GPS failures

[tmmorris]

Another thread got me wondering:

Has anyone seen a GPS error where the GPS reported a completely erroneous position, worse than the typical error of the NDB/ADF combination? I would include handhelds in this question.

The only errors I've seen were where the GPS lost signal/position lock altogether for several minutes at a time. I wondered if actual wrong position reports were a genuine problem, as opposed to a theoretical/regulatory one?

Tim

[High Wing Drifter]

I've seen an incorrect position, twice. Once on a P350 Pocket PC with Pocket FMS which I believe is a specific bug with NEMA sentences on the P350 / P500 (now fixed apparently) and once with a Skyforce IIIC.

[JP1]

In a nutshell your relying on a third party to provide accurate timing data. That means the tracked satellites correctly reporting their position in the Nav data that your GPS receiver decodes.

The satellite is in an orbit and its position is regulary updated (every 12 hours) as it overflies one of 3 tracking stations (if I remember correctly, but this figure could have changed). If its updated incorrectly then it will cause your nav solution to be in error somewhere near to its own postion error. Also the satellite could go u/s and hence before its detected it could be providing incorrect data if it's not setting its health flag to unwell!

Not sure of the probabilities, but it has happened.

When I left this field of work autonomous systems were being developed to enable the GPS satellites to range from each other and correct errors in their modelled orbit to improve the accuracy of the system between the 12 hour updates of orbit data.

Also most GPS receivers are "all in view" systems. This means they can compute many position solutions based on different satellite constellations and throw away any satellite that seems to be giving an erroneous calculation. Not sure if the civillian receivers employ this technology (or even if the military ones have implemented it yet)

So there is robutness being continually build into the system. But as said earlier it only takes human or "mechanical" error and the system could fail. But that's the same with any system.

[cjhants]

once only, garmin 430 flying from shoreham, i passed the midhurst mast, and about 2 miles north the garmin was showing i was still south of it, corrected itself a few seconds later.

[Rod1]

Get within range of Sennybridge when the jammer is working and you will get some very strange results. My friends pilot lll showed 350kn ground speed!

Rod1

[IO540]

Over 7 years of flying long distances I've had several total (and thus obvious) GPS signal losses. These could have been caused by jamming or, in the last case which was today, by the GPS receiver being zapped by the multi-kilowatt radar as I taxied in front of it at a certain airport in the SE...

I have only once seen a subtle error, and that was using the now-discontinued Jeppesen Flitemap product (which is identical to the current Flitestar flight planning product but with GPS input, so you can run it as a GPS moving map) which showed my position a few miles off track, somewhere north of Crete. It was running on a tablet computer and was used for monitoring against a panel mounted GPS. However, this software is so bug-ridden anyway I would not like to rely on it too much. It would not suprise me that the unexpected and commercially nonsensical discontinuation of Flitemap (which suprised most Jepp users) was actually due to this issue. Now, their only moving map product is Flitedeck (which comes bundled with Jeppview 3) which I manage to crash every time I try to use it.

On the whole, GPS is exceedingly good. Problems are usually obvious, and (more importantly) there are few if any human configuration/operation errors which are not obvious. I am talking about a moving map unit with a decent screen, obviously; a GPS without a decent moving map is wide open to human error and that is even before humans start loading waypoints into them manually which is one thing you should have to do only in the most unusual cases (like the pos of your farm strip).

[bigfoot01]

I have 3 learnings about GPS, which may or may not be on thread. Firstly, I always use SirfStar III Chipset in GPS for PDA's, Car Sat Nav or the hand held GPS I use for flying (Garmin 96Cx) this isn't a specific flying GPS, but I use it in conjunction with Memory Map and I like it. (I have tried memory map on a PDA etc and I didn't get on with it!) The handheld GPS is more rhobust than PDA's, simpler and it takes 2 AA batteries, which is easy to change on failure.

So the SirfStar III Chipset seems to be a lot more sensitive than other chipsets and as such maintains the signal and locks faster than other units I have seen. It is now the only thing I will use. I have never lost the signal when flying or driving with it (except through tunnels and I don't fly through those very often!)

