Garmin inaccuracy - Gear - Runner's World

crispylees

07/01/2006 at 20:34

I have been using a Forerunner 301 for a few months now and it seems to be consistently 10% short when measuring distance. That is to say that it tells me I've finished a 10K as I pass the 9K marker!

Has anyone else had a similar experience?
It was bought online from a Danish retailer, is there a UK arm of Garmin who could help?

No.8 ™

07/01/2006 at 21:12

There's always going to be a discrepency with GPS, it is never going to be 100% accurate. That coupled with the fact that you running a 10k, will deviate from the exact (measured) racing line and possibly throw in a bit of cloud cover, trees or tall buildings, then all this will lead to you not recording "exact" distance.

My Garmin is always out, but it is enough of a guide for me for pace and distance.

RunningCommentary

07/01/2006 at 23:37

10% is way off though. I did an analysis of all the races I'd run with a Forerunner, and the discrepancy was something like 1.5% which is very acceptable when you consider the way you meander round a course.

If you're 10% out, you must have a faulty unit, surely? What would be interesting would be to find some other people with Forerunners at the finish line (there are usually a few about) and ask them what readings they had.

crispylees

08/01/2006 at 01:12

Good point, and it's not as though Forerunners are hard to spot!!! I will try comparing with another Forerunner - I KNOW that it is out compared to a measured distance (race, car milometer, map) but I don't yet know whether it is the GPS system or MY GPS device. Thanks

limping along

08/01/2006 at 10:21

Intersting...if you loose GPS coverage (and cloud cover will not effect this) the total distance should be less than actual distance as the unit calculates the distance as a straight line between the two positions (last fix before loosing coverage and next fix after regaining it)

Something to consider is where you wear your unit as I have noticed that this can cause positional accuracy to decrease and throw in some wayward positions - this can increase total distance. If I loose signal something as simple as swapping the unit to my other wrist can regain signal. What sort of terrain are you running in? Tall buildings can certainly effect coverage and accuracy (signal bounce?) as can trees (signal shadow?).

If you use software like (the exellent) sporttracks that can help identify where the signal problems exist and may help better understand where the error is occuring.

Also the position of the satellites the unit is tracking can influence accuracy - a satellite low down to your horizon will provide less accurate data than one that is at say 45degrees (mid way between your horizon and overhead).

On a side note, does anybody know a website that shows GPS satellite positions / coverage for a given location at a given time?

crispylees

08/01/2006 at 12:45

I understand the principle behind calculating distance with missing data and accept that this can only ever reduce the reported distance but is it not odd that the problems I have are consistent in their inaccuracy?

Also, I have found the problem everywhere: at home in Derbyshire where the tallest building is a two storey cottage 50 yards from my path and at most times there is nothing bar the clouds between my and the sky; and races such as along the dock road in Liverpool - again no buildings at all for much of the route - and a fairly rural route in Cheshire. I have also, however, used it in the very built-up city centre of Sheffield with the same result.

What stumps me is that if I look at my routes on a map it is absolutely spot on, so the device KNOWS with relative accuracy WHERE I am, surely it can then calculate the distance accurately. Does this mean then that the problem is nothing to do with GPS and positioning but with the software in the device miscalculating the distance between two accurate points???

I have been using SportTracks, I would be grateful if good,bad,injured could elaborate a little on your idea of looking for signal problems and I will report back.

Little M.iss Happy

08/01/2006 at 14:43

Now I'm technologically challenged to say the least and I love my Garmin (even if all I can do is switch it on and off) and I understand the principle of it drawing a straight line if it loses signal so why does it sometimes measure short? I'm sure it has, well, it's either that or I've grown wings for the odd run:-)

les,mudlark

08/01/2006 at 18:19

If you dont get any help to sort your Garmin out,you can E mail garmin support
& they do get back to you!.I had problems
with my 301 & they exchanged my 2mth old
Garmin for another new one!.

crispylees

08/01/2006 at 22:28

Thanks James. I have now done that. It's good to know that you have had good experiences with Garmin support.

