Head unit accuracy

Page may contain affiliate links. Please see terms for details.

Domus

Guru
Location
Sunny Radcliffe
I did a search and found a thread from 2012. Eight years on have things improved? The reason I ask is my Elemnt Bolt is my only point of reference on the bike, if it reads 50 Kms and 500 metres climbed is it true? When I get home the unit uploads to Wahoo app, Strava, RWGPS and Garmin Connect via Dropbox and they all are different, not by much but a few metres of elevation and maybe half to one Km. How is this possible when they all get the same data from the Bolt? :wacko:
 

Dogtrousers

Kilometre nibbler
The elevation will vary because all sites apply their own smoothing algorithms. The distance probably for similar reasons.

Just pick one and stick with it.

Do not think in terms of accuracy. Can there even be an "accurate" measurement of elevation gained? How would you measure It? Would it sum up all the tiny bounces when you ride over a piece of grit or a pebble? Would an "accurate" distance measure be based on the front or the back wheel? If you stare into the abyss of accuracy of measurement it will stare back.
 
Last edited:

T.M.H.N.E.T

Rainbows aren't just for world champions
Location
Northern Ireland
The exact same route over multiple repeats can even read different elevation. Garmins for example used internal barometric altimeters which are affected by weather and temperature and not compensated.

As above, pick one and go with that
 

ColinJ

Puzzle game procrastinator!
I find that GPS distance measurements are very accurate. I plot new routes carefully on digital OS maps so I know how far they actually are to a good degree of accuracy and 2 different GPS devices usually agree with the plotted distance to better than 0.5%, which is good enough for me.

Only one of the GPS devices (an old Garmin Edge 500) gives me an accumulated ascent value and that is NOT reliable. I have been out for a one hour ride before now and got back to be told that my house had moved up or down by 60+ metres! Obviously if the device cannot measure altitude accurately then it can't log accumulated ascent accurately either.
 

Dogtrousers

Kilometre nibbler
When doing rides for the century challenges I always try to ride at least 101% of the target distance to allow for "GPS shrinkage" where the on bike device says a distance of 100 (or 160.9) but once uploaded the website gives you a maddening 99.9 (or 160.8).

As for what the websites will award you for elevation gained, I find that's pretty much pot luck, but it should be in the same ballpark as the device. Different websites will be consistently high or low. (Same applies to moving average speed and other derived metrics like power or calories)
 
Last edited:
What is an accurate measurement?
What is an exact measurement?
Unless you are in a totally controlled environment all measurements are approximate.
The measuring scales that used to be outside Police Stations (a long time ago) were affected by heat.
The standard metre bar in Paris was so much trouble to keep accurate that they scrapped it.
Now the standard measure is wavelength of some gas or other and how that scientifically controlled measure is accurately passed to the outside world is a mystery to one and all.
GPS and other modern methods rely on signal reception and algorithms.

Take one measuring tool that you like and stick with it. Your measurements be be inaccurate but they will be consistently inaccurate.
 

Tenkaykev

Guru
Location
Poole
Gps? Notoriously unreliable. One of my regular Sunday morning runs was twelve miles over the Purbeck hills.
When GPS watches first became a thing it kept telling me the distance was actually ten and a half miles 😉
 
OP
OP
Domus

Domus

Guru
Location
Sunny Radcliffe
Ta, did my first two pointer for a while in the 50km challenge yesterday. 50 miles is 80.5 Kms so as the Wahoo clicked over to 81 on my driveway I assumed I was fine. (Claimed anyway) however when uploaded Garmin Connect gives 80.9 and Strava 79.8. On the other hand Garmin under valued my climbing but Strava agreed with my Wahoo. :wacko:
 

PaulSB

Legendary Member
Stick with one unit you like, my choice is a Wahoo Elemnt and Strava. I'm interested in comparison and overall numbers. If it's slightly out does it really matter.

I have a Garmin Vivoactive HR which I use when walking. The elevation measure is astonishingly accurate. I can stand next to a trig point and have the exact reading.

Once did a ride using the Vivoactive, Wahoo and phone running RWGPS to record the ride. The devices varied by 37 feet on ascent and 0.1mile on distance. That will do for me.
 
I use a 520 to record my rides. If I use the Strava phone app at the same time the only consistent measurement is time.
Distance, height, speed and even segment performances are all different between the two units.
My mate has a Wahoo, and his measurements are different from either of mine.
 

