Subject:
|
tracking an IR beacon
|
Newsgroups:
|
lugnet.robotics.handyboard
|
Date:
|
Sat, 6 Dec 1997 03:04:47 GMT
|
Original-From:
|
Jeff Keyzer <jkeyzer@ucsdNOMORESPAM.edu>
|
Viewed:
|
968 times
|
| |
|
|
Hello all -
My goal:
I'm trying to determine the feasibility of tracking an IR source with my
Handy Board. My initial scheme calls for using two IR detectors (the ones
with all the internal filters, etc, to provide a clean 5V output upon
detection of a 38-40khz IR signal.) Those two detectors would be situated
3-4 inches apart, on top of an RC servo and sonar transducer. (To be used
for ranging.) If the IR source delivers a 40khz signal with another square
wave modulated on top of it (much like the 6.270 board's 100/125 Hz IR
beacon routines), the difference in phase between the received signals at
each detector should give information about where the IR beacon is, with
respect to a position centered between the detectors where there will be no
phase difference. (The phase difference would be caused by a difference in
the length of the path the IR signal takes to get to each detector.)
With that said, does this look possible using my 68HC11 based handy board
to measure the phase difference, and (hopefully) discreet electronics to
generate the beacon signal? I'm planning to use two TIC lines, and then
use input compare to grab two rising/falling edges of the incoming detector
signals, then compare the times at which those edges were captured. Better
solutions are welcome!
Concerns I have:
1. Is the phase difference between the detected signals measurable with a
TIC line of the 68HC11?
2. Will a simple square wave modulated on the 40khz IR carrier be
sufficient to perform this measurement? Is some sort of signature required
to make the beacon unique? Is a variable frequency wave better? See next
concern.
3. How will background 40khz noise affect the measurement? My guess is
the effects will be quite severe in moderately sunlight/fluorescent lit
room. Spurious pulses from the detectors would be interpreted as edges to
be compared to find a phase - yielding entirely incorrect phase information
a good portion of the time. Thus, some way of distinguishing the beacon
signal from the noise, and then some way of finding the phase difference
between two such signals is needed. What about measuring the pulse width
of an incoming signal to make sure it matches the frequency of the beacon
when measured? This might help, but a small spurious signal might still
blend in with the beacon pulses, yielding a phase value that looks right,
but doesn't reflect the position of the beacon. Ideas?
I'd very much like to hear any thoughts/past experiences with this sort of
setup. I'm confident it has been done before, but perhaps with a more
complicated setup, faster microcontroller, etc. Anyone know of any
websites dealing with this? I'm going to look into how the handy board's
IR decoding routines work, perhaps I can combat the noise problem by using
IR remote techniques - only problem is that the beacon may have to be
powered by a PIC or other intelligent driver.
--------------------------------
Jeff Keyzer
UCSD EE Major
jkeyzer@ucsd.edu
http://sehplib.ucsd.edu/~jkeyzer/
|
|
Message has 4 Replies:
 | | Re: tracking an IR beacon
|
| I'm no electronic wizard which I think you probably would have to be to even consider what you are suggesting. Wouldn't it be simpler to just shield your sensors so that they 'look' in different directions? Have one look a little bit left and the (...) (27 years ago, 6-Dec-97, to lugnet.robotics.handyboard)
| | | Re: tracking an IR beacon
|
| (...) Hmmm...mmm... welllll.. if you are planning to use the GPUI series or equiv. as your detectors.. I think you may run into problems with attempting to detect critical time information about the modulated waverform. Namely, and I'm no expert on (...) (27 years ago, 7-Dec-97, to lugnet.robotics.handyboard)
| | | Re: tracking an IR beacon
|
| The phase difference seen by the sensors would be tiny at any practical frequency. Assume that you have the sensors a distance, D, apart and facing 'forward' on the robot. If there is a transmitting source at 90 degrees to the forward direction, the (...) (27 years ago, 8-Dec-97, to lugnet.robotics.handyboard)
| | | Re: tracking an IR beacon
|
| If you want to get into _all_ the details of tracking you might see if the local library has a copy of "Tracking Multiple Targets with Examples from RADAR" published by ArTech House. This book is crammed with densely worded (and annotated with math) (...) (27 years ago, 8-Dec-97, to lugnet.robotics.handyboard)
|
6 Messages in This Thread:
 
 
- Entire Thread on One Page:
- Nested:
All | Brief | Compact | Dots
Linear:
All | Brief | Compact
|
|
|
|
