|
Hi, I'm a mechanical engineering student doing a project regarding the construction of a team of autonomous robots doing a cooperative search. the mechanical design of the robot is common, a round, can shaped robot on 2 wheels + castor. They communicate using RF through the TWS434 and RWS434 transmitter and reciver modules. they sense surrounding and each other using analog IR sensors(TSL 262) mounted round the robot, which can read the intensity of IR radiation, hence able to tell distance from other robots and objects based on the intesity of IR rays either emitted from other robots or its own IR emission reflected off objects... problem is, i can't successfully construct or find an IR emitter that radiates uniformly, 360 degrees. I tried mounting the edges of the robot closely with IR diodes but the intensity of the IR radiation wasn't uniform enough plus the current consideration was an issue. Also tried using a single filament bulb hoping that IR emitted from the hot filament was more uniform radially. it was but there was a rather slow rise time associated with it. Any kinda help is most welcomed, thank U Regards, Ed |
|
|
omni-directional IR emitter???
Started by ●August 13, 2002
Reply by ●August 13, 20022002-08-13
|
I don't have an answer for you but you might try asking it of an engineer at Lumex. (They make LEDs and IR emitters, etc.) http://www.lumex.com Click on "Sales & Tech Support". On 13 Aug 2002, at 8:44, yellow_monstertruck wrote: > Hi, > > I'm a mechanical engineering student doing a project regarding the > construction of a team of autonomous robots doing a cooperative > search. > > the mechanical design of the robot is common, a round, can shaped > robot on 2 wheels + castor. They communicate using RF through the > TWS434 and RWS434 transmitter and reciver modules. > > they sense surrounding and each other using analog IR sensors(TSL 262) > mounted round the robot, which can read the intensity of IR radiation, > hence able to tell distance from other robots and objects based on the > intesity of IR rays either emitted from other robots or its own IR > emission reflected off objects... > > problem is, i can't successfully construct or find an IR emitter that > radiates uniformly, 360 degrees. > > I tried mounting the edges of the robot closely with IR diodes but the > intensity of the IR radiation wasn't uniform enough plus the current > consideration was an issue. > > Also tried using a single filament bulb hoping that IR emitted from > the hot filament was more uniform radially. it was but there was a > rather slow rise time associated with it. > > Any kinda help is most welcomed, thank U > > Regards, > > Ed > > > ------------------------ Yahoo! Groups Sponsor > ---------------------~--> 4 DVDs Free +s&p Join Now > http://us.click.yahoo.com/pt6YBB/NXiEAA/ySSFAA/dN_tlB/TM > ---------------------------------~ > - > > ">http://docs.yahoo.com/info/terms/ --- http://www.laser.com/dhouston/ |
Reply by ●August 13, 20022002-08-13
|
Do you also need to sense the IR uniformly in 360 degrees? How often do you need to measure the IR signal? Would one or two times per second be fast enough? Let me know, I may have some ideas? Tom yellow_monstertruck wrote: > Hi, > > I'm a mechanical engineering student doing a project regarding the > construction of a team of autonomous robots doing a cooperative > search. > > the mechanical design of the robot is common, a round, can shaped > robot on 2 wheels + castor. > They communicate using RF through the TWS434 and RWS434 transmitter > and reciver modules. > > they sense surrounding and each other using analog IR sensors(TSL > 262) mounted round the robot, which can read the intensity of IR > radiation, hence able to tell distance from other robots and objects > based on the intesity of IR rays either emitted from other robots or > its own IR emission reflected off objects... > > problem is, i can't successfully construct or find an IR emitter > that radiates uniformly, 360 degrees. > > I tried mounting the edges of the robot closely with IR diodes but > the intensity of the IR radiation wasn't uniform enough plus the > current consideration was an issue. > > Also tried using a single filament bulb hoping that IR emitted from > the hot filament was more uniform radially. it was but there was a > rather slow rise time associated with it. > > Any kinda help is most welcomed, thank U > > Regards, > > Ed [Image] > > |
|
|
Reply by ●August 14, 20022002-08-14
|
