Welcome to Netrider ... Connecting Riders!

Interested in talking motorbikes with a terrific community of riders?
Signup (it's quick and free) to join the discussions and access the full suite of tools and information that Netrider has to offer.

GPS Speed Check, accurate?

Discussion in 'General Motorcycling Discussion' started by Doggy, Oct 12, 2006.

  1. Hey all, I got a bit concerned about all this speed camera - speedo accuracy blah blah that I thought "Hmmm, might check that myself".

    I know some people who use GPS sets on their bikes and they give a km/h reading so I jammed the work Garmin eTrex Venture GPS under my screen and went for a bit of a ride.

    Results OMG :shock:

    Steady 40km/h on speedo = 34km/h on GPS

    Steady 60km/h on speeds = 52km/h on GPS

    Well that blows goats, either the bike is way out or the GPS is or both. How on gods green earth are we supposed to know exactly how fast we are going? The cops and cameras expect us to. If the above is to be beleived than my speedo is more than 10% out and does not comply with the ADRs. It's stock like the rest of the bike :roll:

    So the question is:

    Are GPS devices accurate to measure speed?

    How can you "fix" your bikes speedo to be accurate?

  2. GPS is usually very accurate.

    10% error is about standard on car speedos...and they usually err on the side of underestimating speed.
    Tyre pressure can influence the accuracy of the measurement - might pay to check yours

    and I believe there's a gizmo called 'speedo healer' which can recalibrate the speedo to be accurate.

    the el cheapo option is to fit a bicycle speedo, measure your wheel circumference really accurately and get an accurate readout that way.
  3. Speedo is a cable driven jobbie off the front hub.

    How do you attach the magnets on the wheel for a pushy speedo?
  4. Doggy, your speedo, like most bike speedo's, is optimistic. If you ride around at the speed limit as indicated on your speedo, you'll never get pinged.

    My GPS shows my 9R is slightly optimistic under 100km/h, but above, it has a hero factor dialled in, e.g. 109km/h actual = 120km/h indicated.... and get's worse from there.

    Speedo healer fixes the speed, but then your trip meter goes out of calibration.
  5. It's a real pain. Against my Garmin eTrex Summit, I get:

    Actual (GPS) speed............. 60 100 Km/h
    Indicated (speedo) speed.. 65 106 Km/h

    If I travel at the speed limit as indicated, I get hassled by traffic wanting me to speed up.
    So I do the math in my head, which is one more distraction from safe riding. Even then, when riding at the theoretical speed limit, I am constantly overtaken by cars. Never happened in the 4WD, which has a very accurate speedo, as tested against the GPS.
    I reckon people just don't like being behind bikes, so they pass any they can catch. Makes a mockery of the law and the government attitude to speed.

    Oh yeah. There are people on the web that print new speedo faces for cable driven speedos, so that you can "adjust" the indicated speed. Don't have a link, but a google should find one.
  6. Roderick, that's an unroadworthy situation. I think your speedo is broken bro.
  7. Hmmm makes me wonder if my speedo is right.. guess I'll tail a friend and get them to stay at a steady speed.. assuming their speedo isn't similarly afflicted :)

    But don't be too quick to trust GPS speed, its still a technology in development and often wrong. Think of the times it tells you that you are exactly where you want to be to the metre but what you typed in is 100metres down the road. Obviously its a completely different problem, my point is that it does make mistakes.
  8. That's related to crap mapping software. Not the GPS. If you have a strong enough connection to several satelites, most GPS will be accurate under 5m.
    The velocity measurement is incredibly accurate given no change in elevation. If you're riding up or down a large hill then it will be off due to the extra distance travelled vertically.
  9. What OBM said.

    I was just trying to find the GPS thread that explained the ins and outs of speed versus location, but it seems to be elluding me.
  10. depending on the rate at which it takes readings, it is generally only useful on a very straight road and for short intervals, +- 5 or 10 metres at each end could get you speeds that are not even close to what they should be.

    That said, i've not yet owned a japanese bike where I could not happily add another 9 or 10km/h at speeds over indicated 80 and still have the stopwatch tell me the speed was under the limit.

    the german bikes on the other hand have tended to be spookily accurate in comparison.
  11. So if according to the ADRs speedos need to be within 10% but not read lower than actual speed then my bike is unroadworthy as it does not comply with the ADRs. At 40km/h indicated it's doing 34km/h on GPS therefore 15% out. Opps must be a lot of bikes out there that are unroadworthy. :roll:

    I couldn't find any fixed speedo faces on the net for the GS500 so the only other option is a cycle computer. Anyone know the best value one and how you are supposed to mount them. :cry:

  12. around Perth the cops put up little speed testing radars that are connected to a big LCD display that shows your speed as you pass. Now i wouldnt put money on this being 100% accurate but if you calibrate your speed to those things then you'd be fine.
  13. Never seen one of those big billboard radar things in QLD. Thats not to say they are not here. I just can't remember seeing them.
  14. They say, as also mentioned here, that GPS accuracy is at its best when on a flat road at a constant speed. Therefore, this is what I did to install and calibrate a cycle computer.

    1. Set the CC's wheel size by measuring front tyre circumference.

    2. Then take it for a run along a nearby, quite flat piece of road that's about 4 km long. Reset trip computer on GPS and CC before starting your run.

