How accurate GPS

[DERF]

Was in Solent this weekend and for the first time noticed a considerable error on GPS readings.

Chart and plotter set up is correct with right geodetic etc etc...

On exiting Portsmouth ... Balast bouy appreard some 100mtrs from where its supposed to be... As it has recently been removed and replaced I didn't give it much thought..

But the same then applied to all the navigation marks.... All about 100 - 150mtrs away from the indicated course and track on the plotter. Even though I had plotted to be well off the marks I was some 100 mtrs closer than I should have been. (not nice on the return sunday night)

I know the limitations of GPS but I've never had such consistent errors of this magnitude. Wasn't tide either as it was flat calm and happenned in all approaches.

Been using Raymarine for yonks but this is new Furuno kit... what have I missed? or was everyones GPS out? Maybe I'm being a Klutz.... so any suggestions?



<hr width=100% size=1><font color=red>If it doesn't kill ... it fattens</font color=red>

 

[Wiggo]

I think the Solent charts have been corrected to WGS84, but the buoys themselves are still on OSGB...

<hr width=100% size=1>Si hoc legere scis, nimium eruditionis habes.

 

[ccscott49]

I doubt it was the GPS, something in the setup maybe? Was there any Navtex or coastgaurd warnings of GPS anomalies? You could call and ask.

<hr width=100% size=1>

 

[KevB]

I have found 'dead' area's in the Solent with regards to GPS readings. On a couple of occasions I have lost sat reception whilst approaching Portsmouth harbour. Solent triangle maybe?

<hr width=100% size=1><A target="_blank" HREF=http://static.photobox.co.uk/public/images/45/99/10714599.s.jpg?ch=97&rr=16:00:39>Nirvana</A>

 

[Wiggo]

more likely strange ECM stuff from those big grey boats on the right hand side of Portsmuff...

<hr width=100% size=1>Si hoc legere scis, nimium eruditionis habes.

 

[LeytonC]

Was *everything* 100 meters off? cos my plotter has a function that allows u to 'nudge' the chart a bit so it lines up better. I presume your plotter would be the same, maybe its set wrong from the factory or summat.

<hr width=100% size=1>Thanks

Leyton

EXTAR Software - Software that works for you
Phone : 07803 139 736

 

[tcm]

i agree with Leyton. Walk exactly 100metres to the left, and the gps will be correct. Alternatively, move the buouys to agree with the GPS, but this is a lot more work.

<hr width=100% size=1>

 

[DERF]

Don't think its the OSGB thing.. as the charts are already corrected for this aren't they?

I wondered if the system I use needs to have a position offset applied?

Don't recall having this problem with the Raymarine kit and I've been passing those Navy boys for years with no funny stuff on the elctronics...

What about radar /chart overlay interference? Could that be causing an error or offset?

<hr width=100% size=1><font color=red>If it doesn't kill ... it fattens</font color=red>

 

[Deleted User YDKXO]

Could be difference between C Map cartography on Raymarine plotter and Navionics stuff on Furuno plotter. Had Navionics charts with Furuno plotter on my last boat and was'nt too impressed with cartography - poor accuracy and lack of detail. Do you have Navionics Gold charts 'coz they're supposed to be better?
You could try adjusting your position on the chart. On my Furuno, it did have the facility to move the chart to the correct position relative to the boat but then, if you set up for one harbour, it could be well out at the next

<hr width=100% size=1>

 

[mjf]

You do not say if its a hand held or fully fitted unit with plotter.

Last weekend I was showing someone my plotter etc and used the radar overlay on the plotter chart - in Port Solent the outline of the marina / pontoons etc exactly matched the displayed chart with radar picture. So 100%.

That said ; if I leave the track option on - the boat does move about more than actual. moves up the pontoon a couple of berths in the worst case say 20m?

<hr width=100% size=1>

 

[Wiggo]

20m is probably the error margin on GPS these days, unless you have dGPS, so I wouldn't be surprised to find I'm suddenly on the next pontoon according to the machinery.

