Retroreflective sensors

Hi Kevin, could you enlighten me on what you mean by a 'retroreflective sensor' please?

Thanks - John.

Hi John

The light from the LED`s will first pass through a polarised filter, hit a prism type reflector, and and bounce back to the detector having been turned through 90 degrees. The polarizing film in front of the detector is aligned at 90 degrees to the one in front of the LED, this means it will only "see" light returned from the reflector, eliminating false detection.

OK next question. What do you mean by "a wider detection area than is usually available"

Retroreflectors have a triangular shaped cone of operational incidence that is set by the corner cube facet and you can't really change that feature. If you need a larger angular response you could orient individual facets at different angles so that a broader angular incidence would always see a reflection from a portion of the available facets.

Or use a larger retroreflector? I have several over 2".

Well that would increase the effective coverage by the ratio of the effective diameter increase. The size window of acceptance is a function of the distance from the reflector so it would depend on what he is trying to do as to whether the proportionate increase of both the reflector and the acceptance window at his required range were adequate. My original thought would make it possible to keep a smaller device. You could even configure a round omni-directional device but the effective area would only comprise a section less than the diameter of the sphere, for any given direction. How about that did I just invent something? Naw, that would be to easy, lots of hits on google, here's just one patent reference:

http://www.patentstorm.us/patents/6742903-description.html

I will try and clarify what i am seeking to acheive.

I am trying to monitor a zone 2 metres square, In this zone i wish to detect the momentary presence of a moving retroreflective surface. Each detection, however brief, will reset a "presence" timer. Hopefully, the end result will give me a present or otherwise state.

Still inadequate information. You must know the distance from the optical transmission and reception point. The cone of a retroreflector is approximately 90 degrees but the further off axis you get the smaller the incident area and so the less light is reflected.

So it is not infeasible but you must also configure the optical angles of the sensing instrument accordingly. I have some very effective reflective sensing designs that I developed originally for automatically flushing restroom facilities and so they had to use the diffuse infrared mode of reflective sensing and as such are very sensitive. They could be adapted to work for your mode of operation but it would be necessary to have all the optical details and dimensions. If you would like to contract with me to provide this type of design please feel free to contact me with further details by sending me a message.

Kevin,

So you are trying to make a polarized retro-reflective photo-eye with a wide viewing angle. Purpose is to sense a retro-reflective object if it is anywhere within a 2m x 2m viewable area.

It seems to me that your primary problem will be attenuation of the light. The wider the angle of view, the lower your light intensity will be in the area covered by any one reflector. Some commercially available units claim very long distances of detection, which may just be because of a very well-collimated beam.

Idea 1--Use high-output commercial unit with adjustable gain and aim it so the beam reflects off a first-surface convex mirror to give the desired wider viewing angle. The retro-reflected response should reach the detector OK. Comments about the problems with tilt of the reflector away from normal (perpendicular to the beam) are valid, but my experience is that you still get good response at angles up to about 30-deg. This would translate to about a 6m distance from PE to detecting surface, which is pretty far.

Idea 2--Use a pulsed non-polarized light source and have the detector synchronized to only respond to a signal with the same frequency as the source. This avoids the >70% light loss in the polarizing process and would allow the LED to be over-driven. Again, certain manufacturers have pulsed units in their catalogs, but the output level you desire may require you to make your own. Again, the mirror, or a concave lens would be useful to disperse the light output to the requisite angle, etc.

Idea 3--Make your light source a cluster of 7 or more high-intensity LED's mounted with each one aimed a few degrees away from its neighbor, yet so their individual beams overlap. This would eliminate the use of a mirror or external lens. It could work easily with idea-2 above.

If all else fails, and cost is not the major factor, a vision system can do it too.

Besides the corner-cube type reflectors which you have mentioned, you could go to the extremely high-gain reflective materials 3-M sells--the only one worth considering would be the one that comes with a hexagonal pattern on its surface. I can look-up their part number for it if needed. I believe it has a higher tolerance for off-normal viewing angles.

Regards--John M.

Idea 4. If you need more light, go to a 630 to 650 nm'ish laser diode. The whole system including power supply are fairly cheap at 100 mW. You can expand the beam or whatever you want and plenty of light to waste.

Check out "industrial-lasers" on eBay.

Really guys, all that power isn't necessary if you build your receiver right. However, it can make the receiver a bit cheaper. I have a diffuse reflective communications system where you can put the transmitter and receiver anywhere in a room, even occluded from one another and the transmitter is battery powered and will send thousands of events using a standard IR LED and a 3 volt 1AH lithium coin cell.

Rcapper,

In his application, I think (correct me if I'm wrong somebody) he is wanting to detect when the object is within a designated zone or area and ignore it otherwise. Since the object is mobile and I presume the object can be one of many similar ones, I suspect that an emitter/receiver system such as you suggest may not be good. Also, in an industrial environment the ambient noise and light levels may make a diffuse IR system too prone to errors.

Idea 5) If the total number of objects being sensed is reasonable, how about using RFID. The chips are smaller than retro-reflectors, and the sensor antenna can be set-up to check a precise zone.

Regards--JMM

Gentlemen

Firstly let me thank you for your helpful suggestions (particularly John) and wish you a Merry Christmas in case I forget my manners.

Let me try and flesh this out as much as I can (I am limited as this application will be commercially sensitive)

I need to detect a moving target (don`t take that literally!) to the exclusion of all other personell or objects, who will be wearing retroreflective tape about his person (supplied by Reflexite - very effective in tests with conventional sensors).

I need to detect him in a 2m sq zone, however fleetingly, with an overhead sensor mounted approximately 1.5m above his head

Here is where I am

• Light intensity - as John says, up to 70% absorption by polarizers
• Need to stick with polarizers to exclude other reflective surfaces
• I believe I will need to modulate / demodulate the system to attain maximum efficiency, which I will do at a later stage (once I have made the damn thing work)
• Would love to use Infra-Red but I believe this will not work with polarizing filters
• RFID rejected at this stage as not commercially viable

If I may prevail upon you all to once more give of your best!

Many thanks

Kev

I did not mean to infer that diffuse IR would be appropriate, for obvious reasons. What I meant was the circuitry could be readily adapted to the application.

With regard to a limited field of view, if indeed that is a requirement, there are very simple methods of accomplishing this using collimation.