    3. Ride the 4 km, compare speeds, stop. Check trip meters on both. Recalibrate CC as required. Go for another run.

    4. Compare readings. Recal as necessary.

    I have had two Sigma Sport BC800s


    and an Echowell F7


    which is on my current bike. Both worked fine once calibrated.

    I glued a "rare earth" magnet, bought in tandy, to the hub of one of the front discs and located the sensor using cable tied. Glue used was Locktite. Forget the actual product, but it supposedly glues anything.

    The bike's speedo is out by about 5%. This includes both the odometer/trip meter and the speed indicator. For the Blackbird, someone suggested an 18 tooth front sprocket that fixes the anomoly.

    I might try it when the chain/sprockets need replacing. Then there's the speedo healer.

    On the B'Bird with a digital dash, it corrects both speed and odo/trip meter error. Dunno about my analog dash. It too is electronic, so the healer thingy may fix both.

    I've checked the GPS and the CC against a roadside trailerable speed indicator which was on the road to work for around 2 weeks. With the cruise control on at 100 km/h on the car's speedo, the GPS read 99-100 km/h. The speed indicator displayed 99 km/h. It flashed "well done" at me. I then tried it the next day at 101 km/h (a bee's dick thickness on the speedo). It read 101 km/h and "slow down, too fast!". (At speeds higher than 110 it stopped displaying anything other than "slow down etc...)

    With respect to GPS accuracy, I had it (Garmin Etrex) in the car. Doing about 112 km/h indicated on the GPS, car speedo about 1-2 km/h faster. Rounded a bend on the Princes Highway near Orbost. Speed camera got me.

    The fine when I got it read:

    Detected speed: 112 km/h
    Alleged speed: 109 km/h

    So, other than the fine, I was happy with that...

    That is probably the only real world test that we're interested in; what it reads compared to a speed camera. Having said that, I got pinged coming out of a town. I saw the camera, saw it flash as I glanced at the GPS and CC which are next to each other on the top triple clamp. They were reading 68 km/h.


    However, the fine when it arrived said 70 km/h. Which meant that the detected speed was 73 km/h, a good 5 kays over what I saw on the GPS and CC.

    Can't fight it as I can't use the GPS to prove my speed, as it's not considered a legitimate scientific instro for speed detection purposes. I can download the logs and print out the time/date stamp on it. However, I must admit that I'd have to be selective as to what "evidence" I'd produce on the ground that I might've incriminated myself....
  15. Here in a couple of places in NSW we have 5 kilometre stretches of road with a sign every (genuine??) kilometre so you can check your speedo. I didn't see any on my recent trip to Vic; maybe Mr Bracks doesn't WANT you to know how accurate or otherwise the speedo is...
  16. i wonder if this way of detecting speed is a peek of things to come in relation to law inforcement and speeding. if it isnt already a factor? i wonder what sort of google earth the authoritys have?
  17. Yeah Hornet, I remember many years ago seeing a similar sort of set up on the sunshine hwy around the Bribie turnoff perhaps?

    I think I'll go for a new cateye. The magnet should be able to be attached to a spoke on the front disk. I'll start looking around.
  18. According to the old ADR rules, of +/- 10%, it is okay. That is, 60 vs 65Km/h, and 100 vs 106Km/h are both within 10%, and both reading under the actual speed. However, I have heard that the ADR rules changed for vehicles first registered after July 1, 2006, so it may not meet the new rules. My bike was first registered in August 2006. I haven't looked for the new rules yet. Anybody have a link?

    Is that what you meant, or was it just my crap formatting made it look like I was comparing 60 to 100Km/h?

    By the way all (bonox), at least my GPS is extremely accurate in the speed measurement stakes. It even seems to compensate for non-flat roads. That has been confirmed against travel times and freeway speed checks (on bridges etc.) The specifications claim an accuracy of 0.1 Knot RMS (steady State), while the positional accuracy is only 15m RMS. Think about it. My GPS calculates the speed every second continuously. If travelling at 100Km/h = 27.77m/s, and the positional accuracy effected speed calculations in the worst case, ie +15m at the start, and -15m at the end, I would have a reported change in postion of -2.22m over the second, and therefore see a negative velocity. Over the next second, having started -15m from my actual position, worst case is that I finish +15m from my actual position, therefore my reported speed would be something over 200Km/h. I would see wild variations in speed if this was happening. I don't, the speed measurement is rock solid, and responds almost instantly to any actual change in speed. It also reliably allows me to adjust speed by as little as 1Km/h based on its display. The positional errors are averaged out, or are consistent in direction and distance, so the speed is accurate. I'm sure my explaination wouldn't hold up with a math professor, but we don't have any reading this do we?!

    Also, GPS positional information available to private users was once very poor, as the Americans deliberately sent bad data from the satelites, so no other countries would have the accuracy of data they had. They no longer do this. At the same time, the positional calculation methods have improved dramatically. The repeatability of my GPS positions is very good, well within 5m, so when I see my position out by more than 15m against a map, I suspect the map, not the GPS. Map makers admit moving stuff around to make it fit.
  19. There used to be quite a few of those around, particularly on the old Hume Hwy. They don't seem to have made it onto the new Freeway. You could be right about Bracks (and VicRoads).
  20. Yeah whenever I'm in the cage I always drive to the GPS speed, which is about 10% fast. I don't know what I'll do if I'm caught, but its working at the moment. :)