How accurate is the SV speedo?

[kaivalagi]

I think the original issue of speedo accuracy was more centered around the deliberate adjust up in speed of the readings as imposed (indirectly) by law, which there is general agreement on. I think we got a little sidetracked with details

BTW I still need to find out if distance recorded by the odometer is 9/10% higher than actual just like the speedo...if it is that's just plain wrong... different bikes with varying accuracies would record varying falsely inflated mileage...a speedohealer at £60(?) might be worth it if that's the case...

Originally Posted by squirrel_hunter

I've never said that GPS is inaccurate, nor do I think its inaccurate. I have simply been wondering how it is so accurate compared to a speedo. I have previously read that GPS is more accurate than a speedo, but no one has provided the answer as to why. So when it popped up again I asked and kaivalagi and mikerj have been kind enough to provide me with an answer, thanks guys.

For the speed calculation in a satellite based GPS system there are two sets of measurements taken, Frequency Difference of Arrival (FDOA/Dopplar Shift) and Time Difference of Arrival (TDOA). Frequencies shift and become longer as they get further from the source. The same signal travels from your GPS (civilian GPS is two way) to multiple satellites. The satellites pick it up measure the shift in frequency and compares the time of arrival with that of other satellites. That's where the geo-location comes from. Now for speed. the updates are usually going to be on a set timer, say every 3 seconds. If your position changed x amount in 3 seconds, you are going **mph. Add up the most recent 20 or so of those updates and you get an average speed. Yes, the location may be off, but software has programing in it to exclude locational data that doesn't meet it's confidence requirements.

Originally Posted by kaivalagi

BTW I still need to find out if distance recorded by the odometer is 9/10% higher than actual just like the speedo...if it is that's just plain wrong... different bikes with varying accuracies would record varying falsely inflated mileage...a speedohealer at £60(?) might be worth it if that's the case...

I'm a little blurry about the details, but I was working out what my Odo should say for different checkpoints on my trip and it seemed pretty close. I would like to see a larger sample size than 30-50miles though.

Originally Posted by squirrel_hunter

I've never said that GPS is inaccurate, nor do I think its inaccurate. when it popped up again I asked

Originally Posted by kaivalagi

If you still don't buy into the concept I am not sure what else I can say to convince you....

Fairy nuff, it just looked a lot like you were in dissent with the concept.

[squirrel_hunter]

Originally Posted by kaivalagi

I think the original issue of speedo accuracy was more centered around the deliberate adjust up in speed of the readings as imposed (indirectly) by law, which there is general agreement on. I think we got a little sidetracked with details

True.

Originally Posted by kaivalagi

BTW I still need to find out if distance recorded by the odometer is 9/10% higher than actual just like the speedo...if it is that's just plain wrong... different bikes with varying accuracies would record varying falsely inflated mileage...

Now thats an interesting point.

[kaivalagi]

Originally Posted by sath182

I'm a little blurry about the details, but I was working out what my Odo should say for different checkpoints on my trip and it seemed pretty close. I would like to see a larger sample size than 30-50miles though.

Thats interesting, from a design perspective I would expect the speed inflation to only be on the screen and not to affect the odometer too so hopefully that is the case.

Today looks to be a nice day, if I finish work early enough I'll go on a little bimble for half an hour with my tracks running and hopefully have something to go on....even if not very accurate we should see a distance at one end or the other of the accuracy scale

[kaivalagi]

I went out for a ride and although overcast and not great for my crappy gps on my phone I got the results as follows:

Odometer: 37.8 miles
GPS on Android via My Tracks: 36.83 miles

That's around 2% difference which makes me believe that the odometer is not inflated by 9/10% like the speedo is....Also the GPS is most likely reading low because of lost signals around bends etc that would give straight line distances where this is not the case...see the image below, where there is no blue the signal was lost...



I will give this another go when the weather is better, like I say it was overcast today, I nearly didn't bother as I didn't get a gps fix from home, I tried again once I'd travelled for a bit



Anyway I think this raises another question, for all those that have fitted speedohealers are the odometer readings on their bikes lower than they should be because of the reduced speeds on the clock !!!??? That's for someone else to figure out, I'm never going to bother fitting one.....

Originally Posted by kaivalagi

I went out for a ride and although overcast and not great for my crappy gps on my phone I got the results as follows:

Odometer: 37.8 miles
GPS on Android via My Tracks: 36.83 miles

That's around 2% difference which makes me believe that the odometer is not inflated by 9/10% like the speedo is....Also the GPS is most likely reading low because of lost signals around bends etc that would give straight line distances where this is not the case...see the image below, where there is no blue the signal was lost...



I will give this another go when the weather is better, like I say it was overcast today, I nearly didn't bother as I didn't get a gps fix from home, I tried again once I'd travelled for a bit



Anyway I think this raises another question, for all those that have fitted speedohealers are the odometer readings on their bikes lower than they should be because of the reduced speeds on the clock !!!??? That's for someone else to figure out, I'm never going to bother fitting one.....

Sounds like a good next step to me. Another one to try and increase the accuracy might be a long straight line and back. That should avoid some of the signal loss errors.

I'm not sure if an inaccurate odometer would matter anyway. All of the odo's on all of the bikes would be out by approximately the same amount if that were the case. Even if it one odo was accurate, and another was 10% out, you are not going to look at two bikes in autotrader, one with 30,000 and one with 33,000, and choose between them based upon mileage. You could say you would save 10% on your servicing costs, but even at main dealer prices, that's £20 every 4000 miles, and it's not a bad thing if your bike gets serviced more regularly than the schedule anyway.

[kaivalagi]

Very true, I was getting carried away with the detail, I now know that the odometer is accurate where as the speedo id inflated...that's a nagging question put to bed anyway

Atleast if there is a discrepancy it's the right side of actual to help not hinder when it comes to servicing...