what would you use to measure a distance 600..1300mm?

[DiTBho]

hi,
I am building a complex machine that moves a plane vertically and I need to contactlessly measure the position between two movable points that range between 600 and 1300 mm with a precision of 5mm.

I tried the some Sharp reflective IR sensors, but ... the voltage-distance curve is hardly non linear, and it requires too much fixed-point/floating-point computation (which wouldn't be a problem with STM32, but at the moment I am using a less powerful MPU), also the curve decreases like 1/x so at 1000mm the output voltage from the sensor is very very little hence the error very high.

I am also going to try again a couple of new ultrasonic sensors whose output is a pulse, and the relation between the pulse and the measured distance is the pulse width so you need a precise timer. Say, from 100nS to 50mS.

Alternatives?

[mrburnette]

Measuring with rotary encoder...

https://youtu.be/X794KKbXqJ8

[ag123]

VL53L0X lidar?
https://www.youtube.com/watch?v=sU0hr2jtX2Y
it turns out that fancy gadget / chip is a st product
https://www.st.com/en/imaging-and-photo ... 53l0x.html
this is probably as close to the 'james bond' spy thrillers as you can get, tiny, lidar, and measures distance using time of flight

[DiTBho]

Measuring with rotary encoder

This would require to open and hack the motor-box I am using in order to install the rotatory encoder, but I would avoid it because the motor box is sealed.

[DiTBho]

VL53L0X lidar?

I don't know which is real precision of the VL53L0X, I tried the Sharp TOF10120 (SPAD, Single Photon Avalanche Diodes), which is ranging 100..1800mm, and ... well, it performs well in the 100..500mm range, but In the higher points, 750 and 1000mm, the readings are off by a great deal, over 10%. The documentation that I was able to obtain for the sensor claims that the range and accuracy are affected by many factors, including ambient light and the reflectivity of the surface being detected.

[DiTBho]

Measuring with rotary encoder...

The suggested video shows a Measuring Wheel, and it would be a good idea if I didn't need a contactless measurement. I mean, I cannot add a rail just for the measuring wheel, and the only way I can have a rotatory encoder in a contactless measurement environment is ... by hacking the motor box that moves my complex machine vertically.

A quadrature sensor would be fine and I could install it to the shaft of motor. I am sure I could do something similar, but ... I would also need to "Dremel-cut out" the motor box because it's sealed.

I am thinking about that, and I will reconsider it in the future if I will found no other solution.

[DiTBho]

just out of curiosity, has anyone ever hacked an already existing product like a "digital measuring tape" or built one?

[ag123]

VL53L0X lidar?

I don't know which is real precision of the VL53L0X, I tried the Sharp TOF10120 (SPAD, Single Photon Avalanche Diodes), which is ranging 100..1800mm, and ... well, it performs well in the 100..500mm range, but In the higher points, 750 and 1000mm, the readings are off by a great deal, over 10%. The documentation that I was able to obtain for the sensor claims that the range and accuracy are affected by many factors, including ambient light and the reflectivity of the surface being detected.

well these modules are conveniently there from the likes of Adafruit
https://www.adafruit.com/product/3317
https://learn.adafruit.com/adafruit-vl5 ... duino-code
and the usual places e.g. ebay, aliexpress
https://www.ebay.com/sch/i.html?_from=R ... X&_sacat=0
https://www.aliexpress.com/wholesale?ca ... xt=VL53L0X
they may not meet the 'industrial' requirements, hence u'd need to decide if you would play with those

for anything more complicated / precise they'd cost more i'd guess. encoders is 'conventional' and a pretty accurate way to measure distances
but it would take a way to count ticks e.g. against calibrated markings etc.
another thought would be that since the measurements are stochastic, and if the sensors are after all inexpensive, would it help to take several measurements and with different sensors and taking an average? i think it'd improve accuracy. a thing though, would be if multiple sensors can be used concurrently, but even if not i'd guess they can be triggered one at a time

another way though is lasers, since lasers travel in straight lines, it may be feasible to place sensors and figure things out from trigonometry. that i'd guess can be very accurate

[DiTBho]

they may not meet the 'industrial' requirements, hence u'd need to decide if you would play with those

yeah, I need 5mm of precision, but I don't know if the above dispersion data plot is really trust-able. Something tells me this sensor is 10mm of precision if not worse. I'd like to investigate, so I have just bought a couple from Amazon

take several measurements and with different sensors and taking an average?

Kind of sensor fusion? Nice idea.

[DiTBho]

ebay

Little OT/rant. I won't buy anything from eBay!

After 10 years and more than 600 positive feedbacks received as buyer and seller, I deleted my account there since I reported a lot of known scams to eBay without any result, and, worse still, I have a written email from eBay where they deliberately protect a dude who not only declared the false on customs, which is a serious federal crime in the US, but also sold fake goods, and eBay removed my negative feedback because "judging by their algorithms" this dude is supposed to be have "privileges".

That's unacceptable! I asked why they removed my negative feedback since not only I received damaged goods but also the parcel documentation could have caused me to pay a hefty fine.

They replied:

[...]
I checked the feedback you left and although it did not meet the criteria for MANUAL REMOVAL, it qualified for system removal. Due to one of the seller’s privileges which he exercised on this issue.

At the same time, for the purposes of giving you a peace of mind please know that you did not do anything wrong. I assure you that the removal does not have anything to do with negatives or violations against you. It’s just more on the default settings on our system along with the seller’s privileges.
[...]

That's enough for me. eBay is not a serious place, and feedbacks are a joke if an "algorithm" can automatically decide who is good and who is bad, and, worse still, if it can delete one of clue (feedbacks are for what?) a human being has to judge.

When I signed up to their site, I agreed to their end user agreement, but I don't agree with what I found they do, so I think I made the right decision when I asked eBay to close and delete my account. I totally disagree with that point in their policy.

I am looking at Amazon and Bonanza as valid alternatives.