GPS discrepancies

[pheran]

When underway I normally turn on my hand-held GPS (Magellan) to 'second guess'
the boats main fixed GPS/Plotter (Simrad). I cross-check regularly and they both display the same heading and speed, as near as makes no difference. But when I check the log function at the end of the day, there is a difference of some miles in the distance travelled. Can anyone suggest why?

 

[Its_Only_Money]

Is one working in statute rather than nautical miles /forums/images/graemlins/confused.gif if so I don't understand why the speeds displayed wouldn't differ too though.....

Only other possibility is that one is plotting a more accurate track with many more points than the other leading to apparant differences in the distance travelled /forums/images/graemlins/confused.gif

 

[Birdseye]

If they are recording the same speed, then I can think of only 2 alternatives - the one of them is only switched on for part of the time, or one of them has a software error or hardware prob that is zeroing the counter at some point.

 

[BrendanS]

Do you have the 'breadcrumb' trail turned on? This can turn up some anomalies. I've had the trace jump a few miles off, then back again, while stationary. Just a momentary glitch, but puts my max speed at several hundred miles per hour and distance log shot.

The other possibility is that the alogorithms (calculations) used to calculate distance are very disimilar, and that errors are being caused in one or both units. Maybe Tome can throw some light on this - he's the forum expert on all things GPS.

 

[actionoptics]

The GPS is not affected by tide but the log will read distance through the water rather than over the ground.

 

[gjeffery]

Only guessing!

Were you steering a crooked course, or travelling in a straight line? I assume that total distance travelled is determined by integrating the instantaneous VOG readings. If you were travelling a crooked path, the machine with the faster update would record a longer, and more accurate, path.

Which begs the question, does anyone know what the update rates are for GPS receivers?

 

[gjeffery]

Following on, you might expect the handheld to be more power conscious, and to update less frequently. "Battery saver mode", would be more likely to reduce the distance measured on a curve.

 

[gjgm]

good question. If only Gludy were still about we could be heading for a 500 thread post !!

 

[pheran]

Many thanks for the suggestions so far. To answer the queries raised....

- the hand-held is plumbed into the ships electrics and shows 'external power'
so battery saving mode shouldn't be an issue

- I have checked that both GPSs are working in the same measurement units

- 'birdcrumb' track is working in the backgound but is not usually displayed

- travel is mainly in a straight-line in open water. Some changes of course but not
frequent (ie not as if going round bend after bend as on, say, the Thames)

To give you an idea of the extent of the discrepancy, on our last day out, one recorded 35 miles, the other 29.4. I have still to do the obvious thing ie get the chart out and work out which one is 'wrong'

Any other thoughts gratefully received.

 

[Guest]

I have two different Garmin GPS's, one hand held and one fixed. I have often noticed that, if I plug in the same waypoints into both of them, they sometimes give slightly different distances between waypoints. I noticed because I always cross-check one against the other (bearing and distance to each waypoint) because it helps to highlight any error I may have made typing a waypoint in.

Whenever there are any discrepancies, I always go back and triple check each waypoint. There is frequently, but not always, a discrepancy that isn't accountable to me - I have put it down to the something different in the formulas that each GPS uses to calculate distance. I had assumed that it may be a variation in the way that the effect of the curvature of the earth is worked out on each, although in your case a large discrepancy over a short distance does not seem consistent with this theory.

 

[gjeffery]

If both GPS are in WGS84, errors due to transformation computation should be eliminated. If both are set to any other co-ordinate system, transformation computation could be a reason for differences.

Getting the optimum update period is often a problem. Too slow, and any curve is reduced to a sequence of obvious chords, too fast, and timing errors become significant. The effect would be more obvious at higher boat speeds.

My point regarding the handheld is that the designer hopefully sought maximum battery economy, and set the update rates accordingly. This would not be as significant a constraint with a fixed installation. Battery Saver Mode, if set, would reduce the update period to say 5 seconds and the effect of a slower update would be more obvious.

 

[Guest]

Update rates cannot be relevant in my case, because the sets predict a different distance from WP to WP at the planning stage, before setting off.

Both are set to the WGS84 datum. So they both put the waypoints in the same place relevant to the chart, but I do not see that using the same datum should necessarily eliminate the possibility of different ways of calculating distance from A to B. Clearly my two GPS's DO calculate it differently. The differences are generally smallish (of the order of 0.2 or 0.3Nm over a 50NM course, and some of it may be attributable for example to one set rounding figures up in the computation, the other rounding them down. Probably not the explanation, but just one example of how they might compute differently.

 

[Its_Only_Money]

The figures you quote are the discrepancy you would expect if one was set to Nautical miles and the other to Roman miles /forums/images/graemlins/shocked.gif

 

[Guest]

No, both are set to Nm. Also, the discrepancy isn't constant, appears with some calculations, not with others

 

[gjeffery]

I was suggesting that the update rate might account for the discrepancy in the log function (distance run)

With regard to the difference in the distance to waypoints calculated on different machines, I suggested that the most probable cause of differences between GPS systems would be any transformation between co-ordinate systems. Grid transformations are horrible, depend on assumptions, and hence are usually compromised. If you are set to WGS84, this source of error should be eliminated. The method for calculating distances between geographical co-ords should be repeatable - but it would be interesting to know the method that Garmin et al actually use.

 

[Guest]

I take it you total traveled as logged by each GPS ..., and not by the Log !? If by the log - then they will NEVER be same.

If we are talking logged totals on each GPS - then you may have averaging set differently on each machine. Some sets call it damping, averaging, sampling, filtering etc. etc. This is the calculations and display .... the results jump around significantly and you can see that if you remove filtering ... speed jumping to impossible figs., then back to small numbers etc. Imagine the error that creates this .... we are taking readings and calcs over a small time period and the accuracy of the GPS position is anywhere 3 - 50mtrs etc. Increasing the averaging / filtering reduces the jumping around and can also alter the totals as the speed variations are reduced accordingly.

Thats my theory anyway ....

 

[Birdseye]

[ QUOTE ]
Many thanks for the suggestions so far. To answer the queries raised....

- the hand-held is plumbed into the ships electrics and shows 'external power'
so battery saving mode shouldn't be an issue

- I have checked that both GPSs are working in the same measurement units

- 'birdcrumb' track is working in the backgound but is not usually displayed

- travel is mainly in a straight-line in open water. Some changes of course but not
frequent (ie not as if going round bend after bend as on, say, the Thames)

To give you an idea of the extent of the discrepancy, on our last day out, one recorded 35 miles, the other 29.4. I have still to do the obvious thing ie get the chart out and work out which one is 'wrong'

Any other thoughts gratefully received.

[/ QUOTE ]

The difference you quote is very close to the difference between statute miles and nautical.

 

[VicMallows]

As a matter of interest, I put the GPS (Garmin 128) in the car on a recent long trip. Primary purpose was to calibrate the speedo (its spot on on 30mph, overreading by 1mph for each additional 10mph actual). The odometer though agreed with the GPS distance within 1 mile (over 300) .... the limit of the GPS resolution.

Vic

 

[Guest]

But I guess your Odometer won't be nearly that accurate. Presumably it depends on how accurately your Odometer's been calibrated to the circumference of your tyres, which will depend on how much each of them has been pumped up, how warm the air in them is (time of day), how much slippage there is in traction (road surface wet?) etc. etc.

 

[pheran]

We are talking about the log (odometer/trip) function on each of the GPS sets. Thanks for the averaging suggestion. Will investigate.