Searching \ for ' Infra red tracker' in subject line. ()
Make payments with PayPal - it's fast, free and secure! Help us get a faster server
FAQ page: www.piclist.com/techref/index.htm?key=infra+red+tracker
Search entire site for: 'Infra red tracker'.

No exact or substring matches. trying for part
PICList Thread
'[EE]: Infra red tracker'
2000\11\03@094045 by staff

flavicon
face
Infra-red tracker.

I have an idea to use a PIC as a low parts count infra red
tracker. Here is the basic idea. Imagine a number of small
cheap "disposable" robots. Similar to RC toy cars with
brains etc. These are used to seach for metal on the ground,
like gold or land mines etc. They wander around randomly in
a pattern designed for maximum coverage.

When a robot finds metal, it stops and activates an infra
red beacon. Then the base station(s), which is on a pole
nearby, determines the location of the robot by pointing
directly at it, and can log that location. This is what the
infra red tracker needs, to be able to point at the location.

Pointing is important because when the job is done, the same
pointing head points a visible laser at each of the locations.
This will allow the operator to see each location from a
distance and in effect they can be "stepped through" one by
one and checked with binoculars etc. Obviously the cheap
search robots and laptop based base station would be
relocatable to other locations and be quite a saleable product
for ground gold fossicking (good in Australia) and land mine
disposal.

Spec:
2 axis motorised head with laser pointer and infra red
detector
range up to 5km (generally 1km or much less)
each robot on different infra red frequency
accuracy of both systems a couple of metres

Infra red sensor needs to quickly be able to determine
IF a robot is transmitting, and then point straight at
it with minimum hunting.

I imagined an infra red sensor with maybe a lens and
four receiver diodes, giving 2 axis detection. I can use
a high frequency for the signal and hence high gain
amps with high pass filters. And then just connect the
four signal levels to four ADC pins of a PIC and let it
do the rest! (we've all heard that before!!)

I appreciate anyones input, especially if you have used
infra red tracking in daylight at decent distances.
-Roman

--
http://www.piclist.com hint: PICList Posts must start with ONE topic:
"[PIC]:","[SX]:","[AVR]:" =uP ONLY! "[EE]:","[OT]:" =Other "[BUY]:","[AD]:" =Ads




2000\11\03@105652 by M. Adam Davis

flavicon
face
Noise and distance issues aside, I would think that you might have better luck
with the system if you have two  receivers, placed some distance apart, with
just a rotating axis, rotating the sensor around 360 constantly.  When either
one sees a signal, wait for the other one to see it, and solve a simple triangle
problem to determine the location of the robot.

Once you get to past a few meters from the pole, the difference in angle between
500 meters and 600 meters is not much (ie, the receiver will only be moving a
small amount up and down, and if you have a large viewing angle, you're not
going to know the difference between 500 and 600 meters)

Having two poles with alleviate this problem, though you'll still need a second
axis for the laser to point at the robot.

-Adam

Roman Black wrote:
{Quote hidden}

--
http://www.piclist.com hint: PICList Posts must start with ONE topic:
"[PIC]:","[SX]:","[AVR]:" =uP ONLY! "[EE]:","[OT]:" =Other "[BUY]:","[AD]:" =Ads




2000\11\03@111138 by staff

flavicon
face
M. Adam Davis wrote:
>
> Noise and distance issues aside, I would think that you might have better luck
> with the system if you have two  receivers, placed some distance apart, with
> just a rotating axis, rotating the sensor around 360 constantly.  When either
> one sees a signal, wait for the other one to see it, and solve a simple triangle
> problem to determine the location of the robot.
>
> Once you get to past a few meters from the pole, the difference in angle between
> 500 meters and 600 meters is not much (ie, the receiver will only be moving a
> small amount up and down, and if you have a large viewing angle, you're not
> going to know the difference between 500 and 600 meters)
>
> Having two poles with alleviate this problem, though you'll still need a second
> axis for the laser to point at the robot.
>
> -Adam

Thanks Adam, the two head system I scrapped quite early on
due to problems aligning two heads over larger distances and
getting the heads to communicate. I have chosen a one-head
scan and point system for number of reasons. Most likely it
will be used from a convenient overhead position on a hill,
etc and for use in less developed countries with foot access
only a one head setup and communication is much better.

Remember once an object is found there will be hand scanning
of the area with a metal detector, a good operator can
hand search a 5 metre square in a couple of minutes.

