On Sunday, 7 May 2000 08:00:00 UTC+1, Ron Woodward wrote:
In article <38E4A8BB.F3FF257@mirrotek.com>, Boaz Bezborodko <boaz@mirrotek.com>
writes:
I'm familiar with this approach. But what I'm really looking for is to be >able
to point the sensor in a particular direction for communications with only >one
transmitter in the area (there may be a few in the area).
What you need is a little bit of optics in your system. Add a lens or reflector
to make it directional. Place the transmitter/reciever at the focal point of a
lens. Try one of the flat credit card size magnifier lenses. They are only a few bucks and availabe in drugstores. I am not sure how well glass lenses will
conduct IR.
Second method is to use a reflector. Get an old flashlight and place the IR detector/ transmitter where the bulb is facing the reflector (pointing inward towards the reflector not outward towards the other reciever/reansmitter).
With either system block stray light with a tube or box. Using this method it should not be hard to obtain a directionality of a few degrees.
There is a problem with this method. It is dificult to aquire the target. If you can only see the object in a very narrow beam it can take a while to search out and find what you are looking for. Minor variations in the robots attitude due to rough floor or ground will cause the plane of search to tip up
and down. If your robot tips by more than a few degrees the target may not be visible at all.
Think about how the headlamp on a bicycle or car works when driving over a rough road. When you go down a bump all you can see is the ground when you govr
over a hill all you see is the tops of the trees.
A better solution is to use a cylinderical lens or reflector. (harder to find)
I have found long cylinderical lenses in the eye glasses department of pharmacies They are designed to place on a line of text to make it bigger for thoes with eye problems. I beklive they were about $5-$10 in price. Place your
reciever/transmitter behind this
it will give you a narrow horizontal view and a wide vertical view. A cylinderical reflector will do the same thing. The only such reflectors I have
seen are in industiral heater and UV curing systems. You could probabaly make
one that would work OK by using a piece of metalized milar (Baloon) or crome tape (auto parts store) stretched over a bulsa wood frame.
Plot out a parabola with your printer y^2= 4 * FL * x
FL is the focal length. the focal length determins how far away you will need to put the sensor. The longer the focal length the narrower the field of view.
An aproximation would be: field of view = Inverse tangent of ( diameter of sensor / focal length)
Use the plot out of the parabola to cut out two pieces of bulsa wood ( or some
other material). You can use either side of the curve for the frame. It depends
on how you want your reflector to look. Like a D or like a C. The C shape would
be better for attaching the Crome tape. The D shape works better for the mylar.
The D also provides structure for mounting the sensor inside.
Cut 4 or more pieces of wood the same length to use as spacers.
Glue these between the two parabola parts. Attache the reflective material to
the curved surface so the reflective side points to the inside of the curve.
USE a visible LED first to observe the reception pattern on the wall. You should see somthing like a line. The ends will bow in or out depending on where
you place the LED. Position your sensor at what looks like the best focus.
This assembly can be mounted on a Servo or other scanner device to locate the beacons. It should accomodate a large vertical error (+/- 30 degrees) and have
a narrow horizontal acceptance angle +/- a few degrees. It should also have good sensitivity due to the magnifying effect.
Caution: Do not point at sun it could damage the sensor or even catch fire.
Have you ever used a magnifying glass to burn ants?
One more technique. Use a baffel tube to block the field of view. Take a toilet
paper tube. Use black tape to cover the end of the tube leaving a slot 1/8 of an inch wide.
Make a smaller tube that will fit in the larger one. Make a simular 1/8 inch wide slot on the end of it. Place the sensor inside the smalle tube centered in
the tube some distance behind the slot. Place the smaller tube inside the larger tube. Orentent the slots the same way. The sensor can now only see a small area and will be very directional.
The major disadvantage to this method is the low sensitivity. It may still be enough for your application and it is the easiest to build.
And to be able to
get an good idea of the direction the of that transmitter.
Boaz
Alisdair Sutherland wrote:
Boaz Bezborodko <boaz@mirrotek.com> wrote:
Has anyone developed a very directional IR receiver module? I want to >> > be able to locate the relative position of an IR transmission using an >> > IR sensor mounted on a servo, but I need the sensor to be very
directional.
Good luck
May the Mass * Acceleration be with you.
Honestly mate you are such a lifesaver. After hours of trawling forums your post shone out and finally convinced me on going through with my project idea. If you don't mind, I'd like to run it past you and get your opinion on it?
I'd like to mount a narrow-beam IR receiver on a rotating mount driven by a stepper motor (like a 2D LiDAR) that will detect 3 IR beacons. This will be placed on a 4-wheeled robot who's goal will be to navigate a grid maze. The beacons will be placed
outside this maze and will be used to get the robot's XY coordinates as a method of localisation.
So far, the best methods I've come up with is by using beacons and either IR or LiDAR (which is expensive).
Any help you or anyone else on this thread could provide would be much appreciated!
--- SoupGate-Win32 v1.05
* Origin: fsxNet Usenet Gateway (21:1/5)