From: Gerry Schneider
X-Mailer: Mozilla 4.73 [en]C-SYMPA (Win98; U)
Subject: Re: Detecting moment when a fast moving object passes over a line
Date: Fri, 29 Nov 2002 02:28:43 -0800
NNTP-Posting-Date: Fri, 29 Nov 2002 02:39:31 EST
Organization: Bell Sympatico
Richard Sim wrote:
> Hi, I'm working on a project currently that's for automatic lap scoring in
> radio controlled car races, where the cars may be travelling in excess of
> 40-50 MPH at the point where they cross over the start/finish line, and
> several cars may pass over the line within close proximity of eachother.
> The cars will have "transponders" in them, which detect when they cross over
> the start/finish line, and send a signal to a reciever/decoder that's
> connected to a PC.
> I'm having trouble finding a way to implement the start/finish line
> reliably; the biggest problems are that the transponder may be anything up
> to 15cm (6") above the ground with upto 3mm of graphite, plastic or
> aluminium between it and the ground as they will be inside the cars, and
> must be as small and light as possible (1.5"x1.5" is pretty much stretching
> Anything that works on line-of-sight (such as IR) is out of the question
> unfortunately, as the cars may very well be completely covered in mud, dust,
> oil, etc.
> So far I'm thinking of using 2 linear hall effect sensors connected to an
> op-amp to create a very sensitive magnetic sensor, with some kind of wire
> loop or electro-magnet under the start/finish line.
> I'm not very confidant that this will work however, as firstly I think the
> wire loop/electro-magnet wouldn't have a strong enough magnetic field to
> trigger the hall effect sensors, and secondly the hall effect sensors may be
> triggered by the car's motor as depending on the type of R/C car, the motor
> could be very close to the transponder (and the motors have very powerful
> magnets as they can go in excess of 50,000 RPM).
> The tranponders have to transmit their unique ID to the reciever/decoder
> when they pass over the start/finish line, and to do this I'm probably going
> to be using the Linx ES series of RF modules (but the LC series for testing
> however as they're much cheaper).
> I'm definately a beginner at electronics, and this is the first complex
> circuit I've worked on, so any help in finding a reliable way to implement
> the start/finish line & transponders would be greatly appreciated (and
> please be detailed as, like I say, I'm a beginner).
Without getting involved in any on-going arguments about terminology and/or
expertise, maybe I can just just mention a few points.
An AM receiver of the 418/433MHz variety won't work in close proximity to an
electric RC motor, so best to simply transmit blindly (one way) from each car if
possible. Multiple transmissions and some simple data and time coding should
take care of the occasional missed packet at the receive end, and short
(millisecond) randomized packets will take care of possible overlap from
multiple transmitters. Not perfect, but easy to get 99.9% reliability at 10's of
feet range (I push about 150').
A simple wire proximity detector in each car can be built from a small pickup
coil (a telephone type is a good start) feeding a PIC or whatever processor.
This is similar to the operation of the dog collar used in electronic fences.
Just be sure to put a few amps thru the wire (start/finish line) and tune
everything by LC resonance; that will overcome broadband magnetic interference
from the motors. I use about 4 amps at 32kHz using a step-down transformer -
only needs about 1 volt to drive a few feet of wire.
I'm using the above techniques already in an unrelated application - wish I
could give out circuit/software details but it's proprietary to the customer so
can't do it - sorry. However, it isn't rocket science by any means, and it does
work well. One last hint (ass getting close to the boot now) - if you need
fairly precise detection of the line, use a non-symmetrical drive waveform and
an untuned detection coil. The received signal will flip phase as you pass over
the wire and the asymmetry makes it easy to detect.