Evaluating the quality of a cover window by measuring crosstalk

[mf101]

In the production process of a product using the VL53L1X TOF sensor we would like to evaluate the quality of the custom assembled window. There are some mechanical tolerances and we need to ensure that the crosstalk is as low as possible. Is there any way to measure the crosstalk (in kcps) in a test setup ?

In AN5231 chapter 7 a typical crosstalk value of 0.2kcps/SPDAD is given for the hornix cover window. How has this value been measured ?

[John E KVAM]

In the API that ST provides is a crosstalk calibration function.

We generally expected that customers would build their system, use the crosstalk calibration function, extract that cal data and store it on their MCU. Then each time the system is booted they would load that cal data into the sensor and begin ranging.

A major cell phone company used the P-Nucleo-53L1A1 as an incomming test fixture to test the quality of the glass.

By using the crosstalk function in the Graphical User Interface for that device, they could prove that their glass was consistent from lot to lot.

You could do the same.

If you can find a way to place your assemply on top of the P-Nucleo-53L1A1, it would do the job.

This is exactly what horix did.

Or, if you had some way to extract the data from your unit, you could run the Crosstalk calibation during manufacture and examine the crosstalk that way.

• john

[mf101]

Thanks for your response. We are aware of this function, but unfortunately our cover glass setup is so good that we cannot determine any inflection point of the curve within the range of our test setup (max. 2m distance). Therefore we are not able to determine the xtalk calibration distance (xcd), which is required as a parameter for the function VL53l1X_CalibrateXtalk().

To be precise about our intention: We are not looking for a method to qualify the general construction of the cover window, but want to qualify each of our individual product assemblies during the production process. In the ideal case we would like to have a numeric value indicating the quality, lets assume by placing a reflector at a fixed distance. We already thought of using the signal kcps value in such a setup, as we know that for higher crosstalk it will be higher (and the actual distance measurement will be lower -> under-ranging). Unfortunately this value seems to fluctuate quite a lot even if the distance measurement is stable (no moving target).

Do you have any suggestions ? We would highly appreciate any guidance.