Derby Talk is a forum for Pinewood Derby, Awana Grand Prix, Kub Kar Rally, Shape N Race Derby, Space Derby, Raingutter Regatta and other similar races where a child and an adult work together to create a race vehicle and a lot of fun and memories
gpraceman wrote:That is fine in a photocopier, but a different story on a Pinewood Derby track. In a photocopier you can ensure the sensors are tangential to the paper.
A piece of card on the bottom of a car would ensure a surface tangential to the sensor. No wheelbase issues as the card would be aligned with the front of the car.
Regardless, I think there's significant investment in the simplistic fashion in which sensors are placed currently. If my ideas are loopy, it would be nice to explore alternatives. I am leaning heavily towards to optical interrupter with the spring loaded flag that I linked to. Which BTW is another paper sensing mechanism.
I've often wondered why certain car designs don't put a piece of card in the front to show a flat bottom and clear full width profile to the break beam sensors. If you move some cars over the finish line pushed all the way over to one side, the difference between where the point of the nose crosses the line and where the beam is broken seem to be significant.
My dream would be a camera mounted on a post on the side of the track looking down across the lanes. The computer would do vision recognition to determine the cars crossing the finish line. I've seen systems like this for high school/college track events. Check out the "Eagle Eye Track & Field Timing" system. http://www.eagleeyetrack.com/TimingIntro.htm
This would allow for video display of the finishes and collection for the event's videos.
It's a dream for my tracks -- a bit of a budget breaker though! Eliminates the front end design and sun light on sensors and obstruction of the views. . . . They are local so I'm seeing what might be done . . .
Mr. Slick says: Honey, I am doing this for the kids, not myself.
Mr Slick:
There was another thread on finish line cameras.
I found a university project that was using a standard CCD as a line camera but no details as to how it was done. All the CCD sensors I've found have a fixed number of clock pulses per frame and only the pixels sent are changed.
Need a CCD where a reduction in pixels sent could be complemented with a proportional increase in scan rate.
A source of line sensors is flatbed scanners. Some interesting projects making cameras out of them. Would be interesting to achieve optics that would span the track.
Cameras would need to be mounted perpendicular to the track to avoid parallax.
Our track uses simple IR break sensors. In previous years, the overhead IR has been provided by a simple desk lamp with an incandescent bulb, shining straight down on the track from above.
I'm working on a new timer system (USB based), but I'm still planning on sticking with the IR break method for my finish sensors. However, I plan on building a VERY low profile (maybe 3/4" think) overhead bridge with the IR LED's mounted in plexiglass for diffusion purposes. I could actually extend this and make the upright closest to the crown (we set up with the crowd on one side of the track) out of plexi as well. Or, with the right support, you could get away with NO support on the one side, so the emitters hang out over the track. We've found that the overhead bridge makes minimal difference in audience viewing. However, IF you choose to use a halogen work light (as we did for testing purposes last year), you have to keep it off the track at LEAST 5 feet (otherwise you have too much IR flooding / reflection, and the sensors cannot be broken by the cars).
A side note about noses and effects on race times, our Rover Crew actually wants to build a 5 lane wedge (our new track is going to be 5 lanes) out of a single piece of plywood, etc, that we can use to calibrate the sensors location, and experiment with different nose designs for beam breakage, etc. We'd also like to do some extensive testing of different shape and weight placements versus wind and wheel drag, etc. I know a lot of this has been done before, but I haven't found it to the extent we've been postulating.
altontoth wrote:A side note about noses and effects on race times, our Rover Crew actually wants to build a 5 lane wedge (our new track is going to be 5 lanes) out of a single piece of plywood, etc, that we can use to calibrate the sensors location, and experiment with different nose designs for beam breakage, etc.
Fun as a gadget, useless as a tool.
To calibrate, the only useful method is T-squares and straight edges.
To test different front end design wrt to beam breaking, use a solid straight edge to calibrate the sensors, then set the straight edge a given distance back from the beam-break position. The different front ends can be aligned with the straight edge and differences in beam-break can be measured.