Suzuki Burgman USA Forum banner

1 - 20 of 21 Posts

·
Registered
Joined
·
317 Posts
Discussion Starter #1
I know the speedometer is about 5-10 percent optimistic, but is the odometer also off by that much? If it is, then can we accurately calculate real gas mileage, or real mileage for tire wear?
Am just curious about that. Thanks.
 

·
Registered
Joined
·
585 Posts
The odometer is actually very accurate. I've checked it against mileage markers on state highways - those markers are very accurate in metering off one mile at a time. I suspect a straight road will be more accurate. IIRC the odometer on my 650 was off by one tenth in 10 miles - that was with the original stock tire. That is a very small error. I haven't checked it since putting on the new back tire which corrected the speedometer by about half as checked against my GPS. I'll have to do that.
 

·
Registered
Joined
·
1,204 Posts
According to my GPS, the odometer is off about 2% while the speed is off 10%.
 

·
Premium Member
Joined
·
14,278 Posts
The odometers on both my bikes are off about 1%.
 

·
Registered
Joined
·
1,113 Posts
I'm thinking the 1%-2% error is caused by tire wear. I've got an appointment for a new front tire in the morning. I'll check it and see what's up.
 

·
Registered
Joined
·
2,936 Posts
I'm thinking the 1%-2% error is caused by tire wear. I've got an appointment for a new front tire in the morning. I'll check it and see what's up.
That won't help, since the speed signal comes from the rear wheel.

My SpeedoHealer is set to minus 7.9% resulting in a dead-on speedometer, and the odometer then counts 5% to short. So it must read somewhat to long to start with. You do the math.
 

·
Registered
Joined
·
1,113 Posts
Really? Thanks, Eric - today I learned!
 

·
Registered
Joined
·
2,787 Posts
GPS devices are notoriously inaccurate at recording distance traveled; hikers and bicyclists deal with this all the time, with reported errors of up to 10% over relatively short distances (5 to 10 miles). This is due to the inherent inaccuracy of GPS positioning of 3 to 10 meters. Garmin claims "typical" accuracy of <3 meters for it's WAAS (Wide Area Augmentation System) enabled units. Note that this is the position accuracy for each sample taken, a circle of uncertainty around your actual position.

Because of this, and hills and turns where distance traveled is the straight line between samples (typically taken at 1/second), and any internal inaccuracies of the device, GPS distance traveled records are at best 2 to 5% accurate even with WAAS.

Velocity calculations do not suffer from the accumulated positioning errors and are much more accurate. In North America GPS devices with a WAAS capability, which most have, are accurate to +/- 0.05 m/sec or 0.1 mph...
 

·
Registered
Joined
·
2,936 Posts
I did the comparison trip meter vs GPS distance over 3 or 5 gas fillings, and it was consistantly around 5% more GPS vs trip meter distance.

I only add the 5% to the distance when I calculate gas mileage, to get it better than 20 kilometers per liter. :)
 

·
Premium Member
Joined
·
14,278 Posts
I agree with Cliff, distance accuracy on a GPS isn't all that good. Comparing the odo to my gps gives varying results. Sometimes the difference is wide and sometimes it is near the same. It seems to vary the most when riding the twisties.

That makes sense when you consider the GPS calculates distance as straight lines between points which understates how far you have gone in corners. The bike odo counts distance in tire revolutions times circumference of the tire when upright. But when you lean the bike over the circumference is smaller so you don't go as far as the odo thinks you did. That causes the odo to overstate distance in corners. The difference is small in any one corner but compounds for a day of riding lots of corners.

I calculate the error in my odo by comparing it to mile markers over a long distance on relatively straight roads. Using that method I'm consistently off about 1 mile in 100 miles.
 

·
Registered
Joined
·
4,876 Posts
My bikes are pretty consistent with all this. The 400 is about 5% greater than the GPS, and the 650 is about 2% positive.

Now whether the GPS is accurate, aperently that's another matter...
 

·
Registered
Joined
·
2,787 Posts
A picture being worth 1k words, here is a plot representing a vehicle traveling 60 mph over a straightline distance of one mile, with it's actual and "GPS" positions sampled every 88 feet--representing a typical GPS' sampling rate of 1 S/s:



The GPS' circle of uncertainty (COU) is assumed to be <3 meters (9 feet). The GPS positions were plotted with random position "errors" of ±9 feet for each the X an Y axes, with those simulated values being limited to a falling within a 9 foot "circle" surrounding the actual position.

Running several random simulations I found the GPS "accuracy" to range from -0.3 to +3.9% with negative errors being rare, fewer than 1 in 30 simulations.

The average error in 100 simulations was +2.3%.

