Photo cell Detection vs Colour problem

I thought I'd through this out here to see what others have done for this problem. The problem is our banner photo-eyes have trouble detecting blue stain/ or grey coloured wood. We are using 24 VDC Prox type eyes. We have just recently changed to laser type prox eyes and things improved. This week however, we got some wood with a very dark stain, and Im told there is a lot more in the yard. I watched the laser and it now is having trouble. Increasing sensitivity is not an option. Thru beams are not an option. The eyes look down onto a chain running at 900 feet per min. Sales people all say they have the answer, but none as of today has worked when we tested. We have not tested any full spectrum eyes or single frequency other than red. The other mills here are having the same trouble. Has someone here solved this?

It your product thickness consistent enough to use a convergent beam diffuse photoeye? Basically you would be determining presence or absence based on distance from the photoeye not just reflectivity. In this case you could increase the gain without the risk of false detection.


Is there enough of a gap in the chain to allow for a small reflector? If so you could go with a polarized retroreflective photoeye.

Keith

I used "Sunx" color detection sensors in an application that differentiated the different pastel colors of contact lens cases. That project was about fifteen years ago. Can't remember the exact model number, though. Might have been this:

http://www.sunx-ramco.com/SunxPDFFiles/FZ_10.pdf

The sensors were easy to "train" and worked well.

Bruce,
If I understand this correctly you are just looking for present/absence, if so look into a Sick background suppression eye. I use them in the plastics industry on a lot of different colors without problems.
Bob

we have to deal with black surfaces all the time and this is always the challange. what is the sensing range? 1", 5", 50"?
your first choice is polarized photoeye with reflector,
for small distances try background suppresion diffused photoeye
(up to some 15-20" should be no problem),
for bigger distance i'd just go with analog laser distance measuring sensor.

before you choke, let me say this:
they are comming down in price and i just reviewed nice one
from IFM effector that is dirt cheap ($300US list price) and has range from 0.15m up to 4m for black and up to 10m for anything else. you will not need analog input since this device also has programmable digital output.
it has huge lense, visible beam, distance display righ on it,
very solid metal housing (2"x2"x2"), it's bright orange and what more can you ask for...

http://www.ifmefector.com/ifmus/web/pmd_4.htm

btw, the website is wrong, min distance is not 0.2m,
it's 0.15m (6").

What is the sensing distance? If close enough, capacitive might work...

Another thought is ultrasonic.

Since you have variations in color (and therefore reflectivity), it might be difficult to find a 'universal' photoelectric solution. At least not if you don't want to adjust/teach the eye each time you run a different color...



-Eric

Another slick Efector product are these micro laser sensors:


Rectangular Laser Sensors

I haven't found them to be sensitive to color yet.

Blah... I hate prox eyes for this very reason. I'm somewhat interested in the background suppression eyes though. I have several places where prox eyes were the choice because there was no way for a 'two sided' solution such as a retroreflective eye. But sure enough, run a product with high contrast colors and/or plastic film and the whole thing goes to hell. I definately need to look into ultrasonic.

Today I had a banner prox that was acting crazy. I had the gain lowered down to the point that a box 8 inches away wouldn't be picked up. But then I saw it turning it self on and off at 1 second intervals. Soon enough I realized it was picking up a reflector from the line next to it and detecting it as a 'presence' state. The line running next to it was running 30 pkgs/min. at 60ft/min, explaining the 1 second on/off interval. It surprised me that it managed to detect that from over 10 feet away.

Thanks for your input

The laser eyes we are using are mini type. The distance to the chain is less than 6 inches. There is no way to put a reflector or emitter/receiver combination without major modification to the infeed. The ultrasonic that banner has in their new cataloge looks good. It operates at 300 Khz and should be immune to VFD harmonics. ( I hope ). This seems to be the best option. If this wont work, I'll take a cutting torch to the bed plate and mount 30mm thru beams. (evil grin). There are no ultrasonics here so selling this to management isnt going to be easy. Im going to see if I can find a Loaner to test this out. I have no idea what this unit costs, but my guess is it isnt cheap. Thanks guys...

We have the same problem detecting blue pallets. We have resolved this by using background suppression photocells which alway operate at full intensity. I presume that these are similar to the convergent beam devices already mentioned.

I use a Banner QS30LDLQ to detect black product on a grey, white, dark green, and sometimes nearly black conveyor belt. This is the best laser photoeye I have ever used. The first one I installed two years ago and no one has had to touch it since. Very simple to setup also. I install these at an angle to the product out of habit. It always seemed to help when using infrared diffuse sensors looking at shiny black products. I don't know if it helps with this laser or not. It's very fast also. I use it as a registration input to a servo contoller.

QS30