Most of the design will be pretty straightforward apart
from the problems using a one head 2-axis scanning system
to be reliable in daylight with a small infrared source. :o)
-Roman

--
http://www.piclist.com hint: PICList Posts must start with ONE topic:
"[PIC]:","[SX]:","[AVR]:" =uP ONLY! "[EE]:","[OT]:" =Other "[BUY]:","[AD]:" =Ads




2000\11\07@064049 by Mark Hull

flavicon
face
Hi Roman

I've done a fair amount of work with IR, and the best solution I've used is a fresnel type lens and linear CCD array.
The fresnel lens was simply a strip cut from a cheap alarm PIR detector.
The CCD array was placed at an angle to the fresnel.The fresnel focusses the IR, filters daylight, and "segments" the source. By checking which CCD segments are lit, you can determine distance to source. By using 2 detectors 1meter apart you can triangulate quite acurately.
Heat seeking missiles generally use a spinning 'reticule' (which is a disk with masked off regions, starting wide at the periphery and getting narrower toward the center) and a single detector.
If the disk is spinning at a constant known speed, then the pulse width of the detected IR is longer the further off axis it is.
I can's say much more without violating various contracts and agreements.

Maybee these ideas will help you, but one problem to watch out for. Moisture in the air stops IR dead in it's tracks. The more humid or misty the conditions, the shorter the travel (makes sense since IR is essentially heat)

Cheers
Mark

-------------------------------------------------------------------
Mark Hull
Engineering/Technical
EMS (Africa) (Pty) Ltd
PO Box 1026, Melville, 2109, South Africa
Ph +27 11 482 4470 Fax +27 11 726 2552

--
http://www.piclist.com hint: To leave the PICList
spam_OUTpiclist-unsubscribe-requestTakeThisOuTspammitvma.mit.edu




2000\11\07@201921 by John Mullan

picon face
I missed the original post, so I hope I'm not repeating an already asked
question here.

I was thinking of using to motorized cars (mysteriously disappeared from
kids toy box) and fitting them with a PIC, IR LED(s) and IR sensors.  The
idea being that the cars would seek out each other and play a game of tag.

Question: What would be the optimum price/performance configuration.  My
thought was a couple of LEDs and sensors per car.  Perhaps 4 (front, rear,
sides).

Any comments and/or suggestions??

Thanks,

John


-----Original Message-----
From: pic microcontroller discussion list
[.....PICLISTKILLspamspam@spam@MITVMA.MIT.EDU]On Behalf Of Mark Hull
Sent: Tuesday, November 07, 2000 6:39 AM
To: PICLISTspamKILLspamMITVMA.MIT.EDU
Subject: Re: [EE]: Infra red tracker


Hi Roman

I've done a fair amount of work with IR, and the best solution I've used is
a fresnel type lens and linear CCD array.

--- etc, etc, etc ---

--
http://www.piclist.com hint: To leave the PICList
.....piclist-unsubscribe-requestKILLspamspam.....mitvma.mit.edu




2000\11\08@092206 by staff

flavicon
face
John Mullan wrote:
>
> I missed the original post, so I hope I'm not repeating an already asked
> question here.
>
> I was thinking of using to motorized cars (mysteriously disappeared from
> kids toy box) and fitting them with a PIC, IR LED(s) and IR sensors.  The
> idea being that the cars would seek out each other and play a game of tag.
>
> Question: What would be the optimum price/performance configuration.  My
> thought was a couple of LEDs and sensors per car.  Perhaps 4 (front, rear,
> sides).
>
> Any comments and/or suggestions??
>
> Thanks,
>
> John


John, you could get a pretty good result with just two sensors,
a right and left, angled outward at about 45 degrees. Since it
has a PIC brain you can have a scan routine for turning the cars
around and looking for the other car. Mind you, differential drive
is much better for what you want than "car" type steering. It helps
if they can turn around on the spot!

If price is a big issue I would put one transmitter led at the top
center of the car, and right/left "eyes" to sense it.
-Roman

--
http://www.piclist.com hint: PICList Posts must start with ONE topic:
"[PIC]:","[SX]:","[AVR]:" =uP ONLY! "[EE]:","[OT]:" =Other "[BUY]:","[AD]:" =Ads




2000\11\10@040321 by Peter L. Peres

picon face
>front, rear, sides

I think that corners is a better policy. There was a robot project in some
magazine that used this arrangement and it seems to be better. You can
'aim' better if the left/right discrimination occurs on the axis of the
vehicle.

The project used a 16F84 in each robot, 4 sensors (modulated IR), and one
modulated IR transmitter looking everywhere (4 LEDs back to back). The
sensors looked NE, SE, SW, NW and the LEDs N, S, E, W. The drive was two
modified RC servos (potentiometer removed for freewheeling - like that
British school robot project thing).

The tagged robot could know whence the bad guy was coming (both robots
would transmit bursts at constant but different rates). The detectors
doubled as proximity sensors when something was real close (it would
reflect the own transmitter).

hope this helps,

Peter

--
http://www.piclist.com hint: The PICList is archived three different
ways.  See http://www.piclist.com/#archives for details.




More... (looser matching)
- Last day of these posts
- In 2000 , 2001 only
- Today
- New search...