Little Miss Happy: sorry if I've missed the point but if what you are asking is why does it measure short when it draws a straight line then the answer is that straight lines are always the shortest distance between two points. Suppose that you are running in a circle and all is well for the first half of the circle, then you will have an accurate distance for the first half (ie report = actual). Suppose that the signal is lost for the entire second half and only picked up again at the end then the Garmin will assume you did a straight line right across the middle of the circle. You actually ran the same as the first half but it will report the diameter as its measured distance, D'Oh: it says you ran 3 miles in 30 minutes and you actually ran 4 miles in 30 minutes, distance achieved wrong, pace wrong. Just what you wanted!!!

northern snail

09/01/2006 at 00:06

I understand that one of the problems with the cheaper commercially available GPS units such as the Garmin and others is that the GPS satelites are for military use and to get the same accuracy as they require would put the price up so far as to make them too expensive for recreational use.

This was told to me by a friend in the military who understands these things a lot clearer than I do.

Hope this helps

hammerite

09/01/2006 at 11:42

http://sporttracks.zonefivesoftware.com/Discussion/viewtopic.php?t=345

this may help to explain some discrepancies. But as already said 10% seems quite excessive.

crispylees

09/01/2006 at 12:07

That is an interesting article. I had no idea that the splits were calculated based on distance and not time, a point worth bearing in mind when reviewing sessions.

Unfortunately my problem manifests itself everyhwhere in the same way. The data on the Garmin is out, the data viewed in SportTracks is out (by the same amount), and the data viewed in Training Center is out (same).

Amazingly last night I think the degree of inaccuracy was less. I plan to drive the route (new route) tonight to see what the car thinks various parts of it measure, but I am pretty certain that it was out by less than usual.

Maybe the Garmin is learning that things in our part of the world are never quite what they seem!!!

hammerite

09/01/2006 at 12:52

well crispy Derbyshire is my part of the world too and I have no problems!!

Beanz52

09/01/2006 at 12:59

I'm in Derbyshire too!

today's run produced a discrepancy between the SportTracks distance - 8.9 miles and Garmin 9.004 miles

which is interesting - it means they must be interpolating differently as they are using the same data

I assume where the route turns ot a dotty line on the Sport Tracks map is where the signal was weak - in Derwent Valley - steep sided valley with heavy tree cover

crispylees

09/01/2006 at 13:02

beanz52 - your issue is almost certainly covered in the article to which hammerite linked - detailed and takes some reading but well worth it, especially the bit that admin says not to read unless you really want to.

2muchcoffee

09/01/2006 at 13:06

how do they cope with hills and diffreences in allitude. i.e. think of a triangle (hill) it's loger up one side an down the other than it is from the two points (base to base). I've noticed an allitude reading on one of the garmin pages, but it just seems to jump about all over. what experiances have you all had with accuarcy when doing hilly routes??

crispylees

09/01/2006 at 13:17

My altitude is certainly not spot on but I think it is close enough for the software to work out the difference between sea level and local without anything too noticeable happening. Ie at sea level the distance between two points is less than at 2000feet (the earth is a bigger sphere at 2000 ft above).

As for its ability to deal with the triangular nature of hills - well I couldn't comment becuase mine is 10% out on the flat. I don't know whether the Garmin attempts to factor in that info or whether it just reports the distance as though a flat course was run.

As a Derbyshire runner it might be worth a visit to Jacob's Ladder at Edale. With a climb like that and a map to measure the flat distance you'd soon find out!

Inky Fingers

09/01/2006 at 13:32

I'm not sure that the Forerunner does something as simple as draw a straight line from the point at which it loses the signal to point at which it regained it.

When I download my routes to a Tracklogs map, it seems to me that if my Forerunner loses the signal (usually when I've made a sharp turn under foliage or near tall buildings) it assumes that I've continued in a straight line, and then makes stabs at getting me back on track until it's got a good fix again. Sometimes this route can be longer than the one I've actually run.

To answer GBI's question, I doubt if you could get a map of satellite coverage, because they're not geo-stationary. As GPS is an American system, the orbits are optimised for America and are often low in the sky in Europe. When the European equaivalent system, Galileo, is in business in a few years time, its orbits will be better for us.

Inky Fingers

09/01/2006 at 13:36

And don’t worry about allowing for gradients, unless you’re actually running up mountains. If you were to run up a half-mile hill rising 100 feet, you’d run less than two feet further than if you’d run the same distance on the flat.

Roxy Bird

09/01/2006 at 16:47

I got my garmin for xmas and have not been impressed at all...always losing the signel and so far it hasn't measured one run correct. Did a run on Friday and it said I finished running about 2 miles before I actually did!

Am going to use it tonight and try wearing it on the top of my arm to see if it makes a difference.