Split times

[jkale]

I would like to see turns 9 thru 11 or 12. Check point at 11 I think. Dont know what check number.

I think that is check 4 - 5

[Chief Geek]

As many of you know, my dad, Rich, and I will be competing this year in the ex-Estey Miata. That said, I still plan to provide split-time data.

The 2012 system worked OK but wasn't nearly reliable enough and required my constant attention to make work. It also took way too much of my time to compile results.

Here are the planned components for 2013:

1) New photosensors
- a) Improves from 20 msec resolution to 400 usec (speed traps now possible) (won't miss Sherman or Josh D anymore because their head "shadow" will now be detected)
- b) Real time clock integral to the sensor
- c) Memory card slot that is easy to swap card through a car window (not fussy & expensive SD cards, think tic-tac boxes with a plug)
- c) Much cheaper/simpler transmitter transmitter circuit
- d) Trying to get a small batch of "real" circuit boards made
- e) If b, c, d, & e successful, build 6-10 sensor pairs to allow for many sectors & speed traps (speed trap is very short sector)
- f) Better indicator LEDs to help with sensor aiming
2) Keypad "sensor"
- a) used by a friendly start/finish/checkpoint worker to type in the number of car @ start or just past checkpoint
- b) same real-time clock & memory card scheme as photosensors
3) Start tone "sensor"
- a) same real-time clock & memory card scheme as photosensors
- b) can be integrated into a photosensor and placed near start line to act as reaction time/launch guage
4) PC software to save me from hours of spreadsheet tedium.
- a) reads memory cards
- b) compiles "events" into run data
- c) prints formated reports
- d) maybe, feed results to wifi stations at top and bottom for smart-phone users

In my head, it works like this:

1) While workers are getting in place, the sensors are turned on and a memory card is plugged in.
2) Data is gathered.
3) During lunch and after last bring-down the memory cards are swapped for fresh and the "full" cards are brought to me.
4) My PC reads the cards, compiles data, and prints results.

In testing, the real-time clocks I'm using in the sensors all stay within 30-ish msec over 24 hours. So long as I sync the clocks each morning, life is good. Also, I can read each clock in the evening, measure the error, and interpolate between for more accuracy.

I've you've read this whole thing, I fear for your sanity.

Anyone have ideas on how many sectors & speed traps may be useful?

Keep in mind, we don't have to show every sector on every report. Even with 10 sensors & 11 sectors, we can always choose to show the same 3 sectors we used in 2012.

Anyone know any cost-effective, small-batch circuit board manufacturers?

Any ideas, suggestions, and non-mean critiques are greatly appreciated.


Paul

[sachilles]

How do the car numbers get synced up with who trips the sensor?

[KevinGale]

Once you really get all the bugs worked out in a year or two we should talk about using the same technology at finish. No rush since what we have is working fine but we do have similar problems detecting motorcycles at times and the real time clock in the sensor would improve accuracy.

[walterclark]

I use an outfit that batches up boards from a number of small orders to reduce costs. Batch PCB or some such is the name. I am away now and will get you the specifics next week.

[Chief Geek]

I can't believe anyone read all of that. You guys must be "special".

How do the car numbers get synced up with who trips the sensor?

In 2012, I entered them into my laptop where they were recorded as another "event" with a time stamp. I manually selected the car number "event" that goes with the start tone and sensor events. In 2013, I was hoping to find friendly workers that would type in a car's number as they heard over the wire that it was next to start or just started. Someone at start or finish would be perfect, but anyone listening to the comm system or within site of the start line could do it.

Once you really get all the bugs worked out in a year or two we should talk about using the same technology at finish. No rush since what we have is working fine but we do have similar problems detecting motorcycles at times and the real time clock in the sensor would improve accuracy.

Sensors should be pretty easy. When I get into sensor "production", I'll "de-content" a sensor pair or two so they are just sensor pairs and will give you a useable output.

I'm guessing spares are a good idea because, I've heard, sensors occasionally get run over by very wide cars with black and gold paint jobs. Perhaps by other types of cars too.

I use an outfit that batches up boards from a number of small orders to reduce costs. Batch PCB or some such is the name. I am away now and will get you the specifics next week.

I've read about many electronics projects that sourced their boards from BatchPCB. They'll be one of the services I'll look into.

[Challenger392]

Suggestions for speed traps for the first event of the year (ascutney)

#1 the tree of fear, could be fun to see who is the bravest/craziest
#2 the bridge, because its a difficult place to go fast
#3 since we are running the full 4 miles some point in the top fast section could be fun.

[Challenger392]

Had an interesting idea, I dont know what you were doing previously but a 60' time could be a good thing. so you could see how a good launch off the start line affected your overall time/see what techniques improve your launch time.

[KevinGale]

Had an interesting idea, I dont know what you were doing previously but a 60' time could be a good thing. so you could see how a good launch off the start line affected your overall time/see what techniques improve your launch time.

It would be nice and several people have asked. The main barrier is that the hardware we use right now just isn't capable of handling the timing. In the future we might replace the basic stamp with hardware that could handle this. Then of course we run into the same problems we have with the split times. Marrying up the data is a lot of work right now. The light (like the split lights) doesn't know what car is causing the event or even if it is a race vehicle at all.

All things we can work out in the future. I remember when this was all done on paper so progress is being made.

[britracer89]

Paper!? How about the grease pencil and plexi-board leanded aganst a leanto?