before I send it back, is it something I am doing wrong....

[ShipsWoofy]

This was us on Friday sat tied to the fuel pontoon at 1/8mile range.

Look how the boat is jumping about.... It plays havoc with the anchor alarm, never noticed it when under way.

Has anyone ever seen this before or is it a faulty unit? I do not want to make a fool of myself sending it back.

 

[BrendanS]

Yes I've seen it on mine, just once in a while, on several different gps and plotters. It plays havoc with your log for the day - max speed 999knts etc!

Can only assume minor glitch in satellite reception causing spurious errors. Never lasts long, and usually only seen when at rest, not while moving.

 

[ShipsWoofy]

I never saw it on my last (geonav) unit. So, I just have to live with it, I can alter the anchor alarm range but (from memory) it is 0.01nm to 0.05nm step, 0.05nm is quite a bit of drift in a popular anchorage. At 0.01nm it will sound the alarm when it starts doing this.

I might contact RM to see if a firmware update is available, some damping might help when using the anchor alarm.... or an in-between range setting.

Cheers.

 

[BrendanS]

I'll change one part of what I said, having thought a bit more. Only ever seen it whle stationary, never while moving, even slowly, and thinking about it a bit more, only think I've seen it while at anchor, not while on a berth or pontoon.

Al my gps and plotters have been Garmin, so it's not a RM thing either

 

[mikejames]

Its a GPS thing. Keep the plotter and just be thankful they turned off the Selective Availability which would have had you cruising around at several knots permanently while standing still.

As you sit there the satellites move above you in the sky at several kilometres a second and go in and out of view. As they do that their signals bounce off or are obscured by nearby objects. Reflections make them seem further away, so your position will tend to drift away from the reflector. Also the thickness of atmosphere between you and the satellite varies depending on when it is overhead and on the horizon. This alters the speed of the signals and hence makes the ranges vary. The maths tries to use as many satellites as possible but the solution of position is trying to find the centre of a 'cocked hat' in 3-D .

Basically that plot is normal, the GPS is doing well but I suspect there is a building on the quay blocking some of the signals.

I have seen position errors of up to 1km from GPS where a building blocks the view.

I used to write the software inside the GPS units themselves.

In the end, different manufacturers place filters in their software to try to eliminate the effects of this sort of drift. None are perfect.

The antenna installation also has some effect. Placing it where there are no reflections or bits of boat in the way helps. But not putting it at the top of a mast where the rolling and pitching of the boat cause it to be forced to measure high speeds.

The military use two frequencies so they can measure the 'thickness' of the atmosphere and remove that effect.

 

[BrendanS]

If you have programmed these things, can you explain in non idiots terms. Techie stuff would be nice. Doggie is ex military radar, I'm just a geek. Though I thought that the GPS sorted many of the spurious signals by the very nature of the error correcting it has built in.

 

[Ships_Cat]

Only ever seen it while stationary

It may be that the plotter itself (not referring to the GPS part if it has GPS integrated) does some averaging so that smallish course deviations don't show up in the track drawn and that may account for not seeing the effect while underway.

Cannot speak specifically on the point for plotters but our ECS certainly has averaging (the extent of which is user adjustable) so that when it draws the track it is not all wobbly when I do the steering bit /forums/images/graemlins/smile.gif. I don't know if it does it while stationary tho'. But I have never seen any noticable escapades of position on the ECS when in the marina or at anchor so maybe it does - in fact in the marina we seem to sit with no movement obvious in the plotted track (err, rather plotted "spot" in this case) at all. Same in the marina have never noticed any 0.01 second digit or bigger changes in the fix on the GPS's own display (10 year old Autohelm ST50 GPS) which is always running.

John

 

[Guest]

Seen it on my etrex and on the Lowrance ... seen it on Commercial ships stuff.

It is most obvious when stationary. I would suggest damping the position averaging a bit more - you may have it a bit low allowing it jump around.

If you google about GPS position averaging etc. - there are plenty of sites where people have plotted position against time - even PBO did a plot vs different plotters / GPS a few years back ... even with Differential and todays WAAS etc. - it still happens, hopefully over a smaller area !!

 

[john_morris_uk]

You have confused me slightly as the last time (albeit a few years ago now) when I had to read up on the theory of GPS position fixing I thought that the algorithms tried to allow for the atmospheric refraction.

Notwithstanding all this, the physics and maths of how the things work is very very cunning. Furthermore the way the signal is 'found' by the receiver down in the noise is also clever.

Having been brought up without GPS and then with GPS with SA, I am always thankful for any fairly reliable fix within metres of where I am!

 

[tome]

Looks like a poorly sited or faulty antenna. Another possibility is that you have WAAS (EGNOS) enabled?

