Very directional IR receiver
Boaz Bezborodko
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.
Thanks in advance,
Boaz
Gordon McComb
receiver/demodulator. But IR reflects like crazy, especially on white
walls and ceilings. There will usually be a "background" level of light,
which you'll have to compensate for.
A small 20mm +3 to +6 diopter lens should work just fine. Be sure to add
a baffle around the lens and IR module to keep out stray light.
-- Gordon
Mitch Berkson
the output beam. Perhaps you could adapt one of those.
Mitch Berkson
"Boaz Bezborodko" <bo...@mirrotek.com> wrote in message
news:38DFDC07...@mirrotek.com...
Boaz Bezborodko
How would I compensate for the "background" light? I would need this for
another variation of the application for which I was originally looking. I
want to be able to project an IR beam and only have a detection on a sensor
if the direct beam falls on the sensor, not any of its bakground
reflections. Could I use a IR phototransistor and get a voltage signal? I
could then read it through an A/D pin, but I would have to run it through a
bandpass filter first to avoid other light. How difficult would this be.
Boaz
Gordon McComb wrote:
> You could always put a positive diopter lens over the IR
> receiver/demodulator. But IR reflects like crazy, especially on white
> walls and ceilings. There will usually be a "background" level of light,
> which you'll have to compensate for.
>
> A small 20mm +3 to +6 diopter lens should work just fine. Be sure to add
> a baffle around the lens and IR module to keep out stray light.
>
> -- Gordon
>
> Boaz Bezborodko wrote:
> >
Paul Gardner
the same diameter as the sensor. The length of the tube could be calculated
using simple maths to obtain the required directional accurracy. To reduce
background light a piece of filter material could be put over the end of the
tube, this is available in the UK from people such as RS and Farnell.
Filtering light electronically we be pretty difficult.
Just my thoughts...
Paul
Boaz Bezborodko <bo...@mirrotek.com> wrote in message
news:38DFDC07...@mirrotek.com...
> 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.
>
>
> Thanks in advance,
> Boaz
>
>
Gordon McComb
adjust the threshold programmatically. The problem is that you can't use
a compartator or ADC if you're working with one of those modular
infrared receivers/demodulators, like the Sharp units available at Radio
Shack. They have their own comparators built-in, and they will often
trigger on reflected light (in fact, they're meant to do this, as their
typical application is remote control receivers for TVs and VCRs).
If you will be using modulated IR, you will need to design your own
receiver and demodulator. If you're not using modulated light, then you
must also filter for only IR, and try to avoid light from such things
light bulbs and outdoor windows. This approach starts to get fairly
complicated if you can't precisely control the environment.
Boaz Bezborodko
What about putting the IR transmitter behind an IR opaque tube to limit the
transmission angle and reducing the duty cycle?
Boaz
Nick Ray
I was toying with the idea (or rather never got round to it) of using a
detector via a scanning mirror like those in supermarket barcode scanners or
laser printers. You'd effectively get a pulse the timing of which would
relate to the direction of the IR source. You'd still need to build
circuitry to detect the peak of the pulse and to measure the timing of it
w.r.t a reference marker, but it could be pretty accurate.
Boaz Bezborodko wrote in message <38DFDC07...@mirrotek.com>...
Gordon McComb
>
> Gordon,
>
> What about putting the IR transmitter behind an IR opaque tube to limit the
> transmission angle and reducing the duty cycle?
Neither of these really helps the problems I mentioned. If you're
modulating the light, then you need to demodulate it (requires a 567 or
a PLL of some type, plus maybe a bandpass filter, etc.). If you're not
modulating the light, then other light sources (desklamp, sun from
outside), can cause false readings. Mechanically shielding the sensor or
transmitter won't help with either of these issues.
-- Gordon
Boaz Bezborodko
reflections by reducing the intesity of the transmitted light by reducing the
duty cycle and reducing the area seen by sensor. Then the only time the signal
would be strong enough to register on the receiver module would be what is
projected directly at it. Would this be difficult to do and would it work?
Boaz
Gordon McComb
> What I meant was that if I use modulated light I could reduce readings from
> reflections by reducing the intesity of the transmitted light by reducing the
> duty cycle and reducing the area seen by sensor. Then the only time the signal
> would be strong enough to register on the receiver module would be what is
> projected directly at it. Would this be difficult to do and would it work?