the idea is for each robot to give off IR 360 degrees uniformly so that other robots can read the intensity of the IR radiation using 8 IR sensors mounted all round each robot. it has to be uniform in all directions so that the intensity of the IR reading can translate accurately into distance regardless of which direction the robot is facing. each robot will take turns to flash its own IR signal which last for 0.1 sec and repeats twice every second. Also, I tried mounting 24 side leds on a short cirular pillar, each side by side each other. wired 6 parallel sets of 4 in series. Don't really get a uniform enough IR radiation and consumes 2 amps outta a 9V constant DC source. Thank U for replying, Ed --- In basicx@y..., twesthoff <twesthoff@f...> wrote: > Do you also need to sense the IR uniformly in 360 degrees? > How often do you need to measure the IR signal? Would one or two times > per second be fast enough? > Let me know, I may have some ideas? > Tom > > yellow_monstertruck wrote: > > > Hi, > > > > I'm a mechanical engineering student doing a project regarding the > > construction of a team of autonomous robots doing a cooperative > > search. > > > > the mechanical design of the robot is common, a round, can shaped > > robot on 2 wheels + castor. > > They communicate using RF through the TWS434 and RWS434 transmitter > > and reciver modules. > > > > they sense surrounding and each other using analog IR sensors(TSL > > 262) mounted round the robot, which can read the intensity of IR > > radiation, hence able to tell distance from other robots and objects > > based on the intesity of IR rays either emitted from other robots or > > its own IR emission reflected off objects... > > > > problem is, i can't successfully construct or find an IR emitter > > that radiates uniformly, 360 degrees. > > > > I tried mounting the edges of the robot closely with IR diodes but > > the intensity of the IR radiation wasn't uniform enough plus the > > current consideration was an issue. > > > > Also tried using a single filament bulb hoping that IR emitted from > > the hot filament was more uniform radially. it was but there was a > > rather slow rise time associated with it. > > > > Any kinda help is most welcomed, thank U > > > > Regards, > > > > Ed > > > > > > > > > > > [Image] > > > > Your use of Yahoo! Groups is subject to the Yahoo! Terms of Service. |
|
|
Reply by ●August 14, 20022002-08-14
|
I have done something similar to this once, using a laser for a State
Fair exhibit. I used a dot beam laser to project a circle that persons then could walk into. When the beam was broken as a person entered the circle, the exhibit started a recorded message. The technique should be the same, though. What you need is a convex mirror. Not a flat one and not a concave one. Convex glass, bubbles outward and will be fatter at the center then at the edges. This is the kind you want. Position the IR LED on the top of your unit. Aim it upwards. Place the LED inside a small tube-like device that limits the beam so it only fires into the direction wanted. In this case, up and onto the mirror. Place a non-reflective circle in the center of the mirror. Flat-black paper will work. Position the mirror above the LED, centered and facing the LED. It helps to have an apparatus that allows adjustment of the mirror, up or down. This will give you circular peripheral adjustment by raising or lowering the mirror. A shroud along the rim of the mirror also helps to capture and terminate the beams as they are projected and form a circular pattern. When the LED is fired, the beam is projected onto the mirror. Because the mirror is convex, the beam is then uniformly projected down and outwards, not directly back at the LED. The result is a circular pattern that can encompass the unit in all directions and distance, depending on the strength of the LED signal, the adjustment of the mirror position above the LED and length of the mirror shroud. The black circle at the center of the mirror is used to exempt the beams from striking the projecting robot. The result is a doughnut shaped beam, with the bot in the middle. It only uses one LED for transmission (or an array that fire all at the same time for greater signal strength) The other devices, in this case your other 'bots, then see a circular doughnut pattern. Design it using a super bright LED at first, so you can see the light pattern. Then swap out the super bright for the IR LED. If you cant obtain a convex mirror, you can make the same type of reflector using chrome spray paint and any uniformly convex object. (Gumball prize containers work, so to LEGG pantyhose containers)...paint the object with the chrome paint to serve as the reflective device. Homemade ones work but will distort the beam. Not as exact as a true glass convex mirror. Convex mirrors can be obtained from any place that sells Science Experiment Kits. The kind that are for light experiments and have prisms and refractors...stuff like that. Also keep in mind that IR will reflect in other directions away from doughnut