GPS or odometer?

[hoagie]

It seems every Ebike I have owned in is 5%+ optimistic vs the app I use. I’ll take the 33.3.

 

[mschwett]

check the circumference of your tires against the assumed value in the app. that’s a pretty big difference, and it’s unlikely that the GPS track is off by THAT much!

 

[dodgeman]

I got into a friendly argument with my bike shop on this very subject. I’m a retired land surveyor and used GPS a lot for my work. Through experimenting with the survey grade GPS, truck speedometers and my phone I feel phone gps is more accurate.

Phone gps isn’t very accurate, they say 10 meters but on average it’s probably more like 3 meters. Picture yourself standing in one place and your position is jumping around 3 meters, that’s not very accurate. I feel that through software advancements they have ”smoothing” techniques in the phones that make your location and speed more stable and accurate.

Sorry for the long answer but I feel the gps is more accurate.

 

[harryS]

A few bike displays and most all $10 bike computers will take the wheel diameter or circumference to 1 cm precision. Then compare the actual distance with the reported distance and refigure the wheel size. Your weight on the bike, and tire pressure will affect that. Maintain both and it's always going to be the same.
.
My limited use of GPS on phone is via Strava, which I didn't find that useful, but it was probably my phone losing connectivity. Half my attempts to use it rended with only partial recording or mileages I knew were too long, but I liked the silly names that Strava would assign to the little stretches of my ride.

 

[6zfshdb]

There are errors associated with both GPS and wheel circumference odometers. GPS measures the distance between acquisition points which are never in a straight line. The "snap to" or smoothing feature in many GPS receivers does help as @dodgeman stated above. At best, they are a series of straight lines which do not exactly follow the natural curves of a road or trail. This picture gives a visual representation of the effect:

The yellow center line of the road is the true distance measurement. The blue, red & green lines are GPS tracks made on four different trips along the road. The small squares are the acquisition points and the GPS distance is measured between them. It's easy to see why they don't match.

Wheel odometers rely on accurate circumference measurements based on the rolling radius of the wheel. This is very difficult to measure due to changes in surface conditions, tire pressure and rider position on the bike. You also have to take into consideration the minute distances added each time you turn the handlebars. No rider rides in a truly straight line.

Due to the difficulty in setting an accurate wheel circumference, I tend to agree with @dodgeman that, under most conditions, the GPS measurements are likely more accurate.

 

[mschwett]

we’ve had this (interesting) discussion quite a few times here over the years. the device itself really plays a huge role - a state of the art phone’s location services will use much more data than a cheap GPS navigation device, integrating a lot of information to arrive at your position with pretty good accuracy. the one second sampling can definitely undercount distances on very windy courses with sharp corners, but for typical road curvatures the error is vanishingly small.

another factor can be that distances are typically measured by projection to the WGS84 geoid, which is an approximation of the earth’s mostly spherical shape. in other words, it’s flat except over very long distances. if you ride 10,000 feet horizontally at a 15% slope up, and then back down, from a land surveying and distance measurement standpoint you’ve traveled 20,000 feet. but your wheel has actually rolled 20,225 feet - so even a perfectly calibrated wheel sensor in a completely straight line will be more than 1% different than a navigational measurement. add some wiggling back and forth within each one second sample period (the straight line between two points one second apart will always be shorter than the actual path) and the wheel-sensor result, when perfectly calibrated, should always be very slightly more.

however, the nearly 10% variance of the OP indicates either very bad GPS reception or an incorrectly calibrated wheel sensor. 1-3% is very normal.

 

[dodgeman]

I would say my odometer on my bike reads about 4% to long and the speedometer is about 8% to fast. The picture above is a pretty good depiction. In my mind since the gps is zig zagging it should read a little bit longer than a theoretical accurate distance.

 

[Jim1348]

I got into a friendly argument with my bike shop on this very subject. I’m a retired land surveyor and used GPS a lot for my work. Through experimenting with the survey grade GPS, truck speedometers and my phone I feel phone gps is more accurate.