Atmospheric refraction is taken care of by modelling within the receiver, although intense sunspot activities can play havoc with the model in some equitorial regions. In our receivers we use dual frequencies to eliminate this effect

It isn't normal and you should investigate the cause. Take a look at the satellite constellation and see if your signal levels are ok and that you are tracking 5 or more sats

 

[Birdseye]

[ QUOTE ]
Having been brought up without GPS and then with GPS with SA, I am always thankful for any fairly reliable fix within metres of where I am!

[/ QUOTE ]

Need to be careful here. The GPS might be able to replicate a position fix to within a few meters but that is the order of accuracy of your charts even after the recent corrections to the current datum. Add in that much of the surveying on marine charts dates from the last century or even earlier and the accuracy could be even less. Mind you, none of this would cause the boat to jump about on the screen.

Any systems has inaccuracies / tolerances and I guess the jumping about reflects those plus any issues from bounced signals, poor reception etc. It should be easy enough to smooth out at design stage

On a practical level, try moving your GPS aerial location on the boat. Try switching off other electrical kit to see if you have a problem with noise. I wouldnt rule out a faulty set at all - in fact I'd assume a faulty set unless I found some other cause since it seems unlikely that the set would be designed to have an irritating issue like this. In any event, what do you lose except postage by sending it back for checking?

 

[philip_stevens]

Is WAAS (EGNOS) up and running officially & stable yet?

 

[Oldhand]

The position variation looks high in comparison to what I experience, 6 to 10m variation is typically the most I see in my marina berth. Is it consistently like that or just sometimes?

You make no mention of your installation. The GPS antenna must be outside and anything that will obstruct its view of the sky must not be close to it. I would suspect from your plot that your antenna is not fully seeing the available satellite constellation and the available satellites are not providing the geometry for accurate fixes. Does your equipment have a satellite monitoring page? If so, check how many satellites are being used, if the signal strength is good and how many satellites which should be in view are not being used. If the latter is consistantly more than 1 or 2 then I would suspect your antenna's view is being obstructed. A minimum of 4 satellites in good angle (30 degrees +) in relation to each other and having elevation between 15 and 60 degrees above the horizon is required for a good GPS fix. The reason 4 are required rather than 3 is more than one 3 way fix needs to be computed so that the clock frequency error of the receiver can be calculated and compensated for.

If you are close to large buildings or structures, a whole sector of the sky will be permanently unavailable. This places a restriction on the avialbility of good fix angles between the satellites in view and degrades fix accuracy more often than not. However, there are periods of poor geometry in the viewable constellation at various times which causes fix degredation even with a well sited antenna.

Feeding NMEA position sentences into a PC running the downloadable software "Visual GPS" would allow you to better analyse the problem. This package provides a large scale static position plot, a "sky map" plot of the visible satellite constellation and a bar graph of the signal strength of the satellites in use. One can usually see the reason for high fix deviations visually from these graphics.

One major problem with modern multi-channel low cost GPS receivers is they make no attempt at rejecting the use of a satllite which is in "bad angle" with others. They operate on the mistaken principal that averaging position data computed from as many satellites as possible will give the best fix solution. If one of the satellites in use geometrically has a very low angle relative ot other satellites as viewed by the antenna, the influence of this on fix error is hugely higher than an averaging calculation assumes. This I believe accounts for much of the larger static fix deviations one sees.

In conclusion it is unlikely your equipment is at fault. It is more likely to be an antenna siting problem, a location problem or a combination of both.

 

[PuffTheMagicDragon]

[ QUOTE ]
I I can alter the anchor alarm range but (from memory) it is 0.01nm to 0.05nm step, 0.05nm is quite a bit of drift in a popular anchorage. At 0.01nm it will sound the alarm when it starts doing this.



[/ QUOTE ]

Re: Anchor Alarm....Well, maybe not if you consider that 0,01 would be less than 20 metres; that's even shorter than my drive. Surely you will swing more than that if you're anchored in 5 metres of water?

Re: Popular anchorages. Wouldn't they all swing round in unison (assuming some wind and / or current)? If that is doubtful, then I don't think that I'd be happy to stay there.

My set is a Garmin and, while most times it locates itself at more or less the correct distance from the pontoon, there are sometimes occasions when the boat goes 'walkabout' even though well-secured, fore and aft, bows-to the pontoon!

I would just live with it.

 

[srm]

As others have implied, you are simply seeing the natural errors of the GPS system - which come from a variety of causes. I have not been able to source any definitive values since SA was set to zero but basically you should be able to expect uncorrected GPS to be within 24 metres about 90% of the time, for the other 10% it will be worse. Trouble is we don't know when the 10% will occur. Differential systems such as WAAS /EGNOS can improve on these figures when working but there will still be system errors.

