I love my iPhone!

[Kendog]

I've been thinking about this. I'm surprised that the sampling interval on a GPS would be accurate enough to accurately time a lap.

If you crossed the finish line doing about 100Km/h and the GPS had an error of 3m, then that would be an error of about 1/10 of a second. And if it only took GPS samples every 200ms, then that would be another 2/10 of a second.

I guess that is fine if your consistency is such that you lap time varies quite a bit from lap to lap.

What does the application say its accuracy is on an iPhone?

From what I could see it is not going to be the most accurate measure of lap times. But if you have a iPhone it is a cheap way to get some form of lap timing working.

Here is some info about the app.

[TZ350]

So we downloaded a lap timer for $16 onto my iPhone which works off GPS coordinates and times your laps for you.
Taped the phone to my bike this morning and went for a ride around Kaitoke track to see how it worked. It was wicked!!

Hi Trudes, that's very interesting, thanks for sharing it. Getting lap times is a real issue for us too.

.

[F5 Dave]

Hiya, sounds good measuring stick.
I've been having a think, so I hope things can work out for Tues arvo, I've got some ideas to instill.

[Trudes]

Hiya, sounds good measuring stick.
I've been having a think, so I hope things can work out for Tues arvo, I've got some ideas to instill.

If it's fine I'm keen!

[p.dath]

Found this interesting article about the GPS accuracy for the iPhone.
http://blog.teeshotlive.com/2008/08/...s-limitations/

I suspect you might find the timing data will vary from lap to lap quite a bit, even with exactly the same lap times.

[p.dath]

Been doing some Googling. I see this GPS based lap timer seems highly regarded, but is has a purpose built GPS for the task:
http://www.racelogic.co.uk/?show=VBOX

And this unit also seems highly regarded.
http://www.aimsports.com/products/tg/index.html

[Trudes]

Thanks for that p.dath, I did look into getting a proper transponder type unit, however, considering I am an extremely slow bucket racer, not competitive and never will be, I kind of figured I didn't need to spend $500-$1000 on a piece of equipment I would probably use for about two hours every month, it just doesn't matter that much to me.
The readings I got off the iphone seemed accurate enough for me, I rode around slowly and got a time, sped up a bit, got a faster time, sped up some more and got an even faster time, went about the same speed for the next four laps and got roughly the same reading within a second or two for those laps. So it seems to be working. I don't need all the flash information about acceleration, g-force etc etc, all I need is some indication of about how fast I rode in the last lap to encourage me to go faster in the next. And it apears to be doing that from what I can tell... all for $16.

[cmac]

I am building my own GPS lap timer and have done quite a bit of research and experimentation. What I have found is:
(1) Accuracy of location.
From my experiments, I have found that even if the absolute accuracy of out by say 5 meters, the relative accuracy from measurement to measurement is very high. When I plot the readings I get, there is no apparent error between readings. For example, if I travel at constant speed in a straight line and then plot the readings, the plotted readings show equal distance between plotted points, and no lateral deviation. The same happens if I travel in a constant radius turn at constant speed - no deviations from the expected. If I accelerate at a constant rate, I also get accurate plots.
(2) Accuracy of determining the moment the Finish line is crossed.
Even though the GPS location is detected once per second, interpolation is used to calculate the finish line position. If the acceleration at the finish line is zero then this will give a very accurate position (time). Given that most finish lines are at a point where there is very little acceleration (or braking) this gives an accurate lap time.

So ...
Given that we are not up with Rossi or Stoner et al and can crank out lap times that vary by only a few hundreths of a second, the information provided by GPS lap timing can be very useful.
Go for it Trudes

[Trudes]

Good info cmac, and thanks!

[p.dath]

(2) Accuracy of determining the moment the Finish line is crossed.
Even though the GPS location is detected once per second, interpolation is used to calculate the finish line position. If the acceleration at the finish line is zero then this will give a very accurate position (time). Given that most finish lines are at a point where there is very little acceleration (or braking) this gives an accurate lap time.

So ...
Given that we are not up with Rossi or Stoner et al and can crank out lap times that vary by only a few hundreths of a second, the information provided by GPS lap timing can be very useful.
Go for it Trudes

Except that you accelerate off the line when you start ... If you are doing lots of laps this error will become smaller. If you are doing a single lap it could miscalculate the start line by up to 1s.

[cmac]

Except that you accelerate off the line when you start ... If you are doing lots of laps this error will become smaller. If you are doing a single lap it could miscalculate the start line by up to 1s.

You raise a very good point - I overlooked this in my description above. The accuracy of the all lap times except the first lap is mainly dependant on the interpolation code that is used. The accuracy of the first lap is not only dependant on the interpolation code, but also dependant on the position of the finish line compared to the start line - for reasonably accurate measurement of the first lap, the finish line must be past the start line so that you have speed measurements before and after the finish line. At present, I am undecided as to whether I should treat the first lap differently to the remaining laps - experimentation will answer this I guess.

[Trudes]

I don't understand what you're talking about, but does it have anything to do with why the lap timer on the iPhone gets you to set the start finish line then move 30 meters up the track where it sets something else and where it gives you the reading of your lap? I can't figure out why it gives me my lap time at this 30 meter mark and not the start finish line.

[cmac]

I don't understand what you're talking about, but does it have anything to do with why the lap timer on the iPhone gets you to set the start finish line then move 30 meters up the track where it sets something else and where it gives you the reading of your lap? I can't figure out why it gives me my lap time at this 30 meter mark and not the start finish line.

Sorry Trudes....
Its a bit complicated to explain ... but here goes.
The reason you have to mark a point, then move 30 meters down the track and mark a second point is because.....
The GPS unit calculates a point on the earth. Each GPS reading (taken every second for most GPS units) calculates where this point is. When you cross the finish line you probably won't cross at the exact same point as the previous time you crossed the finish line (in fact is is very unlikely that you will be at the same point at the time the GPS takes the position).
Because of the width of the track at the finish line, the finish line is a line (not a point) and you won't cross this line exactly at the time (or point) that the GPS takes a reading. To overcome this problem, the program has to calculate a line that is at right angles to the track. To calculate a line at right angles to the track, the program must know what direction you are heading (your heading) when you cross the finish line. To work out the heading, you take one reading, move down the track and take the next reading - that gives you the heading. By the descriptions you give, it also seems that your program also uses this second point as the finish line (by the way, I plan to use the same method with my GPS lap timer). As d.path points out, using the start line as the line where the laps are measures leaves room for an inaccurate result for the first lap. The way to overcome this is to have the finish line after the start line so that you are moving when you cross the line used for measuring - hence the finish line (the place where the lap times are measured) is past the start line.

I hope this helps you understand - if I had a white board I could explain it a lot easier!

[Trudes]

Thanks cmac! Yep it does make some sense to me. That was one of the issues I'm wanting to iron out next time I have a play with it as when it gives me my lap reading (30 meters from the start finish line) it just happens to be on a corner, so I'm trying to look at the lap timer and corner and set myself up for the next corner right after all at the same time. Was thinking of taking the first reading 30 meters back from the start finish line and have the start finish lne as the point where it gives the lap time to combat that problem some. Wonder if that would work. Trial and error I guess.

[cmac]

Having the 30 meter mark on a corner does mess things up a bit!
What track do you race on? Is it Kaitoke? I see that you are going to be at the endurance race there. So am I - can't wait. I broke my wrist a few weeks ago so it will be my first race back.

I am guessing that as long as the 30 meter mark is after the start line, it should work OK, so perhaps try taking the first mark 20 meters before the start and the second 10 meters after the start and see if that works.