Phone gps isn’t very accurate, they say 10 meters but on average it’s probably more like 3 meters. Picture yourself standing in one place and your position is jumping around 3 meters, that’s not very accurate. I feel that through software advancements they have ”smoothing” techniques in the phones that make your location and speed more stable and accurate.

Sorry for the long answer but I feel the gps is more accurate.

I have three Garmins, which are consumer grade. They all claim 10 foot accuracy. They receive GPS and Galileo (no GLONASS).

They are:

Garmin Zumo XT

Garmin DriveLuxe 51

Garmin DriveSmart 65 with Amazon Alexa

Someone suggested going to a nearby USCGS Classic Horizontal Control benchmark.

I might go find one on one of my bike rides and compare my Garmin Zumo XT to the benchmark and see how close it is.

National Geodetic Survey Data Explorer

geodesy.noaa.gov

 

[dodgeman]

Keep in mind some marks are horizontal only, which would be latitude and longitude (or state plane coordinates) and some marks are elevation only, and some are both.

 

[Jim1348]

Keep in mind some marks are horizontal only, which would be latitude and longitude (or state plane coordinates) and some marks are elevation only, and some are both.

This is what was suggested:

"Find a USCGS Classic Horizontal Control benchmark that has been surveyed to a high degree of precision using professional GPS surveying equipment. The accuracy is usually within .5 CM.

You can usually find one one of these benchmarks fairly close to your location by going here:

https://geodesy.noaa.gov/NGSDataExplorer/

Type your Zip Code in the search box and click go. Zoom in to view the benchmarks. Choose only those marked as "Classic Horizontal Control". Click on the mark and print out the datasheet.

Find the benchmark location in the field and center your GPS on top of the mark. Leave the GPS stationary for a few minutes and read it's location coordinates. Compare with those listed on the datasheet. The calculated offset is the true accuracy of your GPS.

Keep in mind the results obtained here are made under ideal conditions and are the best your GPS is capable of producing. They will vary greatly with location.

Finding these benchmarks in the field can be a challenge but they are always in open areas and usually well marked by local surveyors."

 

[dodgeman]

I wouldn’t limit my choice to “classic horizontal control”. The data on those isn’t any better than some others but might be. The classic marks are ones that were built and surveyed in the 1930’s where they did it all by measuring angles and calculated using triangulation. Most of these actually had towers built above them to see over obstructions. I’m a retired land surveyor so I’ve studied this in the past. A picture of a tower and a light on top that would have been sighted from another tower. All done at night to eliminate heat waves and have better visibility.

 

[PCeBiker]

I downloaded a GPS Speedometer app to my android phone and used it to calibrate my speedometer/odometer.

My hub motor is a mystery and I don't know the gear ratio or how many magnets it has, so I can't enter those parameters to set the calibration.

I rode down a perfectly flat straight road (easy to find around here) at full speed (32 kph) and adjusted my settings until the speeds matched.

I set my speedometer to read 33 kph when I was going 32 kph to keep me on the happy side of the law, then I deleted the app.
As soon as I go over 32 kph I'm risking $5000 in fines.

I broke the law today.
I was going 34.7 kph with a tailwind.
I'm a Hard Core Bad Ass.

I rode 45.4 km today.

It may have actually been 45 or 46 km.
Close enough for me, as long as I don't run out of battery.

 

[Stefan Mikes]

I was going 34.7 kph with a tailwind.

I rode 45.4 km today.

I can see a correlation above

Low average speed, long distance. Necessary to mention my top speed was downhill, not pedalling.
Of course you cannot ride that far on a single battery!

 

[Jim1348]

I downloaded a GPS Speedometer app to my android phone and used it to calibrate my speedometer/odometer.

My hub motor is a mystery and I don't know the gear ratio or how many magnets it has, so I can't enter those parameters to set the calibration.

