Is my calibration environment good enough?

[HHaje.1]

Hi there,

Following this topic, I've made a box out of black paper, perhaps it work for calibrating xTalk.

We couldn't get the customized paper, but here's the box with side's distance = 25cm.

 

 

 

 

The results of both offset and xTalk are inconsistent, given I'm ranging at night with room's lights off.

Any suggestion to improve the setup to accurately perform both offset and xTalk?

Best Regards,

Hamza Hajeir

[Anne BIGOT]

Hello Hamza,

The light doesn't matter as long as there is no sun light. Only 940nm light makes a difference. So if you have a LED, don't take care of it, it will be fine. If you don't have calibrated paper, mate finished black spry-paint paper it is OK. The trick is to find that distance where the sensor under-ranges by 10 or 15%. If you have a good cover glass, it could be 1 meter or 2. With a bad cover glass, it could be 30cm. The size of the paper has to be at least as large as one half the distance to the sensor. If your paper is large enough it is fine.

You are saying your results are inconsistent. If there are 5% different, it should be fine. Use the worse number. The penalty for too higher numbers is a slight inaccuracy. The penalty for too lower numbers might be a ghost image.

You don't need a box to do the calibration of the cross-talk. You need a larger distant (moreover if you have a good cover glass).

For offset, you need a brighter target. White paper at 140mm works fine.

Best regards

Anne

[HHaje.1]

Hi Anne,

Thank you for the information.

So the first step is to get a matte black spray, then spray out a suitable size paper.

Starting with Offset calibration, then trying to get the under-range distance.

I'll do feedback with results.

BR,

Hamza

[HHaje.1]

UPDATE

Hi @Anne BIGOT​ ,

I've obtained the offset calibration value as (2034).

Regarding the xTalk calibration - after many trials with spraying papers trying to calibrate at a short distance (Up to 50 cm), I've came up with less than 5% error in measurement.

Then I've gone with a large black paper set to a distance of 155 cm apart the sensor, the sensor was ranging 132-140 cm, which corresponds to 6.6-12% error.

At 50cm calibration : the obtained cross talk value is (367);

At 155 cm calibration, it ranges in between (400-503).

Here's a sample log of calibration process (The log reports a measurement twice per second):

Distance(mm): 1338
Distance(mm): 1326
Distance(mm): 1374
Distance(mm): 1360
Distance(mm): 1381
Distance(mm): 1349
Distance(mm): 1317
Distance(mm): 1324
Distance(mm): 1374
Debounced button
        currentXtalk=0  calibratedXtalk=451
Distance(mm): 1546
Distance(mm): 1533
Distance(mm): 1580
Distance(mm): 1454
Distance(mm): 1578
Distance(mm): 1581
Distance(mm): 1552
Distance(mm): 1583
Distance(mm): 1531
Distance(mm): 1495
Distance(mm): 1452
Distance(mm): 1592
Distance(mm): 1545

Is the calibration values reasonable?

Or Is there something strange or wrong along the process?

Thank you,

Hamza Hajeir