View Full Version : GPS accuracy?

7th March 2009, 01:16 AM
Hi all..
I've got myself a Locosys GT-31 GPS for measuring speed on my boards.
Even when I don't move the GPS it shows speeds up to 3 knots at times.
This must affect the accuracy a lot (up to 10 %). Did I really sail 31 knots? Or maybe 28?

Aren't these GPS quite unprecise, or should I change something?

Any feed back is very welcome


7th March 2009, 01:40 AM
Hi Per
I work as at captain on board High Speed Crafts.
Among a lot of equipments, we got 3 different GPS, used for speed measurements.
I own several GPS to be used handhold.
When I compare the equipment on board the vessel with my personal GPS, the accuracy is exactly the same.

I trust in the readings (most of the time....;.)


7th March 2009, 03:33 AM
Hi Bensen.
I'm a sailor myself, and I trust my GPS too (most of the time - I still keep a good look out).
Anyway, for trimming my boards speed 2 knots make a lot of difference.
Let's say you have a certain trim (all other factors being excluded) giving you a speed of 30 knots. Your GPS though says only 28 knots (due to unaccuracy). You change trim, head for another run, and this time your speed is only 28 knots but your GPS claims 30. You think your trim made you faster, but the fact is you lost speed. Ithink a GPS should have an accuracy within 0.25 knots to be reliable.
Anyway I'm still working to understand my new toy..


7th March 2009, 03:44 AM
Have you ever tried this: http://trail.motionbased.com/trail/player/view.mb?episodePk.pkValue=7713213

Quite interesting.


7th March 2009, 05:19 AM
Hi Per,
Rest assured that the GPS is accurate.....very accurate.
Try these 2 tests:
1/ Turn on the GPS (on shore) and set it down on something solid.
Give it a minute to locate all the satellites it can.
Take a look at it after a couple of minutes.
Does it still show motion?
Probably not would be my guess.
The 3 knots you are seeing could be one of the following:
Current induced.
Drift due to the wind moving you and your board along.
Your hand motion (if it's attached to your arm).
2/ Place the GPS in a visible place on the dashboard of your car.
Give it a minute to find the satellites.
Drive along slowly, then increase your speed a bit.
At < 10 mph, the GPS should be very accurate to 1/10 mph-m/s- knots.
Since we are no longer "burdened" with selective availability (SA for short), the positonal
accuracy of a GPS unit is down under 1 meter if the unit is allowed clear access to the full range of satellites in it's viewing area.
Even in the days of SA, (a slow sine wave error signal that varied and caused the positonal accuracy to decrease to around 30 meters (factor of 10)) the CHANGE in position accuracy was still very accurate as the SA signal was very slow.
The intent was to cause deliberate positonal inaccuracy.
Now, with no SA, and better GPS units that can see more satellites, it's become very accurate.
Oh, if you check the accuracy when moving with the GPS in your car, and the speedometer and GPS don't agree, the inaccuracy is most likely "built in" to the car speedo system by the factory.
My Mercedes Sprinter Van speedo (analog) has a built in 3 mph (reads 3 mph high) to keep me a little scared of getting a ticket and give me a 3 mph "cushion" to prevent tickets.
I've had as many as 3 portable GPS units running in my van at the same time, and all of them indicate within 0.1 mph while the speedo reads 3 mph higher.
So, if I'm cruising at 60 mph on the speedo (steady speed on the cruise control with RPM steady with in 10 rpm) the GPS units will all read 57 mph +-0.2 mph.
I used the recorded and printed GPS readings recently to have a traffic ticket for 70 mph in a 55 mph zone reduced to 62 mph for a greatly reduced fine and no points on my license.
Hope this helps,

7th March 2009, 01:42 PM
Hi Roger.
You're always good for a detailed answer. Thanks a lot.
I made the first test. I placed my GPS outside under a clear sky on top of my litter bin. It didn't move at all. Anyway it measured a max speed of 1.98 knots and an average speed over 10 seconds of up to 0.98 knots - while NOT moving at all!!!
I guess something is completely wrong with it. I have a friend who made a run with his old Garmin and this new GT-31 and they disagreed with as much as 6 knots!!!
My old Garmin E-trex never showed any sign of being more than 0.2 knots unaccurate..
Does anybody know the GT-31? Is there anything in the set up that I may have forgotten?


7th March 2009, 04:13 PM

7th March 2009, 05:54 PM
Hi Romas.
Thanks A LOT ;-).
Anyway I tried to bring my GT-31 for a run in my car. As soon as the speed increased to above 10 km/h it appeared to be very precise indeed. Actually I was not able to keep the car on an exact speed precisely enough. It looked like the GPS vas precise within 0.2km/h.. Strange.. I will check your links even more and take care that the GPS is updated..
My first goal is 32.5 knots on my iSonic 122 this week-end..


12th March 2009, 02:16 AM
Sounds odd. You shuld ask this question on http://www.gps-speedsurfing.com/

Try testing it in a position a meter or so from where your bin was. There is a grid effect (which I find hard to explain in words) that may explain what happened on that occasion.

In my tests so far the gt31 is very accurate.

Ian Fox
12th March 2009, 02:37 AM
Hi Phill,

"Grid effect" can affect accuracy of speeds derived using Lat / Long methods,
however it is eliminated by the use of Doppler based, which is what (and why) GPSSS
now choose to base all accurate and record timing on Doppler results.

The speed accuracy of the GT-31 (correctly configured and using Doppler) is very precise.
-and significantly free of errors seen from older or lat/long based units/results.

Cheers ~ Ian

12th March 2009, 05:20 AM
I really don't know what I was thinking there. Of course the 31 is doppler. This proves that I am totally barking.

Good to see you are getting the is44 up to some great speeds.

12th March 2009, 10:58 AM
Welcome back Mr. Fox

Craig Bergh
28th March 2009, 10:41 PM

You have asked a very good question.

The GT11 and GT31 are one of very few GPS units that display AND record Doppler Speeds.

This GPS records satellite location and strength for up to 12 satellites every second. This information can be studied on the computer for every single trackpoint on a run across the water.

When the GT31 displays speeds on screen even when the GPS is not moving, that is due to static from satellites.

The GT11 had filters in the firmware that filtered out this static.

The Windsurfing Tech Group removed the filters on the GT31.

I agree that it seems odd for a GPS to display speed when at rest.

Once the GT31 gets moving, that static does not appear. And the results are very smooth once the file is brought into the computer.

The GT11 and GT31 are extremely accurate tools to measure speed, and the GPS files are quite smooth and precise.

Please let me know if I may provide more information, or samples of files for others to examine.

Thank you everyone. I am lookng forward to a great year of windsurfing.

Craig Bergh
Worthington, Minnesota

email: cbergh@iw.net

web: http://www.midwestspeedquest.com

1st April 2009, 02:18 AM
Thanks a lot Craig. Your information is very useful.
Yes it seems odd that a very precise unit shows 2.83 knots when not moving, but my observations are like yours: as soon as the GT31 comes into play at a steady speed it becomes quite precise and updates very quick indeed.

I guess my GT31 doesn't have any problems, so if I dont clock 33 knots on my iSonic this week-end the problem is with... me...... (-;