You have nothing to worry about, but don't trust your anchor alarm. Most manufacturers seem to assume that everything works according to the highest possible specs. But then they rarely take the kit into the real world.

Remember, any electronic equipment is only an aid to navigation, and always has system errors. All those decimal points on the digital display may only be an illusion of accuracy.

 

[duncan]

[ QUOTE ]
basically you should be able to expect uncorrected GPS to be within 24 metres about 90% of the time, for the other 10% it will be worse.

[/ QUOTE ]

am I the only one who doesn't seem to be experiencing such variation in practice?

whilst 24m shouldn't be an issue for basic navigation I have a number of fishing marks that basically a small rock on a flat bottom (say 3m round) and I would be looking forever at that accuracy.

I have differential units, older non diff units and a new Lowrance unit (3500C) that is almost certainly picking up any old WAAS/Egnos transmission it can. From time to time I cross check the position they are giving and they agree - equally if I save a position on 2 units I can revisit it from either.

As has already been said the biggest error is likely to be in relating position to charted objects but again experiences last week in and around the rocks south and west of Alderney would suggest that this is pretty damm accurate too.

The latest unit (3500) certainly does have an option in software menu to adjust sensitivity to changing position readings but setting that to a high figure doesn't seem a clever solution the the initial poster's problem!

 

[Guest]

WAAS ..... interesting item

When playing with my 3500C ... I changed the WAAS setting to ON and I lost "lock" ... it wouldn't lock on sats and psoition kept flashing. Day before no problem. I then reset to factopry defaults - cause I couldn;t figure out what was wrong and bingo - spot on again. Checked through menu items to see what was different and that was it.

 

[Ships_Cat]

am I the only one who doesn't seem to be experiencing such variation in practice?

We seem to be the same Duncan. Our GPS only displays (sensibly I think) to 0.01 degrees and in the marina it never changes the last digit (the sentences it sends are to more decimal places and I can display them but have not taken much notice of them).

Again, on the ECS when in the marina the plot does not wander around and stays within a boats length (so 12 m or so) in our berth as is expected from the stated accuracy of the GPS. However, it is possible our ECS averages the position even when the boat is not moving so hides any excursions. But as said above, I have never noticed any on the GPS display itself either and it is always on when we are on the boat in the berth which is very often.

Both the ECS and GPS pick up quite fast movements in big wind gusts when we are at anchor so I suspect GPS positioning excursions would show up as well unless very fleeting - but we don't see (notice?) any.

I have downloaded the program OldHand led us to (for which thanks are due) and will run it on the boat Friday and Saturday when there to see what happens.

John

 

[ShipsWoofy]

This is a 435i, it is a stand alone unit (internal aerial). My previous Geonav never did this, hence my interest. The two units are in approximately the same position on the boat, they are almost under the mast. If I had completely relocated with the new set I might suspect a shadow, but surely every yacht is going to shadow the gps with masts or booms.

The mast base can just be seen in the top right.

This has happened about 4 or 5 times now, normally at anchor, it has happened when we dried on the beach, dead still absolutely no reflections except from sand hills. Just the mast.

I do have WAAS/EGNOS activated, I will try switching it off and see what happens.

This is not drift, it jumps up to 50m in one shot, think like the old TV game pong, bip bip bip.... backward and forward see above in thread. But it is not constant, it will jump, pause, wait, jump, pause and so on.

The top photo may be a little confusing, the long line going off the screen is the ships heading not the track, just in case anyone is thinking that is a bad jump.

Also on passage at least once during the voyage I will lose differential data for a minute of so and then it will go back and be fine, I thought? this is normal. My Geonav would occasionally lose lock, but unlike the RM it did not have an on screen error message, the ship would just go to 4° and 0.0kts for a minute.

Maybe, I should start suspecting installation, settings etc.

As to the sat page, always 5-6 greens (sats in use, good signal) with 4-5 blues (spares) and 2 empty, never found 12 satellites (as far as I know). When the differential lost alarm goes off the satellite page resets and a full search ensues.

Last year I posted a message to the RM fault reporting pages, but received back a wishy washy answer. It was also a problem as RM add a 48 hour self destruction to any answers they give so the page/service is useless unless you are on the boat. At the time I would not see the boat for a few weeks and as this is intermittent I may not catch it failing over a weekend on board.

That is I think all the relevant information I can recall right now. I would like to thank everyone so far. A little confusing though, to some this is normal, to others not. My biggest beef, the Geonav which was about 4 years older technology wise was steady as a rock when stopped. I would hate to think I have gone backward in technology.....

 

[rickp]

I'm with Tome on this - I would not enable WAAS or EGNOS at the moment, as the EGNOS system isn't active so you may end up getting atmospheric corrections which are only pertinent to North America (via WAAS).

Rick