Why this occurs should be obvious from the plot, most positional errors cause the GPS' calculated distances traveled to be greater than the actual distance.
 

·
Registered
Joined
·
1,204 Posts
That makes sense when you consider the GPS calculates distance as straight lines between points which understates how far you have gone in corners. The bike odo counts distance in tire revolutions times circumference of the tire when upright. But when you lean the bike over the circumference is smaller so you don't go as far as the odo thinks you did. That causes the odo to overstate distance in corners. The difference is small in any one corner but compounds for a day of riding lots of corners.
The vehicle GPS does not calculate based on straight lines (the GPS distances used for calculating advertised distances to stores and things use straight lines. It tries to send me to places over an hour away through the mountains when places within 20 minutes of me on the interstate are listed later). I calculates based on the road. I ride almost exclusively on very crooked mountain roads. The distance between Marion, VA and Tazewell, VA (Back of the Dragon) is 32 miles on Rt 16 and the GPS calculates and measures it that way. The distance between the towns is probably no more than 12-15 miles as the crow flies. I think the error between the odometer and the GPS is more to do with the size/wear/inflation of the tire, the inside of the curves vs the outside of the curves vs the center of the road, etc. I think the deviation of 1-2% is amazingly accurate.
 

·
Registered
Joined
·
2,787 Posts
^ This is correct for most automotive type GPS' when navigating as they use an amalgam of "map mileage" from the internal map database and calculated distance between sampled positions.

However when not navigating and just tracking trip info the device's calculate and accumulate distances between sample points--often with a lot of "fudge" added to some algorithms to reduced aggregate error.
 

·
Premium Member
Joined
·
14,278 Posts
The vehicle GPS does not calculate based on straight lines (the GPS distances used for calculating advertised distances to stores and things use straight lines. It tries to send me to places over an hour away through the mountains when places within 20 minutes of me on the interstate are listed later). I calculates based on the road. I ride almost exclusively on very crooked mountain roads. The distance between Marion, VA and Tazewell, VA (Back of the Dragon) is 32 miles on Rt 16 and the GPS calculates and measures it that way. The distance between the towns is probably no more than 12-15 miles as the crow flies. I think the error between the odometer and the GPS is more to do with the size/wear/inflation of the tire, the inside of the curves vs the outside of the curves vs the center of the road, etc. I think the deviation of 1-2% is amazingly accurate.
By straight lines between points I was talking about the lines between sample points on a curve not a straight line from beginning point to end point of a route. The GPS sees the curve as a series of connected straight lines not as a smooth arc. Each one of those straight lines is slightly shorter than the arc it represents. Individually those differences are minute but collectively over long periods they can be significant. Throw in the inherent problems with exact position of the sample points that Cliffy talks about and the equation becomes even more complicated.

By comparison the calculations the odo is doing are much more straight forward. It's just assumed circumference of the tire times revolutions. Complicating factors are tire brand, tire wear, tire inflation, and lean angle all of which can affect circumference.

Lane position in a corner would only be a factor if you are comparing the calculated distance to some predetermined measured distance like map distance or distance between mile markers. If you are comparing calculations of distance actually traveled between odo and gps they should both be affected the same way by lane position.
 

·
Registered
Joined
·
709 Posts
Yes, the tests I ran over 10 mile markers indicate that the odometer is equally inaccurate to the speedo... Your MPG's are also exaggerated because of it. When you're bike reaches 100,000 miles, it's actually only about 90,000 miles... If your bike reads 56 miles per gallon it's probably more like 50. Which isn't bad!
 

·
Registered
Joined
·
2,787 Posts
The odometer on my '03 400, with a nearly new Avon Viper Strike, was quite accurate (+0.5 miles) over 100 miles (versus mile markers) on a stretch of I-95 between Saint Augustine and Cocoa Beach--0.5%. Mile markers on the Interstates are laid out using rolling wheel odometers, as such they are quite accurate.

As Suzuki doesn't want or need any lawsuits over warranties and odometer accuracy over the covered 12,000 miles that is what I would expect. The speedometer was off by +10%.

That was before I installed the SpeedoDRD of course...
 

·
Premium Member
Joined
·
14,278 Posts
Same with the odometers on all three of my Burgmans. Accurate to less than 1 mile in 100 miles when checked against the mile markers on the Interstate. That would make the difference in 100,000 miles more like 1,000 miles not 10,000 miles. In other words I have gone at worst 99,000 miles instead of 100,000 miles.
 

·
Premium Member
Joined
·
6,179 Posts
Buffalo as your up to 103k now that means you are in the TRUE 100k miles club. :D
 
1 - 20 of 21 Posts
Top