<hr width=100% size=1>Si hoc legere scis, nimium eruditionis habes.

 

[DERF]

I'm using C Map NT charts not Navionics and I am happy with the detail.

I agree with you that its easy to dial in an offset on the chart for a particular harbour/area, but it may only apply in that location and knowing me I'll forget to dial it out again and be miles out somewhere else!!!

I do wonder however how accurate GPS should be.. I'm sure I've read that it is now accurate to within 10mtrs..... I seem to recall that sort of error with my old Raymarine kit...

Didn't seem to have problems with this Kit in the Med this year.. I guess I've forgotten to do something obvious....will have to trawl trhough the huge manual and see...

<font color=blue>So back to my question<font color=blue>.. <font color=red>Anyone know what the accepted accuracy of current GPS is? indeed may it vary dependant on lat? <font color=black>

Someone out there must have some data?




<hr width=100% size=1><font color=red>If it doesn't kill ... it fattens</font color=red>

 

[DERF]

Fully fitted with plotter

Come to think of it mjf .....in actual fact in Port Solent it shows me on f4 not f1...

Hmmmmm..... something is amiss!

Are you using Furuno? if so you don't fancy jumping on board and giving it the once over do you? (I'm not down this weekend)

<hr width=100% size=1><font color=red>If it doesn't kill ... it fattens</font color=red>

 

[mjf]

Just seen its you (derf).

You might find that your unit is badly set up - esp ht above the water. The lat/long will not affect the accuracy here but the antena ht will throw it out by quite a bit I believe.

I have spent literally hours and hours with the Furuno book and I too use the CMap NT which I am very happy with. If you want I will take a look for you..



<hr width=100% size=1>

 

[BrendanS]

Most manufacturers claim under 15m, but in fact most of the time it's down under 8m, or even less. Lots of test sites on the internet, some showing that 12 channel receivers can achieve consistent results under 5m if they have clear access to the satellites

With Waas or Egnos, expect well under 3m (0.8m to 3m is claimed)

This is horizontal (2d), height (3d) is less accurate

<hr width=100% size=1>Me transmitte sursum, caledoni

 

[britemp]

P'raps the yanks turned SA back on for a while - they tend to do this if they think someones going to guide a missile at a US target using their own gps system. They also bugger up gps immediately around their ships for obvious reasons. The dGPS system I used to sell had 72 ground stations to track the sats so even them buggering about with SA didn't upset it, but I gather your common or garden chartplotter just uses free gps then?

<hr width=100% size=1><< Sig removed at request from Kim Hollamby - he thought it might scare forumites! >>

 

[DERF]

Thats what I thort(sic)

EGNOS .....so my 100-150m mtrs is definitely summink queer!

<hr width=100% size=1><font color=red>If it doesn't kill ... it fattens</font color=red>

 

[BrendanS]

Something def not right if it's that far out!

Having said that, check if it's OK next time out. There was a problem earlier this year, where one satellite went a bit bonkers, and was giving incorrect info. This caused an increasing error over several hours for anyone inline of sight, before the satellite was taken offline. Quite a big error was caused I understand.

I'll have a look and see if there were any problems reported over the weekend.

<hr width=100% size=1>Me transmitte sursum, caledoni <P ID="edit"><FONT SIZE=-1>Edited by BrendanS on 07/12/2004 17:53 (server time).</FONT></P>

 

[Talbot]

Would be surprised if the problem was not a datum selection problem. If the plotter was running in one datum and the chart was another, then you would have a consistent error.

Same problem for people using a GPS and any form of paper chart. - make sure the datums match, and understand the differences between them.

<hr width=100% size=1>

 

[brianrunyard]

Check it again next time out, if the error still exists then investigate.
Could be an error caused by a satelite clock going haywire as happened earlier this year.

<hr width=100% size=1>Brian
http://homepage.ntlworld.com/brunyard