Oddly we found this out by experiment. We fully expected this to be an issue. It just isn't. Unless of course you point them directly at each other at a close distance.

Also by experiment, one will get the occational bad status when facing another sensor. The sensor just cannot figure out what's going on, and returns a error status. It is assumed that the clocks line up particularly poorly during this collect. Appears to be a fleeting issue, and the sensor is fine on the next collect.

 


If this or any post solves your issue, please mark them as 'Accept as Solution' It really helps. And if you notice anything wrong do not hesitate to 'Report Inappropriate Content'. Someone will review it.

View solution in original post

2 REPLIES 2
AlexCloned
Senior

I'm working with multiple VL53L5CX sensors, spaced just a few millimeters apart within the same setup. There are situations in which the light emitted by two or more of these sensors hits the same object, and I can tell you that accuracy and measurements are not affected. I reccomend you do your own experiments about this. I've only noticed a variable increase in ambient light noise measurements (in L5 you get one ambient measurement by each zone!, plus signal strength and sigma, if enabled) when receiving laser reflections from other sensors, with measurement performance remaining virtually unchanged.

Let's see what the ST team has to say.

I understand that each sensor has its own clock and an intelligent clock strategy for triggering pulses with its VCSEL, possibly with pseudo-random periods. Thousands of pulses per second. Considering that the travel times are extremely short (the time from when a photon is emitted to when it's received in the diode-array), coincidences between measurements from different sensors are so improbable that these noise reflections will be easily filtered by the histogram system. It is, I assume, the same way that other ambient light noise is handled, primarily coming from the sun. It's worth noting that you should consider the influence of direct or indirect sunlight, which can potentially completely invalidate the measurements because of its strenght compared to that from other sensors.

John E KVAM
ST Employee

I tell people that the sensor emits the light, and stops the stopwatch when the light returns. While true, this is an over-simplification of what is truely going on. The light pulses - maybe a million times or so - and the sensor is almost continously collecting photons. So just prior to the light turning on it's known how much ambient light there is. So that amount of light can be subtracted and what is left is signal over ambient. 

Another thing going on is the clocks aren't running at precicely the same rate. This is accounted for inside the sensor, but externally that phase shift averages out, and you are left with the right answer.

Oddly we found this out by experiment. We fully expected this to be an issue. It just isn't. Unless of course you point them directly at each other at a close distance.

Also by experiment, one will get the occational bad status when facing another sensor. The sensor just cannot figure out what's going on, and returns a error status. It is assumed that the clocks line up particularly poorly during this collect. Appears to be a fleeting issue, and the sensor is fine on the next collect.

 


If this or any post solves your issue, please mark them as 'Accept as Solution' It really helps. And if you notice anything wrong do not hesitate to 'Report Inappropriate Content'. Someone will review it.