Hey you
On a (solid) convey-belt, I have a small (one way moving) object of approx 8mm x 5mm, that I need to either move to one lane or another, depending on a 0.1mm tab placed asymmetric on the top of it.
Hence I would like to have a horizontally mounted optical sensor, that outputs a simple high/low value, if the tap passes by.
...I don't need any distance sensoring.
Could you please direct me to a good/reliable sensor with such an accuracy?
Hopefully not to expensive
Thanks!
Kind regards
I just measured the thickness of the thinnest wire I can find: a 1/4W resistor lead is over .3mm thick. Are you sure about the 0.1mm thing?
I just measured a sheet of laser printer paper and it is about 0.15mm thick. I have to ask the same question as JimboZA.
Is this a different color than the belt?
if the thing on top is from metal and the rest not a metal detector could do the job.
Yes, either by coil or by capacitance. Capsense could probably do it, in that case. Two probes, one grounded.
If the belt isn't very wide, you might be able to use a linear CCD sensor of some sort, with some sort of magnification optics.
Sorry guys (and girls if any)...
I had a fly in my eye and an elefant stomped on my toes, while measuring the tab. The tab is 0,6-0,7 mm
The convey (which is actually a shaped chute) is made in blank stainless stell and my object is in white pvc.
NB. The tab is 1,0mm wide, and I would like to register approx. 1-2 objects per sec.
...if sensing horizontally is on the x-axe, I would only be able to use 2,2mm on the z-axe and 0,6mm on the y-axe.
cr0sh > I can mount the sensor as close as I want to, and I can also mount a reflector, reader or likewise on the opposite side.
So basically... does anyone know good photoelectric sensor either through-beam (opposed) or retro-reflective, that can do the job in these dimensions.
This is where a picture paints a thousand words. Is it possible to show pictures of the two items your trying to separate?
If one version of your object has a tab sticking up and the other version does not then could you just use a bar angles across the conveyor to catch the tabbed one and slide it to the side?
If not then for such a thin tab width you might me able to use a laser beam break to detect tabs passing by but how will you then separate them?
Sure... here is a drawing of the object outside its shaped chute.
I've tried to make a simple seperator on the chute, with some sort of opposite tabs, but I could not get it reliable enough... also I think this is much more joy 
A laser beam break, photoelectric sensor, ultrasonic sensor... I'm pretty much game for any suggestion.
I've attacehed a small drawing of how I plan to seperate the obejcts into two different lanes, when I have detected which way the object is entering, based on the assymetry of tabs.
The entire disc wil be approx Ø2-3cm
(Yes I know the proportions are not correct... it is an illustration)
Thanks!
If this is stable enough, I think I'd set a laser to just hit the 0.6mm tabs sticking up, along with another sensor to detect that the object is in the beam so it knows when one without tabs is there.
If the laser beam is too wide, just shot it through a pinhole in a sheet of heavy duty aluminum foil.
I wonder if you could use an optical mouse as a close-up camera, suspended so the objects pass just below the mouse. But how is the positional repeatability of the object on the belt?
An optical or laser mouse is really just a tiny, low resolution camera. I don't know if it would have enough resolution for this. Side lighting could help pick out the tabs.
http://frenki.net/2013/12/convert-optical-mouse-into-arduino-web-camera/
polymorph > doh thats a simple solution...
I have spend many hours looking for a component that would shoot and sens a tiny laser beam, but your quite right, I just need a light sensor on one side, and a laser on the other side where I narrow the beam through a small hole...
It is going to take some calibrating to get i right, but it should work.
yeah I've been reading about those optical mouse tricks too...