Drago

Legendary Member
Total accuracy is not achievable. However, near total consistency is - pick a brand and stick to it if a few metres over many miles is of a concern to you.
 

pawl

Legendary Member
I have a Wahoo Bolt and have noticed a stretch of road I enter at15 mph drops to 12 mph.
I assume this is due to a reduced satellite signal Does this affect the accuracy of average speed and mileage recorded.I assume it Doesn't bother me unduly .Just a thought
 

Drago

Legendary Member
That'll be a quirk in the programming. GPS satellites are essentially highly accurate clocks combined with radio beacons. They send nothing more than massively accurate time signals. The device receives those signals, compares timings between the two, quickly does a sheet load of maths, and gives you a result. The receiver either receives the signal or it doesn't - its digital in nature, so a weakened signal still means the same thing as a strong one.

People think the the more satellites the unit can 'see'the more accurate it becomes. That's sort of true, but not quite for the reasons people think. No matter how many birds the unit car see it only does the calculations between 2 at a time, then it'll move on to 2 more, etc.

And the final consideration, GPS was only ever intended to define a location on the earth's surface. Speed came along not as a consequence of any changes to GPS, but because receive units were developed that could do continuous multiple calculations, co,pare themnto the previous ones, do the maths to work out the difference and calculate a time. GPS can be used to calculate altitude but is not especially accurate because it lacks a fleet of birds at a second, much higher, altitude. Truly accurate altitude calculations require the unit it be perfectly stationary for long periods of time, such as bolted to concrete blocks like the Ordnance Survey do. Aircraft that require a super accurate altitude reading get theirs by cross referencing GPS with barometric and radar data to improve the resolution.

The Galieleo system was designed from scratch to have a constellation of birds at a much higher altitude, but unfortunately the system is compromised by financial and political meddling, and is proving unreliable. In a significant number of the unit launched the hyper accurate super cooled clocks have failed, and the units have had to revert to more conventional time keeping units so none of the promise of additional accuracy or functionality is materialising. The US air force were simply doing it better 45 years ago.
 
That'll be a quirk in the programming. GPS satellites are essentially highly accurate clocks combined with radio beacons. They send nothing more than massively accurate time signals. The device receives those signals, compares timings between the two, quickly does a sheet load of maths, and gives you a result.

So far so good.

The receiver either receives the signal or it doesn't - its digital in nature, so a weakened signal still means the same thing as a strong one.

Well, not really, noise processes will become more dominant for lower received signal strengths thus reducing the tracking accuracy.

People think the the more satellites the unit can 'see'the more accurate it becomes. That's sort of true, but not quite for the reasons people think.

I'm not sure what the reasons are that people think, but primarily it's the geometry of the satellites with respect to the receiver and the ability to calculate an over-determined solution.

No matter how many birds the unit car see it only does the calculations between 2 at a time, then it'll move on to 2 more, etc.

I'm not sure what you mean by "2 at a time" but it will depend on the receiver software - most multi-channel receivers will use all measurements in an over-determined solution.

And the final consideration, GPS was only ever intended to define a location on the earth's surface. Speed came along not as a consequence of any changes to GPS, but because receive units were developed that could do continuous multiple calculations, co,pare themnto the previous ones, do the maths to work out the difference and calculate a time.

That's simply wrong.

Yes, a receiver may determine speed by differencing locations but most will determine it directly from range-rate measurements.

Receivers track phase (range) and frequency (range-rate).

Similar solutions are used to produce x, y, z & t (position) and the derivatives (velocity).

GPS can be used to calculate altitude but is not especially accurate because it lacks a fleet of birds at a second, much higher, altitude.

What's the orbital height got to do with it?

The reason why GPS vertical accuracy is typically poorer then horizontal accuracy is because the geometry is one-sided, receivers cannot track satellites below the horizon.

The Galieleo system was designed from scratch to have a constellation of birds at a much higher altitude, but unfortunately the system is compromised by financial and political meddling, and is proving unreliable.

There not that much different: SMAs of 29,500km (Galileo) vs 26,560 km (GPS).

But you are quite right that the primary driver behind Galileo is political/financial.

HTH
 

Ajax Bay

Guru
Location
East Devon
Gps? Notoriously unreliable. One of my regular Sunday morning runs was twelve miles over the Purbeck hills.
When GPS watches first became a thing it kept telling me the distance was actually ten and a half miles
But how did you know it was "twelve miles"? Had you measured it with a steel tape or (less accurate, especially over the Purbecks) a measuring wheel?
 
Top Bottom