I have a degree in computing, so I never read instructions. I pottered out on a 'local flight' to test it. I had the unit in Course Deviation Indicator. I now know that this will point a relative bearing along the line between two points and the VOR like break in the middle indicates how far off course you are. I became disoriented because I was expecting it to point at the next waypoint, but it wasn't I thought I really new where I was, but I was getting information I did not understand and I did not like it. Now I was tracking the M1, so nothing was really lost, but I about turned and got back to the VRP Pronto and was pleased to see it. I will tell you a valuable lesson was learnt there. This kit is complicated, you need to understand how it works, make sure it is programmed correctly (and be ready for a mistake!) and you need to be practised in it's use. Just like any other piece of flying kit. I now use it as part of my navigation tool set, along with may watch, compass, DI, VOR, DME, NDB, on board GPS, etc. etc. etc. When I jump in my car with my Tomtom, I don't give it a second thought - but it is definitely different in the air! I practiced with it in the car, parks etc. etc.

I wish they spent as much time teaching about GPS including handhelds as they do about the VOR, during PPL.

[Nipper2]

In answer to the original question, no.

I am always surprised by the who-ha surrounding GSP failure modes given that NDBs (which we do rely on for non-precision approaches) have a whole host of known subtle errors and no possibility of integrity checking.

Don't get me wrong - I'm not saying there are not nor can there be any GPS errors. I'm just pointing out that we do already use something that is basically know to be unreliable at best an yet we get very excited about much less likely problems with GPS.

User errors are another story but don't get me started on that....

[Droopystop]

Have lost the RAIM a couple of times, but not for more than a few moments and the GPS unit did loose its ability to calculate the w/v which is really frustrating but no great disaster if you have a whiz wheel ;)

On a different note, is there any feedback on the GPS approach trial that was being conducted last year?

[A and C]

I fly about 900 hours a year with the GPS being the primary navigation tool and have only seen two GPS failures in the last ten years.

The GPS flagged the failure both times.

I have had no "incorrect" position problems if this had happend it would have been picked up by the other inputs to the FMS system (2x IRS 2x VOR 5x DME).
The results in the "big" aircraft are reflected by the performance in my light aircraft with an IFR approved GPS fitted, in my opinion most of the problems are with non-approved systemes and I would guess that poor aeriel position and getting tangled in the yards of wire that conect these things to the cigar lighter socket are the most lightley reasons for faiure. . . . . not the actual GPS unit!

I have flown a few of the UK CAA GPS trial approaches in my light aircraft and it all went very well, the sooner that ADF approaches are replaced by GPS the better.

[IO540]

On a different note, is there any feedback on the GPS approach trial that was being conducted last year?

Last I heard is that the CAA is going to introduce them this year.

Presumably, at the airports that have the existing NP approaches.

No progress (that I am aware of) on the real key issue: doing away with mandatory ATC for an IAP. Mandatory ATC would cost somewhere into 6 digits per year even for day coverage and this prevents most GA airfields getting GPS approaches. I know people are plugging away on this one.

[neilcharlton]

If you have 2 airplanes to choose from; one fitted with a Garmin GPS the other with a ADF , VOR , DME ...
I know which one i would take !

[Droopystop]

If you have 2 airplanes to choose from; one fitted with a Garmin GPS the other with a ADF , VOR , DME ...
I know which one i would take !
So do I, the one with the ADF, VOR, DME and carry on my hand held GPS. But then again I would take a map too and do in manually anyway. More fun!:p

IO540,

I agree, GPS approaches with mandatory ATC doesn't really equate to progress, but maybe, just maybe, this is the thin end of a wedge.

[IO540]

If you have 2 airplanes to choose from; one fitted with a Garmin GPS the other with a ADF , VOR , DME ...
I know which one i would take !

Neither would be of much use. Both would be illegal for airways. The 2nd would be OK with a handheld GPS added.

[jai6638]

Why would the second one be illegal if it has a VOR and an ADF in there?

[IO540]

Because, in Europe, above FL095 (which is where most airway routes will be), you need BRNAV, and the only way this can be met practically in the GA context is an IFR GPS.

You can meet it legally with a KNS80 and antenna filters for FM immunity, but it won't be much good for real because airways ATC treat VORs as waypoints and will happily give you a DCT to a VOR 300nm away.

[London Mil]

I saw an 8km error in Hungary a few years back. We never figured what had gone wrong but suspected that the database differed from the charts we had been given (not helped by the non-WGS84 grid on the charts).

[172driver]

Had one a couple of years back. Bimbling around in an area I know well north of the Alps in Austria, I noticed a 2-3 mile error on the Garmin 430. Never been able to establish what happened and why.

Other than that the occasional loss of coverage on the handheld, usually for a few seconds only.