VL53L1CB behind 0.7 mm cover glass – reliable 2–3 m person detection?

Question

Hello,I am designing a wall-mounted touchscreen panel using a VL53L1CB for wake-up and person/presence detection at approximately 2–3 metres. High absolute distance accuracy is not required, but detection must be reliable without frequent false triggers.The sensor will be mounted behind a single 0.70 mm cover glass. I have read AN5231 Rev. 2 and UM2133.The proposed optical construction is:Single-layer 0.70 mm cover glass	Sensor mounted parallel to the glass	Air gap kept at or below approximately 0.30 mm, if mechanically possible	Full-height black gasket/light barrier between the Tx and Rx sides	Separate clear apertures for Tx and Rx in the black printing/coating	Target transmission above 87% at 940 nm and haze below 1% at 940 nm	Crosstalk and offset calibration performed after final assemblyThe first production batch will consist of 15 mechanically identical units.I would appreciate clarification on the following points:AN5231 recommends a cover-glass thickness below 0.6 mm, an air gap below 0.4 mm and a combined thickness of no more than 1.0 mm for fast and accurate ranging above 2 metres. Since my glass is 0.70 mm but accurate distance measurement is not required, is reliable person detection at 2–3 metres still realistic with an air gap of no more than 0.30 mm, a gasket and proper calibration?	Table 3 in AN5231 gives a maximum air gap of 1.0 mm with a gasket for the VL53L1CB, while the long-range guideline gives a maximum combined glass and air-gap thickness of 1.0 mm. Which limit should be followed for this application?	Should the gasket completely fill the air gap and form an opaque barrier between Tx and Rx all the way up to the cover glass? Are there recommended gasket materials?	Is ST’s cover-glass exclusion-zone spreadsheet fully applicable to the VL53L1CB? Are two separate Tx/Rx apertures preferred over one oval aperture for this application?	For a batch of only 15 identical assemblies, should crosstalk and offset calibration be performed individually on every unit, or may validated crosstalk calibration data be reused when the glass and mechanical tolerances are consistent?	Which ranging mode, ROI and timing-budget settings would you recommend as a starting point for detecting a person indoors at 2–3 metres?The panel will normally be used indoors, although daylight may enter through nearby windows.Thank you for your assistance.

Zhiyuan.Han · Answer

HiOverall VL53L1CB ranging performance will be affected by CG if the CG and airgap is not well controlled, and if your design well controlled CG quality and airgap, then it has minor impact onranging capability.Back to your question your 0.7mm thickness and 0.3mm air gap combination should be no big issue. but to make sure that the transmittance andHaze value meet the target.transmittance andHaze are the key parameter for the CG itself.Regarding to the gasket, it’sbetter to block TX and RX totally, “completely fill the air gap and form an opaque barrier between Tx and Rx all the way up to the cover glass”. be careful,the barrier doesn’tapply to much press force to the TOF top surface. and about the material, sorry we can’t help here, as they are the key confidential information for Cover Glass vendor.ST’s cover-glass exclusion-zone spreadsheet fully applicable to the VL53L1CB? About the shape, the separate Tx/Rx apertures relative better than the oval aperture, but they are both fine as we have both a lot ofsuccessful stories with these two designs.About the calibration, we suggest to part-to-part calibration at beginning, if all the data are aligned, then may consider the generic xtalk method.At last, I suggest you use the standard ranging mode with longest timing budget,increase timing budget will help sensor to ranging further.BrZhiyuan.Han

Sign up

Click below to sign in

Login to the community

Click below to sign in