Accuracy of non-TSO aviation GPS receivers

[ForkTailedDrKiller]

Been reading the latest Flightsafety Australia. The article "Will it RAIM" states in relation to non-IFR GPS receivers "Errors in excess of 600 nm have been reported", and got me thinking.

I have now sat in front of a Garmin non-TSO 296/496 and a TSO Garmin 430/430W for more than 600 hrs (ie 90,000+ nm or the equivalent of 4+ times around the world!) while flying all over Australia (every State and Territory bar Tassie!), and in that time I have never seen them disagree by more than 1 nm.

So I am curious! Has anyone seen significant errors when using top-of-the-range non-TSO GPS units, either panel mounts or portables?

Dr :8

PS: Here's a gem from Flightsafety Australia.

"To undertake these arrivals correctly, the ground must be available and used".

I am going to add an item to my TOD checklist.

GROUND AVAILABLE - Check!

[Sunfish]

GPS does not go loopy very often, but when it does.....

Keep your paper chart, pencil, rubber and protractor.

[Monopole]

Keep your paper chart, pencil, rubber and protractor Keep your what:}:}:}

[RadioSaigon]

Interesting. Likewise, I have spent lots of hours/years with "high-end" handheld GPS on the stick (currently a 296), but never seen any significant errors at all! A lot of that time was in extremely mountainous terrain in dirty wx too, when you most need the GPS to be accurate. Even then with much of the visible sky obscured by terrain, it never let me down.

I've had the occasional drop-out, but the last one of those was so long ago now I'm not even sure I could yell you accurately what year it was! Probably around the time that Bill Clinton had his finger on 'the button' (amongst other things :}) and got rid of SA. 600NM error too... that's getting up there and something most pilots would surely recognise... come to think of it, isn't 600NM around the 'magic' figure for GPS where they need to do a full (and lengthy) almanac update???

Haven't read the article yet, but will have a bit of time on my hands tonight to do just that!

[18-Wheeler]

Keep your paper chart, pencil, rubber and protractor.

Pffft .... my mobile phone has a GPS in it. As long as I know the address of the airport I can just follow the directions. ;)


600NM error too... that's getting up there and something most pilots would surely recognise

I suspect the reality of this is someone entering the data with a touch of finger trouble, e.g. north instead of south, etc. That's the only time in many years of using GPS where it's come up with bad info - We were going to Palau (near the Phillipines) from PNG and the FO typed in the latitude of the island as being south. Distance was about right, but the track was way wrong.

[startingout]

tomtom car devices may or may not work just like that aswell :ok: works a treat if you get bored one day and are wondering how fast you are going in kms without needing the conversion chart

[VH-XXX]

Have never had an accuracy problem. Any accuracy problem that I have ever thought that I may have had turned out to be a user error. That is across the half dozen units I have owned plus permanent units in aircraft I have flown over 7 years.

[Biggles_in_Oz]

Errors in excess of 600 nm have been reported That's a great throwaway line in support of the expensive TSO'd gear, but it needs some substantiation because the GPS signal nowadays will give your horizontal position to within about a 10m circle with a confidence of 95%. (vertical accuracy and confidence is a lot less.)

600nm is a heck of a long way away from 10m, so I'd be inclined to ask questions such as,
- when ? (did the problem occur with an early model receiver)
- where ? (was it near a place that was testing GPS jamming or near a conflict zone that the USA had an interest in ?),
- for how long did the the error persist for ? (was it a just a 'one off' or did it persist for a decent time)
- what sort of unit was it ? (GPSs have different position and velocity filters depending upon their intended usage., a marine unit has a lot of smoothing and filtering otherwise one ends up on the rocks.., an aviation unit intended for approaches can't have such heavy smoothing therefore external augmentation (ala some form of SBAS) or a more stringent accuracy (RAIM) is required.)

[OZBUSDRIVER]

Forkie, if anyone is going to see an event I think you would have to be the prime candidate.

[18-Wheeler]

600nm is a heck of a long way away from 10m, so I'd be inclined to ask questions such as,


.... all those, and "did I type in 33°S instead of 23°S?", etc.

[RadioSaigon]

...entering the data with a touch of finger trouble...

A valid assumption, but in this case the article is definitely stating a case of a 600NM error in a GPS whilst navigating. Remarkable stuff. I can't help but wonder, is this just another case of the regulator trying to build a case against a magical 'new' box in favour of a 60-year old technology that as likely as not will point at any stray electron in the sky? Why??? The article struck me as subtly skewed, whilst still managing to provide some pertinent cautionary notes particularly relevant to the TSO units and RAIM messages.

