VL53L1X Low Range issue

[spiralbrain]

We use the VL53L1X in one of our products and have been doing that for over 2+ years now. As this device is not easily solderable, we have been relying on the Chinese supply of breakout boards used in the hobby market. 

Just recently we are facing range issues with entire lots of this sensor based boards. All of them exhibit range issues where the Sensor is unable to detect beyond 30 to 40cm. I thought it could be a calibration or offset issue. I've tried getting offset values using the 17% Gray sheets as recommended. Any help in this regard will be appreciated. Thanks.

[John E KVAM]

I've seen this before. Someone forgot to take the protective liner off the sensor.

We put that bit of tape over the sensor to prevent debris from getting into the sensor. And unfortunately, it's almost transparent. 

You would not be the first. A cellphone vendor did this and had to take apart 100 phones. 

what is happening is that the liner is causing a huge number of photons to be reflected back into the sensor. With a million photons at zero-distance, all your numbers are low. 

Please check it carefully.

- john

[spiralbrain]

Hi John,

Thank you for a quick reply. I did check and the Kapton tape protection is not on, I've verified that. 

The difference I've noticed is that now the tape is removed by who ever is assembling the breakout boards. Many people must have faced this issue in the past. 

Thanks,

-raman

[John E KVAM]

There are three ways the sensor can range 'short'. One is offset - where every range is short (or long) by some constant value. 

The other is a crosstalk issue where some target is at about the right distance when the target is close, but as the target gets farther and farther away, the distance gets shorter and shorter. 

The third thing that can happen is that the light gets obstructed.

On some of the VL53 sensors debris gets into either the transmit hole or receive hole.

This cannot happen on the VL53L1 due to the external filter. But if the filter is covered by flux residue, it will behave the same way.

So do you have an offset issue, a crosstalk issue or and obstruction issue?

And once we know what the issue is, we can find a solution.

- john

[spiralbrain]

I think that's the issue. Is there a way to remove crosstalk calibration completely as there is no glass in front?

Are there any default values that could be tried to see if there is any change? I am very sure its the calibration or the offset but setting it does not make a change but the range reduces further.

Its not obstruction for sure. I have checked and both the glass are clean.

[John E KVAM]

Is there a way to remove crosstalk calibration completely as there is no glass in front?

With no glass, you should have no crosstalk. and by default, there is NO crosstalk set. 

Find a target where the range is 10-20% short and find the real distance. 

Then do the crosstalk calibration.

That will tell you how bad things are. 

Then try cleaning the sensor.

Try a puff of air. Try a wipe with a soft cloth. Try a SMALL amount of water. 

Harsh chemicals (Isopropyl alcohol) can melt the plastic, so be careful.

Then run the crosstalk again. 

See if it improves. 

Then you need to taik to your manufacturer and ask them when the pulled the liner. 

Taking off the liner causes the condition which leads to your poor ranging results.

- john

[spiralbrain]

Thanks for your help John.

I went through various of your replies to people having issues with this sensor and decided to try recalibration with ST's software. I hooked it up to a Nucleo board and recalibrated in a dark box with the sensor at 140mm from a 17% gray sheet as recommended in the manual. 

Please advise on the correct calibration procedure in case I am doing something incorrectly. I have attached photos of the setup and software screenshot.

What I have noticed is that that the sensor can detect a white sheet at 50cm in Short mode, 30 cm in Medium Mode and 18cm in Long mode. I find this very strange, perhaps this could shed some light on what's going on?

Q: Does the sensor need a calibration in 3 modes separately?

Could these sensors be damaged during reflow? Is there an onboard firmware which is playing its part? 

It worries me because we bought these breakout boards from 2 different vendors and all exhibit the same problem. I suspect there could be thousands of boards with similar issues.

Thanks again.

[John E KVAM]

Use white paper at 140 cm to do the offset correction. 

Use the dark gray or black to do the crosstalk correction. Distance is highly dependent on your coverglass. Do some experiments - range until you find that place where you are under-ranging by 10 or 20%. Then do the crosstalk from that distance.

In your case without a coverglass you should not have anywhere you are under-ranging. If you do, your sensor has an issue. 

If you cannot range beyond 40cm, you either still have the Kapton tape on, or there is debris on the filter. 

Double check for the tape. 

Do the calibration, then clean your filter. use a gentle chemical. Just water, maybe a little alcohol in it. Maybe some ammonia, but go easy. The do the calibration again. 

Does anything change?

- john 

[spiralbrain]

Tried all that you suggested.

The Range is still the same, no improvement. Found one in the lot with the KAPTON still on, removed it and this again detects a white sheet up to 60cm in short mode, 33cm in medium mode and  17cm in Long mode. So its not the glass or the tape.

Do you think the entire lot is faulty? Uncalibrated or damaged laser?

I also suspect the voltage regulator and the associated caps. Can you advise if the power to the laser is different in the three modes?

[John E KVAM]

We can assume something is wrong. So, let's do an experiment. 

Take a good sensor and range on something. 

get the signal strength and number of SPADs.

Then take a 'bad' sensor and do the same thing in EXACTLY the same.

If you conclude you are getting 20M photons and using 62 SPADS on a good sensor and are only getting 10M photons using 162 spads then we know that there is not as much light returning from the target.

Two ways to make this happen. A blockage on the transmit side, or a blockage on the Receive side. In a naked sensor this is hard to do. With a coverglass or with an obstructed housing it's easy. 

Is your sensor installed - or out in the open?

in your picture, could it be that the black object be covering part of the Field of view?

Re-adjust the sensor so that black bit is well away from the sensor. The light goes out in a code of 27 degrees, but there is roll-off at the edges. Perhaps that holder is reflecting and making a mess. 

- john