Subject:
|
RF ideas for the RCX
|
Newsgroups:
|
lugnet.robotics
|
Date:
|
Tue, 14 May 2002 03:15:09 GMT
|
Original-From:
|
Rob Limbaugh <rlimbaugh@greenfieldgroup.[Spamcake]com>
|
Viewed:
|
690 times
|
| |
|
|
Hello everyone,
My apologies for the rather long post, but I implore those interested to
read on. My hope is that someone can help offer solutions to the following
ideas.
For the past several months, I've been attempting to find a reasonable
wireless radio solution to RCX communication. Unfortunately, due to time
and abilities (lack there-of, on both counts), I've not been able to fully
complete the project.
The heart of the concept is to build a wireless infra-red repeater link
(WIRRL) that merely replaces the IR signal behavior with RF. There are
commercially available solutions to do this, but to spend over $80 on
something to hack apart and retro-fit didn't seem feasible, so I decided to
approach things on a "component" level, especially since the parts are
"cheap".
Just last month, I found this site speaking of a similar project, but seems
to go about it in a more complicated way (but very insightful!):
http://www.umcs.maine.edu/~pbrick/rBrick/papers/prelim100200.pdf
Practical home-brew IR-RF schematics seem almost non-existent on the web,
but I found this one to base a low-cost design from:
http://www.webelectricmagazine.com/00/1/rremote.htm
The original design uses a Visitect RF300XT RF Transmitter and a Ming RE99
V3.0 RF receiver. Since I could not locate a supplier of these parts (at a
reasonable cost), I chose to use the Linx LC family of devices
http://www.linxtechnologies.com/ldocs/modules/m_lc.shtml) which are
available from Digikey (http://info.digikey.com/T022/V5/0384.pdf). Two pair
of transmitters and receivers are less than $45.
An ideal solution does not require modification to any Lego hardware and is
compatible with all existing Lego IR hardware. [With the more expensive
Linx components, which have higher baud rates, a more custom solution could
be made allowing communications to be addressable using Holtek (or similar)
encoder/decoders. One data line would be for IR data, the address lines for
selecting RCX's. By tying the address lines to the parallel port, RCXs
multiple robots could be controlled individually.]
Also, RCX-to-RCX/Scout/Tower communication should still be attainable via
IR. A modular design is key since it may not be practical in a robot design
to have an RF transceiver block butted up to an RCX. Parts for the solution
should be cheap and easy to obtain. PIC's (or similar devices) shouldn't be
necessary.
In my "experimenting" with this project, I purchased some parts from
Reynolds Electronics (http://www.rentron.com) and Digikey to experiment with
IR and RF communication. The main components are the IR Demodulator, IR
LED's, RF Transmitters, RF Receivers, 555 timer(s) and Schmidt triggers.
(At least, that's what I've been using.)
For simplicity in prototyping a solution, I've been using the Lego Remote
and an RCX with a single motor attached. Instead of concentrating on
circuit to handle two-way communication, my intent was to focus on
communication in one direction and then just build a 2nd circuit when I
worked out the bugs.
Decoding the 38KHz IR signal and re-broadcasting it via RF is quite simple
and requires very few parts... the problem is rebuilding the 38KHz carrier
signal and placing the data on that carrier. All of my attempts have
resulted in blown out IR LED's or they send signals when no data is sent.
Basically, I took the schematics from the WEM site and only replaced the
transmitter and receiver modules and tried debugging from there.
Another idea:
Recently, I picked up an iPaq Pocket PC. It appears that the IR tower can
be used directly from the iPaq. Since the iPaq can be used as a wireless
network device and NQC is available for it, one could conceivably go that
route as well. This is a rather expensive solution, but it should be
possible, right?
Lastly, an idea for positioning:
In HAM radio, there are contests for using "Fox Finders". Basically (for
those that don't know), this is where a transmitter emits a simple beacon
tone on an RF signal and special antennas are used to determine the
direction of origin. The signal hits the two halves of the antenna out of
phase and thus "right" or "left" can be determined. When the signal is
equal at both antennas, the beacon is directly ahead or behind.
Predecessors to GPS use 3 RF beacon towers and a single antenna (mounted on
a vehicle) to triangulate position.
So... what if you had 3 antennas and 1 beacon with the antenna array mounted
on a robot and the beacon plugged into a wall outlet? Could a measurable
shift be determined from say, a 6 inch spacing between the antennas arranged
as an equilateral triangle? Would it be possible to use the sensor port as
a data bus for such a device? (e.g. - turn on the sensor port as a "Request
to Send" and then a raw value is sent regarding "x"... then another "Request
to Send" is initiated to retrieve a raw value corresponding to "y"). When
the sensor is returned to passive mode, perhaps switches/bumpers could still
be used.
In any event, I'd be more than willing to share my theoretical notes with
those much more experienced in electronics than I am. If anyone would like
more information, please email me. I'll try to get as much info up about
what I've learned ASAP.
Rob Limbaugh
|
|
Message has 1 Reply:
 | | Re: RF ideas for the RCX
|
| I'd suggest simply buying a spare rs-232 tower, taking it apart and using the guts. You'll have full duplex IR, just connect it to your RF xceiver. "Rob Limbaugh" <lego-robotics@crynwr.com> wrote in message news:002101c1faf5$8f...imbaugh... (...) (22 years ago, 14-May-02, to lugnet.robotics)
|
3 Messages in This Thread:
 
- Entire Thread on One Page:
- Nested:
All | Brief | Compact | Dots
Linear:
All | Brief | Compact
|
|
|
|
