SparkFun Forums

[Hydra7403]

Hi everyone,

I need to detect the distance of a point in space and i was wondering if using a IR Sensor or a Sonar.
I need an accuracy of around 2mm (i read that IR are a bit nasty when it comes to precision ).
Here's an example of the detection:

Basically,giving the facing direction of the sensor,i need to know the distance between the sensor origin and the
point A in a straight line.

What should be better?

Thanks in advance.

[jremington]

Sonar ranging modules are much more accurate than most IR distance sensors, but can be confused by echos and surfaces that are not perpendicular to the line of sight, such as your drawing suggests. Their range is also quite a bit wider than 10-150 cm, whereas many ranging IR sensors are limited to less than 100 cm.

[Mee_n_Mac]

I need to detect the distance of a point in space.

Is this a point on a surface or a lollipop suspended in air ? How big in diameter is this spot (or lollipop) ?

[TastyResistor]

It depends on your application.

IR sensors wont perform as well outdoors due to interference from the sun.
My experience with hobby priced IR and sonar sensors tells me that sonar sensors would be more accurate, with a higher distance too.

You won't achieve 2mm accuracy for any medium sized distances. A laser might be able to.

[Hydra7403]

Thanks for the replies.

It can be pointed at 'any' objec.The project is to make a sort of scanner to reproduce the object in 3D by taking multiple samples.
The spot,the smaller it is,the better,a spot of 1-2 mm would be great.

It's an indoor project,@TastyResistor what do you mean by "medium sized distances"?
The problem with lasers is that they cost a ton,and i'd prefer to use something not visible (not a laser beam) if possible.

Mmm i suppose there's no way to make that sonar spot more narrow to avoid the interferences that "jremington" posted isn't it?

Thanks.

[Mee_n_Mac]

Sonar is definitely out, the beamwidth is too large. Laser is the way to go. You might find this approach interesting.
http://techcrunch.com/2013/11/10/meet-s ... he-masses/

[jremington]

You've probably heard of the Kinect hacks for 3D scanning, but if not, here is one link: http://www.wired.com/design/2013/03/kinect-dev-kit/

[TastyResistor]

Lasers aren't visible. They'll also give you distance measurements to all visible points within its field of view.

With sonar you might be able to achieve a cm accuracy at best. Nowhere near 2mm. 2mm accuracy is bordering on industrial robotics precision.

The Kinect might be the best (and cheapest) approach to recreate the object in 3D.
I would personally wait for the xbox one kinect camera to come out. It has a depth sensor in it which sounds suitable for your application.

Otherwise there is really no substitute for a laser.

[Hydra7403]

I wanted to do a "personal project" so i though to be fine with a limited budget (<100$)
but i see that it's kind of impossible since the components are too expensive for the precision needed.

About the kinect,yeah,it's an handy solution but those models unfortunately are still "rudimental" meaning that they
can't be used in a real case scenario (it`s those kind of models suitable only for video-calling and similar).

Just a last question,is it also impossible to get a 2mm precision using a infrared module
reducing the distance range to 10-90 cm?

Thanks again.

[Mee_n_Mac]

Just a last question,is it also impossible to get a 2mm precision using a infrared module
reducing the distance range to 10-90 cm?

It's not just the range resolution but also the vertical and horizontal resolution, aka the beamwidth. Just like pixel size affects picture resolution.