GPS lap timer

Heyou · Posted: Tue Jan 20, 2004 4:37 pm Post subject: GPS lap timer

Is there any software for TTN2 on a Dell PDA that would work as a circuit lap timer?

cheers

Del.
_________________
Good judgement comes from experience. Experience comes from bad judgement

lbendlin · Posted: Tue Jan 20, 2004 5:09 pm Post subject:

No, there isn't. Don't forget that GPS information comes in once a second, and has a diluton of 3 to 15 meters. Not good enough for a lap timer.
_________________
Lutz

Report Map Errors here:
TomTom/TeleAtlas NAVTEQ

Heyou · Posted: Tue Jan 20, 2004 5:59 pm Post subject:

Ah! i wondered if the resolution wouldn't be high enough, OK thanks for that info. saves me searching high and low for that :D

Cheers

Del.
_________________
Good judgement comes from experience. Experience comes from bad judgement

DavidW · Posted: Tue Jan 20, 2004 8:42 pm Post subject:

If you want an excellent stopwatch program, try Timekeeper from Conduits. It has nothing at all to do with GPS, mind you - it's just a very programmable stopwatch and timer program. There is a trial version, so you can make up your mind before paying out.

Details at http://www.conduits.com/products/timekeeper/

David

Heyou · Posted: Tue Jan 20, 2004 11:13 pm Post subject:

Cheers David, good idea, you must have read my mind :o i'd decided that was the only other way to go, to have a big stopwatch/lap timer on the PDA so it's easy to hit the screen each time i cross the S/F line. I've just got to make sure i don't get caught timing myself at the event, or i could get taken off the track.

cheers

Del.
_________________
Good judgement comes from experience. Experience comes from bad judgement

DT2 · Occasional Visitor Joined: Jan 31, 2004 Posts: 4

Actually, I think it should be possible to build something with a gps. I don't expect something that will show you 100th's of a second accuracy, but if you need that you should buy a professional system.

There are a couple of issues that will impact the quality of the system:

- accuracy of gps (lat/long). This is not such a huge problem, because the absolute accuracy might not be great, but relative is much better. If you want to make a plot of location on track for lap analysis it could be a problem, though.
- data rate. A normal gps only delivers a 1 Hz signal, so that reduces your time accuracy to full seconds only. You have two options to enhance this: data interpolation or a gps with a higher data rate (Garmin 16A, for example can do 5Hz).

I'm in the process of writing a lap timer application. I've got a couple of ideas to do this, not sure yet which idea will work best but I'll test all of them and see what works best. All ideas are based on marking the start/finish while driving. This is very important as it will give you a location and heading. This also assumes start/finish is on a straight piece of track.
One way to do it is by plotting a virtual line at a 90 degree angle from the finish lat/long point. Width of this line should be at least as wide as the track, but not so wide it would cross the circuit at another point!
Another way is to compare the last two coordinates and see if the finish point is in between those two. If the finish point was selected while driving at the outside of the track and another lap is driven on the inside it is possible that either lat or long positions do not end up above and below the finish point. So you would need a check to see if it could be close enough.
Now all that is left is to make the result better than just full seconds. I think the solution could be to check lat and long delta between the finish point and the previous measurement.
You then know which one has the biggest delta, for example latitude. When you pass the start line you take the latitude of the last measurement before the line, the first one after it and the one for the finish line itself. And since you know the time between two consecutive measurements is always 1 second you can determine the fraction. But so far this is all just theory, so the accuracy still needs to be tested. :-)

If you have a look at http://www.racelogic.co.uk/2003/vbox/ you can see it can be done very accurately. Ok, they're using much better equipment, but they aim for a much higher accuracy so that's not very surprising. For trackdays I think normal gps equipment should be enough.

simontdc · Posted: Mon Feb 02, 2004 6:40 pm Post subject:

I agree with the last post, simple interpolation can be carried out.

I have developed a 0-60 timer using GPS with 0.1s resolution which is pretty accurate, interpolating the start point within the starting second and the 60 crossing point automatically.

http://www.tdc.co.uk/marketing/press/0603_mdt800_article.htm for more information.

DT2 · Occasional Visitor Joined: Jan 31, 2004 Posts: 4

I've got a working version now and the results are not bad. I've modified the ideas a bit, and the thing that works best is:

at the time of marking the start finish point I check which has the biggest difference since the last measurement, lat or long. I also store the lat/long position and the heading.
At every point I now check only the direction with the biggest difference at start/finish and see if the finish is between the last two measurements for that direction. If it is, the other one just needs to be close. How close is something I'll make a user setting (depending on width of the track and if you want to add the pit-lane, for example). Now, the last check is the heading. This needs to be close to the original one, let's say within 5 degrees. So doing donuts on the finish line does not count as a lap :-)
The main reason for this last check is tracks with a hairpin at the end of the straight. If you make the width of the virtual finish line too big you could end up with a false finish-crossing if you don't take the heading into account.
As long as the start/finish is on a straight piece of track this works very well. Had a fun Sunday, driving laps through town all day to gather data and then back to the pc to test :-)
Hopefully I'll soon be able to test it out in a car with a decent measurement system and see how accurate it is.

	PocketGPS
	Web