Not sure I follow you on "reducing the duty cycle." Your best bet is to
follow how the Sharp IR demods work. You could probably use a commonly
available 567 tone decoder to do your experimenting. It's a touchy chip,
but it will work for you. Read up on its data sheet; it includes the
formulas needed to set the bandwidth and center frequncy. Some advice:
don't breadboard the circuit. Use a soldered board, and use short
lead-lengths for the components. If you've done the math right and
picked close tolerance parts, you shouldn't have to do much re-work.
Keep in mind that if some other light source is as bright as the IR
transmitter the detector will be "swamped" and won't see the IR light.
This could happen in daylight outdoors, for example, or with an
incandescent lamp.
-- Gordon
Dmitri Katchalov
Gordon McComb <gmc...@gmccomb.com> wrote:
> If you're modulating the light, then you need to demodulate it
> (requires a 567 or a PLL of some type, plus maybe a bandpass filter,
> etc.).
How about synchronous detection?
Say, a signal from the detector (amplified by an op-amp if necessary)
is fed through a 1-to-2 analog demultiplexer (a pair of FETs) which
is driven by the same clock that you use to drive the LED.
So now you have two signals, corresponding to "off" and "on" states
of LED. Then you can filter them and feed them into + and - inputs of
another op-amp. The output will be detected demodulated signal.
If the ADC is fast enough you can do it all in software.
LED on. delay. Take a sample. LED off. delay. Take another sample.
Delay is calculated to produce around 40Khz frequency.
Repeat the process a few times and average the results.
Then compare the averages of "on" and "off" values.
How does this sound?
Dmitri
Sent via Deja.com http://www.deja.com/
Before you buy.
Gordon McComb
transmitter is on the robot; the clock that drives the IR transmitter is
somewhere else in the room.
-- Gordon
Dmitri Katchalov
As for directional sensor, how about replacing the sensor temporarily
with a LED of the same shape and check the size and shape of the spot
it produces?
In article <38E164...@gmccomb.com>,
--
Boaz Bezborodko
set to be on 50% of the time and off 50% of the the time. If the 'on' time is
reduced and the 'off' time is increased then the receiver module would get less
energy to work with. This should decrease the chance that what it is seeing is a
reflection for light from the emitter as opposed to direct view of the emitter. If
the sensor also has a lens or tube to limit its field of view then that would further
increase the possibility that any detected signal comes from direct view of the
emmiter.
Boaz
Gordon McComb
has an inertial and inductive energy store) or a capacitive load, but IR
detectors are fast-switching semiconductors. They are either off or on,
and a duty cycle at the frequencies we're talking about doesn't affect
the amount of "energy" they receive.
You can certainly do things to make the common Sharp IR
receiver/demodulator less sensitive, and therefore more likely to only
trigger on direct light rather than reflected light. One is to operate
the transmitter at a higher or lower frequency than what the module is
rated for. You could experiment with changing the duty cycle; I don't
know what that will do with these units. A neutral density filter to cut
down the light might also help.
Remember that all things considered, if the sensor has a lens or tube
you can still just pick up a reflection off a nearby wall. That alone
won't help you discriminate between direct and reflected light if the
sensor is overly sensitive.
-- Gordon
Boaz Bezborodko
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). And to be able to
get an good idea of the direction the of that transmitter.
Boaz
Alisdair Sutherland wrote:
> Boaz Bezborodko <bo...@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.
>
> How about using four sensors each looking at one quadrant. Take sample
> values from each sensors and compare levels. The strongest signal should be
> comming from the transmitter. This would enable you to seek the most likely
> direction of the transmitter and home in.
>
> \ /
> \O/
> O O
> /O\
> / \
Boaz Bezborodko
What about increasing the resistance tied to the transmitting LED so as to reduce the LEDs
power? If the level is set right then the reflection in the background wouldn't be enough
to activate the sensor, but direct sight of the LED would.
Boaz
Ron Woodward
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 <bo...@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.
Boaz Bezborodko
Thanks for the belated post.
Boaz
jdra...@googlemail.com
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!