Loooong distance IR detection, and IR collimation

Hi,
Every body know about the IR remote controls that we use everyday for our TV, music set etc. They are short range, almost omni direction (due to
reflections from almost any surface) controllers.
Imagine that we have a tube/pipe of some sort with radius R, and lenth L. We have positioned the IR transmiter at the end of this tube and alligned it in such a way that its illimunation axis parallel to the tube so that the IR source can only be visible if the observer/receiver aligned with the tube/pipe. BUT the inner surface of the pipe/tube must be coating with an IR absorbing material so that the tube will not act as a wave guide. Or the tube/pipe must be made out of an IR absorbing material.
Do you know any IR absorbing material?
Is it possible to collimate the IR light source with the above mentioned method to 2degrees? OR is there a way of creating very narrow IR light cone ie. collimation around 2degree?
If we use an ordinary IR remote control receiver (or some other low-cost IR sensor) to detected the IR light about 100 meters away from the collimated IR source how much IR power (how many IR LEDs or what else) we need?
Thanks
Rico Maxle
Add pictures here
<% if( /^image/.test(type) ){ %>
<% } %>
<%-name%>
Add image file
Upload
Rico Maxle wrote:

If i didn't get the impression that you were trying to figure out a new way to set of an IED I might offer a few suggestions. I must be getting paranoid in my old age.
Shawn
Add pictures here
<% if( /^image/.test(type) ){ %>
<% } %>
<%-name%>
Add image file
Upload
Rico Maxle wrote:

Sure. Lots. Black paint, for example. It won't help you.

No problem, your collimation scheme will work fine. But it won't help you. You will wind up with a very dim, approximately collimated beam.

All that will do is to absorb some of the light before it leaves the instrument. This won't do anything to increase the amount of light you will receive at the far end, because the light you absorb would have missed the receiver anyway. If you put it at the receiver end instead of the transmitter, it will help quite a bit in reducing background light, which will blow you out of the water otherwise. (Sunlight is *very* bright.) If you use lenses, you'll do much better, but success won't happen by accident--do it by the seat of your pants and you're liable to be off by 6 orders of magnitude from practicality.
It isn't difficult to calculate how much light you'll receive--the detected photocurrent is proportional to the optical power received, and the optical power just spreads out in a cone from the transmitter. The transfer efficiency is just the ratio of the detection area (lens or photodiode) divided by the cone's cross-section at the distance of the receiver.
Assuming you're using kilohertz modulation like a TV remote, your noise will be dominated by the shot noise of the background light, so that's pretty easy to calculate--just measure the photocurrent I_BG caused by the background in a realistic situation and compute the 1-Hz noise as i_N = sqrt(2*e*I_BG), where e is the electron charge (1.6e-19 coulombs).
Then you can figure out what detection bandwidth will give you the signal-to-noise ratio you need. On your first try, it will be a very small fraction of a hertz.
When you find out how frightfully bad your SNR will be, you can improve it in several ways:
1. Lenses 2. Very small detectors 3. Lots of transmitter power 4. Baffles to exclude stray light 5. Laser sources
Unlike hobby laser radar, this is not at all impossible--it just takes work and thought.
Cheers,
Phil Hobbs
Add pictures here
<% if( /^image/.test(type) ){ %>
<% } %>
<%-name%>
Add image file
Upload
Phil Hobbs wrote...

[ snip ]

There's Rico Maxle's answer, use an IR laser.
--
Thanks,
- Win
  Click to see the full signature.
Add pictures here
<% if( /^image/.test(type) ){ %>
<% } %>
<%-name%>
Add image file
Upload
You probably won't find a paint which does the job. Most paints are reflective at grazing angles. For a surface coating some kind of flocking (a fuzzy surface) is better. Better still is to put a series of thin black painted baffles (like washers) inside the tube. The last one should be slightly larger diameter so no energy can directly illuminate the inner edge.
You really need to define the system better. Whether you can pass the needed signal depends on: 1. transmit power 2. power density at the receiver (determined by distance, atmospheric absorbtion, and level of collimation (beam angle). 4. detector acceptance angle is as important as beam angle 5. detector area 6. ambient illumination (usually solar energy in the receiver optical bandpass) 7. detector noise -sets ultimate detection signal-bandwidth limit. 8. required signal bandwidth. Matched modulation and detection schemes can extract very weak low bandwidth signals out of broader bandwidth noise.
To get better signal to noise it's common practice to use optics to narrow the beamwidth, but that may not be practical for hand held or vehicle mounted systems. Other methods use more power and/or narrower bandwidth signals. Modulating the transmitter can allow detection of your signal in the presence of optical or thermal noise. Reducing the optical bandwidth using a partially collimated laser plus interference filters on the receiver can minimize optical interference (usually from sunlight). Cooling can reduce detector thermal noise. You're requirements and budget will determine what is possible and what is practical.
Rico Maxle wrote:

Add pictures here
<% if( /^image/.test(type) ){ %>
<% } %>
<%-name%>
Add image file
Upload
RicoMaxle wrote:

Hi,
Is this a theoretical exercise, or do you have some intended use? If you tell us just what you are doing with this, we may be able to provide more directly useful answers.
It is somewhat of a mystery to me why so many posts here have no reference to the actual inteded application.
Luhan
Add pictures here
<% if( /^image/.test(type) ){ %>
<% } %>
<%-name%>
Add image file
Upload
"Luhan"

** It is characteristic of usenet that posters secret the real purpose behind asking a question.
There are three major reasons for this;
1. The purpose involves doing something illegal.
2. The purpose is highly dangerous.
3. The purpose is completely stupid.
Sometimes the purpose is illegal, dangerous and stupid all at the same time.
Since folk cannot get free, expert advice on illegal dangerous and stupid ideas in the real world - they naturally head straight onto usenet.
Where they find dozens of obliging fools ready and willing to help them on their way.
....... Phil
Add pictures here
<% if( /^image/.test(type) ){ %>
<% } %>
<%-name%>
Add image file
Upload

NOTA (None of the above.)
The majority of such posts (on sci.optics at least) appear to emanate from EEs or MEs and the like who appear to have a very limited understanding / knowledge of optics in general.
Add pictures here
<% if( /^image/.test(type) ){ %>
<% } %>
<%-name%>
Add image file
Upload
Phil Allison wrote:

I'm only slightly less cynical. Perhaps they cannot say because they are doing it for a company that doesn't want its secrets revealed. Even in that case, it would be nice to preface the question with 'I cannot reveal exactly what I am doing here.'
Luhan
Add pictures here
<% if( /^image/.test(type) ){ %>
<% } %>
<%-name%>
Add image file
Upload
Phil Allison wrote:

Thanks Phil...I feel like I'm not alone. I try to use common sense when judging the validity of a post before responding. Unfortunately some people can't overcome their egocentric desire to spew onto the world their knowledge. In this unfortunately sometimes ugly world, a little common sense and proper judgment can save lives and limbs. If I knew how to make a bomb, or even a small component of one (which I do), I would not tell some guy I have never met how to make either of them. I'd rather be paranoid than the reason for another human(s) losing their life. Thanks for getting it.
This is an open forum, and if your question has nothing to do with robotics or an air of illegal or questionable moral standing, or it you can't post your project in detail, you'll get no help from me. Sorry.
Shawn
Add pictures here
<% if( /^image/.test(type) ){ %>
<% } %>
<%-name%>
Add image file
Upload
Shawn B. wrote:

<rest snipped>
The first thing I thought of was the OP is wanting to create YAO (Yet Another Opticon) device for the *illegal* control of traffic light signals. You see these types of requests all the time on the hacker forums. Of course, it could be a legitimate question, but since collimators are common finds in any optical catalog (or use a spotting scope in reverse) you have to wonder. Why go to the obviously lossy approach of a baffle tube, other than the fact that it's smaller and more stealthy.
-- Gordon
Add pictures here
<% if( /^image/.test(type) ){ %>
<% } %>
<%-name%>
Add image file
Upload
On Thu, 11 May 2006 11:32:03 -0700, Gordon McComb wrote:

I saw some guy who was allegedly using one of them, but got caught when he kept showing up on the red-light camera. ;-)
Cheers! Rich
Add pictures here
<% if( /^image/.test(type) ){ %>
<% } %>
<%-name%>
Add image file
Upload

Is an interesting topic I've never heard of or thought of before.
First, what is an Opticon?
Then, how does a (self-actuated) traffic-light work?
It contains a *camera* of some kind?
(I had imagined that it used some kind of sonar to tell when a car appeared.)
It "photographs" in red? Or infra-red (heat)? Or what?
Interesting stuff -- please do elaborate a bit.
Thanks!
David (usually just "lurking" here)
Add pictures here
<% if( /^image/.test(type) ){ %>
<% } %>
<%-name%>
Add image file
Upload
On Tue, 06 Jun 2006 22:17:44 +0000, David Combs wrote:

Well, an "Opticon", in this context, to the best of my knowledge, is a photocell (electric eye) mounted up on the stoplight bracket, and ambulances and fire trucks (and, presumably, the cops) have some secret code that they flash through their windshield with a strobe - the Opticon picks up that signal, and changes the cross-lights to red, and changes the light for the emergency vehicle green. Some guy had decoded that code, and was breezing through town, but presumably got caught because the red-light cameras were timed to the moment when the light was _supposed_ to have turned red - there was this car in the intersection, and the "switch to red" signal had been overridden!
They caught the guy. :-)
Cheers! Rich
Add pictures here
<% if( /^image/.test(type) ){ %>
<% } %>
<%-name%>
Add image file
Upload
David Combs wrote:

The answer to this is "it depends". Some use coils embedded under the road (often used at highway on-ramps in Long Island, for instance).
The more modern scheme is a camera in the traffic light that exploits the highly reflective license plate on a car to detect if the intersection is occupied.
I read something - here maybe - about a motorcyclist who was tired of not triggering this type of detector. He put a strip of orange reflector tape on his helmet and all was well.
Add pictures here
<% if( /^image/.test(type) ){ %>
<% } %>
<%-name%>
Add image file
Upload
<Rico Maxle> wrote in message

I see no reason why this can't be readily achieved. As someone has already mentioned glancing blows tend to reflect better, so a tube with an inner surface with texture (ie: black cloth or velvet lined for example) will probably work somewhat better than a shiny interior surface. To achieve a two degree cone your LED light source will need to be mounted in the center of a tube of radius R, with a length L of greater than or equal to R/0.0175. In other words, select R, then solve for L>= R/tan(1degree) = R/0.0175. For example, if you select a tube with a 3 cm diameter (R= 1.5cm), then the length needs to be at least 86cm long. The LED will need to be located at one end mounted at the very center of the tube and pointed straight down the tube. Using this method will not however increase total range, it will only make the light source more directional. In order to get any gain in range, you would need to use at least one lens.
However... Given that 100m is fairly modest and not too hard to achieve with the right semiconductors, I would probably avoid pesky optics and lenses and whatnot.
The SFH 4503 infrared LED by OSRAM (available from Digikey) claims to ouput typical infrared power of 40mW (at 100mA), and probably around six times that at 1A pulses. The LED claims a half angle of intensity of +/- 4 degrees. If we assume half of the power is contained in the 8 degree main cone, then at 1A it should provide something like 120mW (typical, min. is much less) into the cone. At a distance of 100m, you can find the output in mW/square meter by using some basic trigonometry. It should be something roughly like 100meters * tan(4 degrees) = r. (assuming a fixed circle projected at a distance 100m away). The area of the circle is approximately 3.14*r^2 = 154 m^2. So the energy delivered at that range will be around 120mW/154m^2 0.78mW/m^2.
The TSOP1256 infrared receiver by Vishay (available from Mouser) might be able to receive a signal of that strength under ambient irradiance levels of up to perhaps around 10W/m^2 (which equates to around 1.4klux of incandescant illumination or 8.2klux of sunlight like light). Those levels of illumination are much weaker than direct unshaded sunlight, however, they are probably quite a bit brighter than your house is lit, or even your office or grocery store is probably lit. You will definitely want to put a hood over your receiver module to make sure it is shaded from direct sunshine if this is an outdoor application.
Datasheet for the TSOP1256 can be found here: http://www.vishay.com/docs/82013/82013.pdf
Datasheet for the SFH4503 LED can be found here: http://catalog.osram-os.com/media/_en/Graphics/00029741_0.pdf
Your milage may vary in practice. These are all very rough numbers. Multiple LEDs will require a much larger diameter tube (and consequently much longer as well), so if one LED isn't quite enough for your task, a lens might not be totally out of the question.

Add pictures here
<% if( /^image/.test(type) ){ %>
<% } %>
<%-name%>
Add image file
Upload

Dear Fritz,
THANK YOU for your very informative posting. I'm sure many people who read your reply feel the same as me and appreciate the info and your help.
Regards,
Rico Maxle
Add pictures here
<% if( /^image/.test(type) ){ %>
<% } %>
<%-name%>
Add image file
Upload
Actually in practice, if you are using the half angle as the measure of collimation, the tube you describe for a 4 LED even assuming a reflectivity of only 50% for glancing blows over 85 angle of incidence will only reduce the 1/2 angle cone to approximately 3. If you used a highly reflective material you would get practically no change. That's a long way from the 1/2 angle cone of 1. My own estimate is the tube would have to grow to at least 500 meters to actually reduce the signal to a 1/2 angle of 1. Not too practical and not much signal left.
Add pictures here
<% if( /^image/.test(type) ){ %>
<% } %>
<%-name%>
Add image file
Upload
On 2006-05-10, <Rico Maxle> <Rico> wrote:

a convex lens would be a good start. the focus fo IR will be behind the focus for visible light.

Edmundson Electronics in the 80s made a 100m infrared security beam using a pulsed signal between a single LED and a single photodiode they had plastic lenses at both ends.
--

Bye.
Jasen
  Click to see the full signature.
Add pictures here
<% if( /^image/.test(type) ){ %>
<% } %>
<%-name%>
Add image file
Upload
Take a look at
http://ronja.twibright.com /
I had my senior design students build this and the LED based link works at about Ethernet for about 1 km. We had difficulty in making it ethernet compatiable, and pointing the system. We also didn't test for a full KM, just indoors since the semester ended.
However the short answer is that about 1 km with 1 LED at 10 Mbps.
john muth
Add pictures here
<% if( /^image/.test(type) ){ %>
<% } %>
<%-name%>
Add image file
Upload

Polytechforum.com is a website by engineers for engineers. It is not affiliated with any of manufacturers or vendors discussed here. All logos and trade names are the property of their respective owners.