Bike speedo v GPS

[rmurphy195]

Well, I made it this morning - one of my getting fitter targets was to complete a 25 mile ride on a Sunday morning, and I did it today - YAY!

BUT- my bike speedo says 25.25 miles, and my Etrex (worn on a belt clip) says - 24.82!

So - should I rejoice in my 25, or recalibarate my bike speedo and admit to only 24.8 - decisions, decisions ...

[NATURAL ANKLING]

Hi,
I think you will find that that's about right in comparison between the two.

I normally measure the tyre accurately, radius or circumference, input size for speedo, take off 2cm, on circumference.

Compare your measured course with an accurate plot on a route planning website.

Say nav can be way off (mine comes short by 0.6-7 over speedo 36 miles).

Adjust your speedo to match webroute planner, I always err on the pessimistic side so I know I have done a real mile.

[robgul]

Chuck away the bike computer and stick to the GPS for just one reading!

Riding with a group we seldom find that the GPS gadgets have more than a whisker of variation on the same ride - that's Garmins, Brytons, Wahoos and even one of the MemMap satnavs.

The mere fact that you have to calibrate the computer with tyre size is a weak point - and a GPS with a separate mount on each bike you have (You only have ONE ) makes life simple.

Before anyone says it - blasting down a steep hill where there is heavy tree cover can sometimes trick the satellites and give a slightly false reading - but again, very minimal.

Rob

[RickH]

With the GPS units that can make use of them, wheel sensors can give you the best of both worlds and reduce/eliminate the speed anomalies you can get when you, say, ride through a tunnel. (My old Garmin Edge 605 - which didn't link to sensors - tried to claim I did over 50mph in the Chirk tunnel on the Llangollen Canal once! )

[bikes4two]

> I would think the bike speedo, if correctly set up in the first place (measure tyre circumference on a path using chalk marks etc) is more accurate than a GPS
> The bike speedo will record distance travelled rather than the distance between points, i.e. when taking a less than straight line along the road or zig-zagging (for what ever reason).

> Errors from the GPS due to processors speed rates, app accuracy and app algorithms, signal fade (buildings, trees, atmospheric conditions) and others.

> Having said that I use both and don't really care about discrepancies after all my years of cycling, even to the extent that I rarely look at the speedo ODO before setting off, relying on local route knowledge and how tired I feel after a ride which is good enough for me (but each to their own).

> Bottom line - as others have indicated, there are differences in the measured values but they are not of an order to both most folks.

[geomannie]

> I would think the bike speedo, if correctly set up in the first place (measure tyre circumference on a path using chalk marks etc) is more accurate than a GPS
> The bike speedo will record distance travelled rather than the distance between points, i.e. when taking a less than straight line along the road or zig-zagging (for what ever reason).

> Errors from the GPS due to processors speed rates, app accuracy and app algorithms, signal fade (buildings, trees, atmospheric conditions) and others.

> Having said that I use both and don't really care about discrepancies after all my years of cycling, even to the extent that I rarely look at the speedo ODO before setting off, relying on local route knowledge and how tired I feel after a ride which is good enough for me (but each to their own).

> Bottom line - as others have indicated, there are differences in the measured values but they are not of an order to both most folks.

I agree. You can measure the actual circumference of your loaded bike tyre to within a mm or 2. If your tyre circumference is around 2155mm, then even assuming a measurement error of 5mm you are within 0.25% of the true value, a trivial error. A 2mm error in diameter is about 0.1% off the true value.

Your computer will then reliably measure 2155 (+/-2mm) every time your wheel turns giving a very accurate distance.
With the GPS the fractal nature of measurement comes into play. The GPS calculates locations at timed interval and calculates distance by assuming a straight line between the points. While this is a good approximation, bicycles move in curves and curves are always longer than the straight line approximation.

For example, if you try to take the distance from a recorded GPX track, you find that it is actually made of straight lines segments. My Garmin Etrex gives segments from ~5-100+m in length. If you compare the segments to the road, the mismatch of the straight GPX segments on curved road is apparent. (see image, GPX track green, nodal points shown)

IMHO, GPX will generally under-read distance slightly, but not by a lot.

[rmurphy195]

Thanks for all your replies - they gave me some thoughts, resulting in my deciding to trust my speedo rather then the GPS.

Why? - several reasons

First of all , when I set up the speedo I reset it with the tyres currently on the bike (I do this each time)m - using the "chalk mark" method, rolling the wheel along 3 times then dividing the result by 3. So it should be accurate enough.