Further examination of memory last night revealed what I thought an interesting aside: the greatest "lack of accuracy or misleading direction from a GPS" award would definitely go the database! Some of the information is just completely wrong (not to mention incomplete)!!! At one airport I know rather well, the GPS airport coordinates were about 1.5NM E of the RWY, somewhere in the middle of a river! The 1st time I saw it, I assumed SA or insufficient satellites in view... after that when it was still consistently wrong I thought perhaps a unit fault -but why? Other places were spot-on. So had a look at the stored coordinates. That was where the problem was. Checked the rest and found more errors. Eventually I hacked the database and sorted it for myself. The regulator won't sort it, Jepessen seem to be incapable of sorting it.

Like any other electronic marvel these days, it's 'garbage in, garbage out'. They're a great device, but to get the best of it, make sure that it's using correct information -verify from other published resources where you can!!!

[aileron_69]

I have a fully IFR GPS in the panel of my old girl and it is the most unreliable, useless pile of junk i've ever seen. It drops out and misreads all the time. My handheld boaties Garmin 76 tho is fantastic and I have never had it drop out or lead me the wrong direction yet.
The 24 channel GPS we use for surveying is pretty good too, accurate to 20cm as long as we have 5+ satellites locked on.

[Deaf]

Easy way of checking the repeatability of portable GPS. Accuracy when static at a known point is one thing, moving is another.

Stick it on the dashboard of the car for a few days (obviously not while parked) with your daily travels. Then zoom in on the breadcrumbs especially with regular turns. Something like 30m (no clear double concentration on well divided highways) seems normal with internal antenna. External antenna probably 10m.

Worst I have encountered is YSSY to Sydney city through sunken highway where the road algorithm has you on the side road and tells you to turn where you can't. Thats not really relevant to aviation use however.

[onthedials]

The article should have been clearer about the concepts of integrity and accuracy. There is quite a bit of useful information in that article, so the reason for the bit about inaccuracies in non-TSO is hard to fathom.

As it states, the purpose of RAIM is to provide integrity, that is, to assure the user that the information being displayed by the device meets specified standards and alert that user when it does not.

Accuracy is the concept of solution precision.

Accuracy is a factor in providing solutions with integrity, because it is necessary for the solution to be cross-checked and the user alerted if inconsistencies result. When that happens, the displayed position might (or might not) be accurate, but that accuracy cannot be validated.

The concepts of accuracy and integrity should therefore be separated. The accuracy of a recently-designed non-TSO receiver may well (probably, statistically, measured over a period of time) be superior to TSOd ones that are now 10 or 12 years old. The latter might offer integrity, but be less accurate than a non-TSO one that doesn't. Whether a pilot can discern that difference in accuracy is, of course, another question.

There are so many variables in the process of operating a GPS that anything might have caused the quoted 600nm error. Perhaps it was shutdown or standby in DR mode and then put with airline baggage and reactivated 600nm away without a view of the satellites. Or finger trouble, including wrong waypoint selection. It would have helped if the article gave even a clue as to what that error was and how it arose.

However it seems that the pilot community is now considered too dumb to understand even the most mildly technical subject. Almost everything we read from the regulator has been dumbed-down to be comprehensible to 8 year olds. CASA seems to think that such communications are even better if they throw in some alarming and dissuasive statements in preference to analysis and facts.

I'm certainly not saying that you should use a non-TSO GPS for approaches or any of the other IFR activities reserved for the TSO variety, but I am saying that perhaps highest among the reasons not to do so is one of integrity, rather than accuracy. Also, non-TSOd units are probably lacking many of the other important features (such as terminal and approach mode CDI scaling) that you need to fly an approach safely; but these aren't accuracy issues (of the receiver) either. And lastly, the safety of everyone else depends on pilots knowing when their GPS position lacks integrity.

Hope this helps. OTD.

[the wizard of auz]

I have been using a Garmin 100 for aerial mustering and nav for over ten years now with 0 error and never an outage. I can return to a waypoint (goat or bullock holed up under a tree) after tracking across a 20 mile wide paddock repeatedly all day. it has never quit on me in over 5000hrs of use.
I have used it to run lines in survey work, to crossing the equator and flying in Asia, again with zero problems. dunno why they are looked down on.

[Jabawocky]

Wiz mate

Coz it ain't Doppler nor an NDB! :}

J:ok: