Performance testing accuracy

I appreciate that this function will never be that accurate but I have got very odd results from initial tests.

My car which will do 0-60 in around 3 secs, did 2 runs in the mid 7s. This was with a bluetooth gps receiver being used (BT-368 from GlobalSat). This was clearly miles off so I repeated the test with the internal gps on the phone (Nokia 5800) and it registered 4.4 s on a much less agressive attempt.

Have I got a useless external GPS receiver or do I need to change a setting ?

Thnx any help.



  • edited January 2010
    Unfortunately the feature is only as good as the GPS receiver used. Many GPS are very bad at standing starts. Some "stay put" the first second or so before registering speed correctly. That's why I don't market RaceChrono as performance meter...
  • Thnx reply - So, no setting to change, its just down to the receiver ?

    Dont worry - I am using this for lap timing - so the performance testing is just me playing with it to get used to what it can do. But in that, I wanted to find the best way of getting accurate lap results and if my receiver is rubbish I will get the one u recommend.

    The one I have also gave a very flat line on the graph - clearly wrong as it didnt reflect gear changes, which the internal gps did. So on track if the data isnt coming over quick enough it will be useless.

    I assume best bet is to just buy the Gps receiver recommended ?

  • edited February 2010
    no to hijack but I was planning on trying some drag races with the qstarz 818x. Does this work well or "stays put"?

  • edited February 2010
    it's one of the best cheap receivers.
  • so should I expect it to work for drag racing? Do you have many drag racing users?
  • edited February 2010
    It works. You may or may not get good accuracy. Usually the 1/4 mile accuracy is pretty good, within 0.2-0.3 sec diff to the actual. I have no idea how many users are using it for acceleration testing, but my guess is lot more than track day users.
  • Does it actually matter which QStarz receiver you use ? They all seem to have the same spec ? 818, 890 etc ?
  • They don't. Make sure you buy MTK-II chipset GPS that has 5 Hz, like 818X
  • Just having tried some Performance Testing with an OBD2 reader I'm curious about the precision when not relying on GPS but OBD speed instead.

    Tried some different runs, and the times seem plausible, but it would be nice to know what kind of accuracy can be expected.
  • Depends how much difference there is between the speed reported by GPS and OBD. If the difference is small, the OBD-II speed based performance tests are very accurate. On distance based tests, the OBD-II helps only at the start.
  • edited May 2010
    Hmm, I was using the built in GPS of my E5800 as I was mainly interested in the OBD data, so I'm not sure how accurate that is. The OBD graph certainly LOOKS a lot more accurate, but under heavy acceleration it shows a higher speed then my GPS, while under deceleration (just releasing the throttle after my run) it shows a slightly lower speed, difference of about 3km/h.

    The OBD data is updated about 3 times per second as far as I can tell, while the GPS data looks like interpolated 1hz data (since the GPS is 1hz, that's as expected I guess).

    Still not wiser though... I have slightly bigger tires than the original dimension (at least on my street tires), so that would mean the OBD speed is a bit lower then my real speed, 3km/h is probably in the region I guess (205/45R17 original and 215/45R17 on now).
  • GPS/OBD speeds need to be same on steady pace. During acceleration / braking the time shift affects too.
  • Ok, so to make sure it's accurate I should find a section where I've driven with constant speed (or make a session) and compare the speeds of OBD and GPS. And if the speeds are about the same it's accurate?

    How about if I find a slight disrepance, like one is a few km/h faster than the other? And if say the OBD speed is slightly lower, does that mean I have a slight error on the negative side (meaning my times are slightly worse than actual performance)?
  • Usually OBD speed is lower than GPS speed (real speed) which gives you slightly optimistic results.
  • Ok so I've checked a place when I'm cruising at a constant speed, the OBD shows exactly 2.6km/h slower then GPS speed. Which sounds about right considering my tires are a bit larger then stock. So that means my acceleration time should err on the pessimistic side, right? Since it will take slightly longer for it to reach eg. 100km/h.
  • Yep, pessimistic. You could create 0-98 kph (OBD speed) test to test 0-100 kph (real speed) when using your OBD adapter.
  • Ah, that's so obvious I didn't think about it at all hehe.

    Thanks for the prompt replies!

    Gotta take a look at those "Donate" buttons soon I think, but unfortunately not until next month because my car is going in for some work and it'll probably leave me on rice and beans for the rest of the month :-)
Sign In or Register to comment.