as they strike other objects like the floor and such. The other bots will see a reflection just as well as the true beam. That was one reason why I used a laser in the Fair display. I used optics that were precision tuned to the bright laser waveform, and not the weaker beam reflections. That will be harder to do with IR since an IR wave does not deteriorate and disperse photons as quickly as other waveforms in the light spectrum. That is one reason IR is very reliable. Slow photon decay. The pickup optics would work best, I think, if they are facing more downward then up or outward. They will see the reflections off of the floor then, before seeing the actual beam. You can use that to your advantage as a warning (depending on signal strength)....weaker signals to indicate proximity, solid signals to indicate the doughnut has been breached. The transmitting bot should disable its own collectors each time the device is fired to avoid seeing its own doughnut and enable them again once the firing is complete. Hope this helps. .db. -----Original Message----- From: yellow_monstertruck [mailto:] Sent: Wednesday, August 14, 2002 1:15 PM To: Subject: [BasicX] Re: omni-directional IR emitter??? the idea is for each robot to give off IR 360 degrees uniformly so that other robots can read the intensity of the IR radiation using 8 IR sensors mounted all round each robot. it has to be uniform in all directions so that the intensity of the IR reading can translate accurately into distance regardless of which direction the robot is facing. each robot will take turns to flash its own IR signal which last for 0.1 sec and repeats twice every second. Also, I tried mounting 24 side leds on a short cirular pillar, each side by side each other. wired 6 parallel sets of 4 in series. Don't really get a uniform enough IR radiation and consumes 2 amps outta a 9V constant DC source. Thank U for replying, Ed --- In basicx@y..., twesthoff <twesthoff@f...> wrote: > Do you also need to sense the IR uniformly in 360 degrees? > How often do you need to measure the IR signal? Would one or two times > per second be fast enough? > Let me know, I may have some ideas? > Tom > > yellow_monstertruck wrote: > > > Hi, > > > > I'm a mechanical engineering student doing a project regarding the > > construction of a team of autonomous robots doing a cooperative > > search. > > > > the mechanical design of the robot is common, a round, can shaped > > robot on 2 wheels + castor. > > They communicate using RF through the TWS434 and RWS434 transmitter > > and reciver modules. > > > > they sense surrounding and each other using analog IR sensors(TSL > > 262) mounted round the robot, which can read the intensity of IR > > radiation, hence able to tell distance from other robots and objects > > based on the intesity of IR rays either emitted from other robots or > > its own IR emission reflected off objects... > > > > problem is, i can't successfully construct or find an IR emitter > > that radiates uniformly, 360 degrees. > > > > I tried mounting the edges of the robot closely with IR diodes but > > the intensity of the IR radiation wasn't uniform enough plus the > > current consideration was an issue. > > > > Also tried using a single filament bulb hoping that IR emitted from > > the hot filament was more uniform radially. it was but there was a > > rather slow rise time associated with it. > > > > Any kinda help is most welcomed, thank U > > > > Regards, > > > > Ed > > > > > > > > > > > [Image] > > > |
Reply by ●August 14, 20022002-08-14
|
Assuming you will use this in a dark room, because normal room lights have tons of IR that will fool the didtance measurements. Now, use a fairly heavy amount of diffusion on the LED. You will loose a LOT of light in diffusing it enough to get some uniformity ( one or two orders of magnitude ), but in a non IR flooded room, it might work. The diffuser will look like a white piece of plastic. You want non directional diffuse light if you want this to work. Test the diffuser by using a red LED, and hold a white paper around the light, There should be no structure or pattern to the light. That's what diffuse light is. The inverse square law might work now! Good luck with the project, -Les --- In basicx@y..., "yellow_monstertruck" <wm_edmund@h...> wrote: > > the idea is for each robot to give off IR 360 degrees uniformly so > that other robots can read the intensity of the IR radiation using 8 > IR sensors mounted all round each robot. > > it has to be uniform in all directions so that the intensity of the > IR reading can translate accurately into distance regardless of > which direction the robot is facing. > |
|
|
Reply by ●August 14, 20022002-08-14
|
Modulated IR will overcome normal room reflections and admissions without
