how to calibration 6180

[Zwf.1]

hi，st

I am using a 6180 to measure the height of milk. The height of vl6180 from the milk is about 10 - 60mm，anf the front of the sensor is covered with a layer of clear plastic. I don't know how to do the offset calibration.

1. Remove the covering plastic for calibration or bring the plastic on for calibration?

2. The calibration process on the specification uses a 88% reflectance target of 50mm. Can the value of 50mm be changed? Can it be calibrated directly with milk

3. Does the cross talk calibration have to use a black target 100mm for calibration?

thanks �?

​

​

[John E KVAM]

Offset calibration is easy enough but start by reading the section in the datasheeet. As this summary leaves out some details.

"Position a white target (88% reflectance(g)) at a distance of 50mm (using scaling factor 1) from the top of the cover glass. "

White printer paper is close enough to 88%, and it's cheap. I'd glue a couple of sheets on a backer as it can be a little translucent. Target diameter is at least 1/2 the height so at least 25mm, but there is no harm in making it larger.

So take a bunch of measurements, and average them. The difference between your average and the right answer (50mm) is your offset. Store that in your non-volatile memory.

1) The presence of the coverglass on not does not matter. The brightness of the target will overcome any effects of the coverglass.

2) I would not use milk. I've done studies. Milk is more transparent than you might think. It's mostly water and the photons penetrate it. And it is this fact that is going to give you some issues.

3) When doing crosstalk calibration you want to find a target and location where you under-range by 10 or 20 percent. This point is easier to find with a dark gray target. I'd use a flat spray paint, but I don't have a specification. It's going to take some experimentation. And the distance depends on the quality of the coverglass, its thickness, and the air gap between the sensor and the glass. Remember the diameter of the target must be at least 1/2 the target distance.

So I can save you some work. The photons are going to hit off the bottom and if they reflect back they will make a mess of your result. You can fix this by using a black bottom container. But I'm going to guess that your product designer will not let you use black. Second choice is to use a transparent plastic (PETE works for me) and a flat (as opposed to gloss) black surface on which the container sits. This will limit the number of photons that penetrate the milk.

[Zwf.1]

I found that when calibrating at the same height, different reflective surfaces have different calibration offsets. For example, using A4 printing paper and a white desktop have different calibration values; so I want to measure the distance of milk, do I need to use milk for calibration? Instead of a target with 88% reflectivity.

[John E KVAM]

There are two competing issues.

It's always best to calibrate using the reflectivity of your known target if you have one. So that's a vote for milk.

But you must have a target that overcomes the effects of the cover glass. Cover glass crosstalk will cause your system to under-range if you don't have a very bright target. And so that is a vote for the white paper.

If you have a good cover glass, maybe the crosstalk effect is minor and calibration on a bright target is not required.

I'd spend some time experimenting. After all, when you are done, you can test it. You know the correct distance, and the sensor tells you what it gets.

So anything that works for you, works.

[Zwf.1]

I wish to rule out the cover plate effect. , can this api VL6180_RangeIgnoreConfigure() have a good effect on my device, and how should I set these parameters

[John E KVAM]

The cover plate (or cover glass) will create crosstalk - you must read and understand that calibration procedure. Range ignore refers to what you want to do with the range data after you get it. But if you don't calibrate your crosstalk, you will get the wrong answer and you cannot make a decision about what to do, if you have the wrong answer.

I suggest you go to:

https://community.st.com/s/article/Time-of-Flight-Cover-glass

It will lead you to some good documentation on what crosstalk is and how to eliminate it.

• john

[Zwf.1]

During the working process, there may be tiny water droplets condensed by water vapor on the cover. At this time, my ranging distance will become smaller. Is there any way to reduce this effect?

Is VL6180_RangeIgnoreConfigure() useful for these cases

[John E KVAM]

It's an optical sensor, so anything that blocks the light is going to be an issue. There are 2 things you can do.

You can look into anti fog coatings. These will slightly lower the amount of light that gets through, but they will help with the condensation. In the end, you win.

Or you can use a crosstalk-free coverglass. (www.hornix.com.tw) has them. Go to Product, scroll to almost the bottom. They are the ones with the raised barrier between the TX side and the RX side.

The condensation will still lower the number of photons, but you will never see any crosstalk.

And it's the crosstalk that is causing you to under-range.

You can for sure build your own, but the Hornix is easier.

• john

[Zwf.1]

How to use this api VL6180_RangeIgnoreConfigure(), what does this parameter （IgnoreThreshold）mean, and how to determine it

[John E KVAM]

Your issue is crosstalk. You must read and understand what crosstalk is.

go to:

https://community.st.com/s/article/Time-of-Flight-Cover-glass

It's a short article that will point you to other resources.

Then go to www.hornix.com.tx

This company molds optical quality plastics. They have a crosstalk free coverglass. Go to their products page and scroll almost to the bottom. Look for the coverglass that has a barrier between the Tx and Rx sides. This will keep the photons from short-circuiting the system and giving you that short answer.

Note there are 3 products - the one you want is the one on the right. That ridge is made of opaque plastic and extends all the way to the chip. It prevents the light from crossing over to the receive side no matter how dust, dirt, or milk gets on it. You must either use this, or invent something like it.

• john

[John E KVAM]

The RangeIgnore does not do what you think it does. It insures you ignore the photons from your coverglass.

If your coverglass returns some photons we need to subtract them from the range calculation and insure they do not create a 'ghost' target when there is none.

Generally during crosstalk calibration we set that value to 1.5 times the crosstalk value.

but then if your coverglass gets dirty, you will range on that dirt. So we can prevent a false target by increasing the RangeIgnore value.

-john

[John E KVAM]

The easy way to solve your problem is to go to www.Hornix.com.tw and buy the coverglass with the raised barrier between the TX and the RX side of the sensor. (From the products tab, go almost all the way to the bottom).

this coverglass creates a near-perfect barrier between the TX and the RX side. If you get enough dirt on the sensor you might blind it but you will never get a wrong answer. And if you get no answer, you have to indicate to the user he has to clean the lens.

There is no real trick to the Hornix cover, except the molding of the two plastics. One clear one (to 950nm) above the optical elements and the opaque one everywhere else. That and a very tight fit is what you need.

And sorry about the slow answer, not sure why I was not being notified.

• john

[Zwf.1]

Sorry, your answer was folded, I didn't see it and asked a few more times, thank you very much