Secondly, my GPS is set to the "normal" time interval, which is fine for recording walks but the bike is a bit faster! When I look at the track on Basecamp I can see that the recording points go a bit awry when turning corners (most often cutting them off), and going along curvy roads. The "series of straight lines" shaves a bit off corners and curves, the cumulative effect (probably) accounts for the c1.7% difference. I came across this which was quite interesting https://wiki.openstreetmap.org/wiki/Garmin/eTrex_Vista_HCx, but I don't plan to alter the GOS unit 'cos we use it to log our walks!

My average ride speed is slower than many of you (about 10mph or so), I guess if I rode faster the discrepancy might be greater.

Why am I using the GPS as well as the speedo? I don't do it all the time - or even very often, but I do like to record my routes from time to time then look back on them (I use the GPS for recording what I am doing, rather then telling me what to do, being old-fashioned I still like my paper OS maps!)

[bikes4two]

And of course there's the battery management side of things - the Speedo batteries on my kit last at least a year - it's a different story for the batteries in GPS gizzmos.

Mind you, many people are ditching their speedos which means they are cheap to buy on ebay 2nd hand and I've got them on all my bikes and some spares too

[Witterings]

Add them both up and divide by 2 = 25.035 ..... go celebrate!!!!

Personally I really don't worry about minor discrepancies I just want an approximate of if I've done 10, 15 or 45 miles .... give or take and use just a standard bike computer as it's so much less hassle and only needs the battery changing once a year rather than something I have to keep remembering to charge.

[RickH]

Here's an interesting one I found very recently.

I use a Garmin & wanted an extra speed sensor. LBS got me one in for a reasonable price & happened to be the new(ish) v2 one which includes Bluetooth as well as the "traditional" Ant+.

It turns out that if you pair the Bluetooth side through Garmin Connect Mobile it will auto start/stop record milage data "headless" & then download it direct to GCM (apparently the sensor can store up to 300 hours of riding data). Ray (of DCRainmaker.com) is using one on a Dutch cargo bike to add in the mileage he has done on that, shopping (elsewhere on his site there is a photo of the bike loaded up with quite a few crates of beer) & transporting kids, when he doesn't really need/want GPS data or other metrics (his review is here is you are interested).

[Bmblbzzz]

Don't worry about it.

I've always found the GPS slightly underrecords compared to the speedo; or the speedo overrecords compared to the GPS. Reasons for possible GPS underrecording have been given, possible reasons for speedo overrecording (in addition to inaccurate wheel diameter) could be as simple as incomplete wheel revolutions being recorded as complete; it only records passes of the magnet, doesn't know whether a full revolution has been done, so if you wheel the bike out from parking, then back to manoeuvre, it could record a few full turns when in fact the magnet has been just passing and repassing the sensor. Same if "hovering" at junctions. How much that could add up to, I don't know; probably not much.

In any case, total accuracy is not really needed.

[Mick F]

When I was using a Gamin Edge 705, I also had a speed/cadence unit. Cadence speaks for itself of course.

The speed part of it had a wheel magnet, and providing you had the settings programmed for Auto, the GPS and the wheel magnet conspired to make the speed and distance bang on. The 705 took speed and distance off the magnet primarily, but it was belt and braces to check on it using GPS. Tyre pressure variations, or tyre circumference as it wore, all were catered for. You did nothing. You left it to do it.

Best of both worlds I reckon.

As the magnet was on the rear wheel, it was interesting seeing the variations in rear wheel circumference when climbing up a hill vs sailing down the other side. Climbing makes the tyre deform and decrease in circumference as it puts the power down. Believe me, there was a few millimetres difference.

Just a few millimetres difference as an error with a manual system, would put the speed and distance out quite a way over time. The 705 eliminated the errors.

[NATURAL ANKLING]

Hi,
Adjusting satnav in settings I believe can make it more accurate.
I normally ride a course several times, after a pessimistic input of my wheel size.
Check this with an online route planner.
Has anyone got the satnav to read longer than an online route planner?

[Mick F]

Has anyone got the satnav to read longer than an online route planner?

Yes.
It varies a bit, but it's never the same and often my Garmin Montana reads higher.

I must make a point of checking accurately and following the plotted route accurately.

Maps and mapping programs can only use flat-earth. Wheel magnet systems go up and down the hills, and GPS devices know the elevation profiles. The steeper the hills, the more maps and programs are out, though it's a moot point if it makes much difference ........... but the further you go, the further the error.

My Monty is set to record at 1sec intervals, and is accurate to 7ft.

[PH]

Use them both: Claim the highest and round it up.
Well don on your 26 mile ride