a diffuser. That is how your television and stereo remotes work, even in full sunlight. They are modulated. Modulated and diffused will be even more reliable then just diffused. .db. -----Original Message----- From: v8fd [mailto:] Sent: Wednesday, August 14, 2002 4:04 PM To: Subject: [BasicX] Re: omni-directional IR emitter??? Assuming you will use this in a dark room, because normal room lights have tons of IR that will fool the didtance measurements. Now, use a fairly heavy amount of diffusion on the LED. You will loose a LOT of light in diffusing it enough to get some uniformity ( one or two orders of magnitude ), but in a non IR flooded room, it might work. The diffuser will look like a white piece of plastic. You want non directional diffuse light if you want this to work. Test the diffuser by using a red LED, and hold a white paper around the light, There should be no structure or pattern to the light. That's what diffuse light is. The inverse square law might work now! Good luck with the project, -Les --- In basicx@y..., "yellow_monstertruck" <wm_edmund@h...> wrote: > > the idea is for each robot to give off IR 360 degrees uniformly so > that other robots can read the intensity of the IR radiation using 8 > IR sensors mounted all round each robot. > > it has to be uniform in all directions so that the intensity of the > IR reading can translate accurately into distance regardless of > which direction the robot is facing. > |
|
|
Reply by ●August 14, 20022002-08-14
|
No, that won't help him. He is trying to measure distance with the inverse square law of optical radiation intensity fall off. It's not a data transmission reliability problem. It's an analog issue. That's my understanding of his post. -Les --- In basicx@y..., "Dan Bielecki" <Dan.Bielecki@A...> wrote: > Modulated IR will overcome normal room reflections and admissions without a > diffuser. > That is how your television and stereo remotes work, even in full sunlight. > They are modulated. > > Modulated and diffused will be even more reliable then just diffused. > > .db. > >ed] |
|
|
Reply by ●August 14, 20022002-08-14
|
Modulation will still work. Modulate the signal, period. The receivers look only for the modulated signal. It doesn't have to be for data, he's just looking for pulses. If he modulates the pulses and has the receivers look only for modulated pulses he can eliminate stray interference. He was looking to pulse twice a second. Thats a modulated pulse. The receivers look only for that pulse and ignore anything else. -----Original Message----- From: v8fd [mailto:] Sent: Wednesday, August 14, 2002 5:56 PM To: Subject: [BasicX] Re: omni-directional IR emitter??? No, that won't help him. He is trying to measure distance with the inverse square law of optical radiation intensity fall off. It's not a data transmission reliability problem. It's an analog issue. That's my understanding of his post. -Les --- In basicx@y..., "Dan Bielecki" <Dan.Bielecki@A...> wrote: > Modulated IR will overcome normal room reflections and admissions without a > diffuser. > That is how your television and stereo remotes work, even in full sunlight. > They are modulated. > > Modulated and diffused will be even more reliable then just diffused. > > .db. > >ed] |
|
|
Reply by ●August 14, 20022002-08-14
|
BTW...plain ol wax paper makes for a good home-made diffuser. So does the side of a plastic milk carton. Its a bit thicker but can withstand more punishment then the wax-paper. .db. -----Original Message----- From: v8fd [mailto:] Sent: Wednesday, August 14, 2002 4:04 PM To: Subject: [BasicX] Re: omni-directional IR emitter??? Assuming you will use this in a dark room, because normal room lights have tons of IR that will fool the didtance measurements. Now, use a fairly heavy amount of diffusion on the LED. You will loose a LOT of light in diffusing it enough to get some uniformity ( one or two orders of magnitude ), but in a non IR flooded room, it might work. The diffuser will look like a white piece of plastic. You want non directional diffuse light if you want this to work. Test the diffuser by using a red LED, and hold a white paper around the light, There should be no structure or pattern to the light. That's what diffuse light is. The inverse square law might work now! Good luck with the project, -Les --- In basicx@y..., "yellow_monstertruck" <wm_edmund@h...> wrote: > > the idea is for each robot to give off IR 360 degrees uniformly so > that other robots can read the intensity of the IR radiation using 8 > IR sensors mounted all round each robot. > > it has to be uniform in all directions so that the intensity of the > IR reading can translate accurately into distance regardless of > which direction the robot is facing. > |