I rode down a perfectly flat straight road (easy to find around here) at full speed (32 kph) and adjusted my settings until the speeds matched.

I set my speedometer to read 33 kph when I was going 32 kph to keep me on the happy side of the law, then I deleted the app.
As soon as I go over 32 kph I'm risking $5000 in fines.

I broke the law today.
I was going 34.7 kph with a tailwind.
I'm a Hard Core Bad Ass.

I rode 45.4 km today.

View attachment 158523

It may have actually been 45 or 46 km.
Close enough for me, as long as I don't run out of battery.

I would have expected Celsius on the temperature display.

 

[PCeBiker]

I can see a correlation above
Low average speed, long distance.

I don't really watch my speed too much.
I watch my Watt meter, and try to keep my output around 350 Watts or less.
I was going into the wind at 22-25 kph using 350-400 Watts but going down wind at almost 35 kph using as little as 150 Watts.

Necessary to mention my top speed was downhill, not pedalling.

I never pedal, and there are almost no hills around me to get my speed up.
It is kinda cool to see my Watt meter drop to Zero on a downhill though.
It's like I'm getting something for free.
(Yeah, I know, I spend way more than I get back going back up the hill.)

Of course you cannot ride that far on a single battery!

I have a single 48V 25Ah battery that has taken me 85 km on a single charge with no pedaling.
I'm sure it can go twice that if I pedaled and kept my speed down.

 

[Stefan Mikes]

I watch my Watt meter, and try to keep my output around 350 Watts or less.

So when you drive a car you do not watch the speedometer Well well well.
Why do you watch your Watt meter in the first place?

 

[PCeBiker]

I would have expected Celsius on the temperature display.

Lol,..
I grew up with Imperial before Canada went to Metric.
We switched to Metric when I was in grade 7 which worked out great because the math and science was starting to get difficult.

Metric is WAY EASIER !!
It's all factors of ten. Guess how many meters are in a Kilo Meter ?
I have no idea how many feet are in a mile. Too much to memorize.

Then in grade 12 Chemistry class, I learned that a 10 centimeter cube, filled with potable water at 20°C is Exactly 1 Liter and weighs Exacly 1 Kilogram.

That's Cool !!

If I'm ever stranded on a desert island and my only measurement tool is a ruler that is exactly 10 cm long, I can establish my own system of weights and measures that are legal for trade with the rest of the world.

However,..
I couldn't get used to the metric temperature scale.

I always went to school as a kid knowing that the room temperature is 72°F, then they reduced it slowly to about 68° or 69° to save energy.

Those numbers are more tangible to me.

PS,..
I keep my bike in the house where it's air conditioned.
It was actually 27°C outside today.
Apparently, that's kinda hot ??
I don't really know?

 

[PCeBiker]

So when you drive a car you do not watch the speedometer Well well well.

Yeah, all I do is keep checking my speedo and I don't like it.
I always try to keep it under 10 kph over the limit to avoid trouble and tickets.

Why do you watch your Watt meter in the first place?

If I drove a car that would only go the speed limit, I would ignore my speedometer.

I watch my Watt meter to know exactly how fast I'm using up my battery reserves.
My e-bike is locked at 32 kph (except with a tailwind) so I don't have to worry about the Po-Po.
I always wear a helmet, and I have turned my 25 amp controller down to half, with first throttle gear at 20% of that.

I am well within my 500 Watt legal motor limit, but I don't think about that.

I turn it down to extend my ride time and distance, as well as go real easy on ALL my bikes components.
I want my bike to last forever.

 

[Stefan Mikes]

I watch my Watt meter to know exactly how fast I'm using up my battery reserves.

Does not your e-bike have just Battery % meter?

 

[PCeBiker]

Does not your e-bike have just Battery % meter?

Yes but I don't use it, and I didn't bother calibrating it.
I just use the voltage to know how much battery I've got left.

Knowing how fast I'm draining the battery allows me ride as far and as long as possible.

If I see the power draw climb above 400 Watts, I slow down.