Constant short reading after adding 2 mm bandpass filter – VL6180X & VL53L4CD

[dariodelbono]

Hello,

we are using VL6180X and VL53L4CD sensors in an outdoor setup and we have issues with sunlight (even reflected light).

To improve sunlight rejection, we added Thorlabs bandpass filters:

• FBH850-10 for VL6180X (850 nm)

• FBH940-10 for VL53L4CD (940 nm)

Both filters are 2 mm thick.
The filter is placed ~1 mm in front of the sensor window.

We have four identical sensors per type; only one has the filter.

Observations:

• Sensors without filter measure correctly (e.g. 143 mm → 143 mm, 70 mm → 70 mm).

• The sensor with the filter almost always reports a short distance (~20–30 mm).

• Even when the real target distance changes significantly, the filtered sensor reading remains almost constant.

This looks like strong optical crosstalk rather than noise.

Questions:

1. Is a 2 mm interference filter likely to introduce severe crosstalk?

2. Could the 1 mm air gap be the main cause?

3. Would a much thinner IR-pass filter be more appropriate?

4. Can the crosstalk calibration compensate for this configuration?

Any guidance on proper external filter integration for outdoor use would be greatly appreciated.

Thank you.

[Zhiyuan.Han]

Hi  

The VL6180X and VL53L4CD really don't work very well outdoors. The issue is that the laser light is really faint. We do this for eye-safety and power constraints. and there are 850nm and 940nm light from the sun. The sun acts as noise to TOF sensor, with sunlight level increase (indoor to out door), the signal to noise ratio becomes unfavorable. so sensor can't get proper ranging data.

and inside VL6180X and VL53L4CD, there are already bandpass filter inside, so even through you add external band pass filter, it won't help much on the outdoor condition.

Back to your question, 2mm thick filter and 1mm airgap can both contribute to the crosstalk, with thinner filter and shorter airgap could help on the crosstalk reducing but can't remove it totally.  normally crosstalk calibration will help on your case, you can try to do a crosstalk calibration to see if help.

Br

Zhiyuan.Han 

[dariodelbono]

Hi Zhiyuan,

Thank you for the explanation.

In our case, we do not have direct sunlight entering the sensor. The issue is mainly sunlight reflected by a nearby surface, which increases the background noise seen by the receiver.

We are considering adding a small conical baffle only on the receiver side (not on the emitter). The cone would be just slightly larger than the laser emission cone as seen from the receiver, in order to limit the acceptance of off-axis reflected sunlight while preserving the useful return signal.

In your opinion, could this kind of geometric shielding provide a meaningful improvement in this scenario, or would the reflected solar background still dominate?

Best regards,
Dario Delbono

[HERESHENRY]

This looks like classic optical crosstalk. Adding a thick (2 mm) filter with an air gap likely causes internal reflections, so the sensor “sees” a constant short distance. Also, as mentioned, these sensors already have built-in filtering and struggle outdoors due to low signal vs sunlight noise.

You might get slight improvement with a thinner filter, reduced gap, or calibration but fundamentally this sensor isn’t ideal for outdoor/high ambient light use.

[John_Kvam · Accepted solution]

Your baffle is a good idea, and it will cut down on the reflected sunlight, but it has drawbacks. Any time you limit the field of view, you are going to get less of the transmitted light you want. It's a trade-off. But do give it a try because I've seen it work. Just don't make the opening too small. 

And another note. It's tempting to make such a baffle out of 3D printer plastic. And most of those plastics are translucent to 940 nm light at best. If you try it, print the baffle, then paint it with flat black spray paint. Paint is opaque to 940nm light. 

Good luck, 

- john