Speed accuracy of bike computers vs GPS tracking

petemadoc

I haven't got a fancy Garmin or anything, just a cateye wireless computer.

My mates got a suunto thing that uses gps to track speed and I sometimes use mytracks on my android phone to track the route/speed etc

The gps devices always seem to record about 1mph slower av than the cateye (and yes the wheel size is entered correctly). The only reason I can come up with is that the gps things are plotting a straight line between each recorded point hence calculating a shorter distance traveled. Have I got this right or is my cateye getting things wrong?

mattshrops

dont know the answer to your question.
but you can find out which one is accurate- choose say a 10 mile run ,do it together sit and work out the maths after - boom bada bing.

GiantMike

GPS should record approx once a second and applt a degree of 'smoothing' to the data. Unfortunately, GPS receivers have a 'circular probability of error (CPE) of between 3 and 10m. This means that, for 90% of the time, the recorded position is within 3 to 10m of the rider's actual position. A GPS receiver may also assume that the rider maintains a steady speed and therefore miss the finer details of accelerations and decelerations, weaving etc. However, this should show up as a different distance.

Assuming it is correctly set-up, the Cateye should be 100% accurate because it recieves a confirmed signal from the wheel every time it turns.

It might be that there is a difference between the way they calculate the average speed i.e. moving time vs total time. But basically, because of the different manufacturers and the differences in the way they gather the raw data, I'd expect the 2 to have different readings.

snailracer

GiantMike wrote:

...
Assuming it is correctly set-up, the Cateye should be 100% accurate because it recieves a confirmed signal from the wheel every time it turns...

Nothing is truly 100% accurate, for example tyre size, pressure and tyre deflection due to load all change the effective diameter of the wheel/tyre assembly.

GiantMike

snailracer wrote:

GiantMike wrote:

...
Assuming it is correctly set-up, the Cateye should be 100% accurate because it recieves a confirmed signal from the wheel every time it turns...

Nothing is truly 100% accurate, for example tyre size, pressure and tyre deflection due to load all change the effective diameter of the wheel/tyre assembly.

Fair point. Assuming it's correctly set up it will only be 99.99% accurate.

Unclemonty121

i have s similar problem with the iphone and mapmyride app, however normally it's only about .2 mph different.

One thing I know is that the bike computer (a Polar) stops when I get stuck at lights etc, where as I am not sure the gps allows for that, actually I am pretty sure it doesn't. I tend to adjust some of the stats on the phone to the one on the bike, and get a sort of mid way point which I am happy with.

nwallace

Some of the GPS plots when stationary are hillarious.
In the space of an hour standing speaking to a mate in the street with the bike propped up against the wall I travelled 20 miles.

Also had tracks at 1s intervals showing me weaving all over the road and even bouncing from the top of 1 hill to another.

Could never trust it.

Ands

nwallace wrote:

Some of the GPS plots when stationary are hillarious.
In the space of an hour standing speaking to a mate in the street with the bike propped up against the wall I travelled 20 miles.

My turbo sessions according to my GPS have me cycling all over my neighbours' gardens, as far as about 4 doors down the road.

fossyant

Standard bike computers are are accurate if you set the wheel size OK.

Garmin's are usually OK, unless you get a spike etc.....

petemadoc

GiantMike wrote:

snailracer wrote:

GiantMike wrote:

...
Assuming it is correctly set-up, the Cateye should be 100% accurate because it recieves a confirmed signal from the wheel every time it turns...

Nothing is truly 100% accurate, for example tyre size, pressure and tyre deflection due to load all change the effective diameter of the wheel/tyre assembly.

Fair point. Assuming it's correctly set up it will only be 99.99% accurate.

nwallace

For tonights ride, it seems I took an offroad route through the woods.
Ok admittedly the tree cover on the road there is fairly thick.

I'm using my phone, on the road bike I stick it in the saddlebag, so it's only trying to get a signal through the top of the bag, the saddle and my arse.

Seems to be just as accurate there as in a bar bag though.
Good readings on the West Sands road in St Andrews though, just clear blue sky above me.

bompington

nwallace wrote:

For tonights ride, it seems I took an offroad route through the woods.
Ok admittedly the tree cover on the road there is fairly thick.

I'm using my phone, on the road bike I stick it in the saddlebag, so it's only trying to get a signal through the top of the bag, the saddle and my ars*.

Seems to be just as accurate there as in a bar bag though.
Good readings on the West Sands road in St Andrews though, just clear blue sky above me.

Cycled to St A's on thursday, lovely run downwind from Dundee, arrived and got the phone out to check the route... it had never picked up a gps signal at all, total distance 0, average speed 0.
Reset it & it recorded the upwind return home in gruesome detail...

nwallace

Get that a lot with SportTracker, it just doesn't lock on to the signal in Dundee at all. Yet it's 90% of the time fine in Wormit.

What way do you go?

I'm going through Tayport down by Tentsmuir and then the obvious route through Guardbridge and the railway path. The back by taking the main road to Egyptmuir and then in through Links Wood.

Last Saturday shot off, averaged 30kmh to the Tentsmuir junction then hit the wind, only lost 4kmh in the last 40% though, whihc wasn't too bad!