Imperfections in Garmin's Aviation Databases

 

The following compilation errors are present in the aviation databases sold by Garmin from their website for various aviation handheld GPS units. If anyone would like to checkout the panel-mount models, please let me know what you find.

 

The errors listed are taken from cycle 0308 databases for the GPSMAP196.

 

1. Invisible Waypoints

 

Units affected: GPS 92, 96, 96C, 195, 196, 295, 296, and III Pilot.

Cycles affected: All databases since the model became available.

 

Certain waypoints are not visible on the map page if you fly past (or scroll to) the area in which they should be displayed. They only become visible if they are included in the active route. The reason is that any waypoint which has a longitude roughly between W179deg40min and W180 fails to be included in the internal index, and is due to the fact that the binary resolution employed does not exactly fit the decimal limits of  plus and minus 180.

 

Waypoint        Country            Databases Affected

 

APT NFNM Matei  Fiji/Tonga - NF    pac amr atl

 

INT ALDOZ       Alaska - PA            amr

    FALAV       Alaska - PA            amr

    ARTOP       USA - P            pac amr

    BOILS       USA - P            pac amr

    NEILS       USA - P            pac amr

    MAZZA       USA - P            pac amr

    FROTH       USA - P            pac amr

    IDOCA       USA - P            pac amr

    MILDA       USA - P                amr

    DUNAK       Fiji/Tonga - NF    pac

    POPAS       Fiji/Tonga - NF    pac

    ATOLL       Fiji/Tonga - NF    pac

    PARZO       Fiji/Tonga - NF    pac

    90S80       New Zealnd - NZ    pac

    APOTO       New Zealnd - NZ    pac

    GONAX       New Zealnd - NZ    pac

    INTIB       New Zealnd - NZ    pac

    KADEN       New Zealnd - NZ    pac

    POPIR       New Zealnd - NZ    pac

    BABAK       Russia - UH        pac     atl

 

NDB LK Lakeba   Fiji/Tonga - NF    pac

    OI Ono I Lav Fiji/Tonga - NF   pac

    MT Matei    Fiji/Tonga - NF    pac

    NB Uelkal   Russia - UH        pac     atl

    WJ Mys Schmidta Russia - UH    pac     atl

 

2. Search for Airport by Facility Name

 

Units affected: GPS III Pilot

Cycles affected: atlantic since 9908, and all areas since 0307.

 

This really is due to a database compilation error, which results in the search for an airport by its facility name not working. Do not be put off by tales of it never working, or never being intended to work, it works just fine without the compilation problem. The situation came about when they were forced by space limitations to remove all the brief "out-of-area" airport details from the database (these are still present in other models). It is actually quite a serious error, because they failed to remove all the associated names, and left the facility name internal index full of rubbish entries. When you re-make one of these databases you get an immediate saving of 258k bytes (atl) and the search facility works again. In fact, had they fixed this problem, there would have been no need to split the world up into five lumps instead of three.

 

3. Duplicate SUA entries

 

Units affected: all models

Cycles affected: all cycles

 

This is rather less serious, since the numbers involved are less than ten areas in each database, but nevertheless there are instances of SUAs which are identical in every respect. There is no obvious effect that the user can observe.

 

4. Corrupt SUA Entries

 

a) Units affected: all models

Cycles affected: pacific 0307 and later, perhaps earlier.

 

The first entry in the SUA(b) section - areas not associated with an airport - is an exact copy of the first entry in the SUA(a) section (GOROKA) except that its name pointer is rubbish (too large).

 

b) Units affected: GPS III Pilot, GPS 92

Cycles affected: europe (eer) 0507 onwards

 

They really blew it with this one, they added a whole batch of 529 new SUA(b) areas, but the compiler didn't recognise the airspace type and overwrote the pointers to the names.

See complete list here: SUACorruptNames0509.txt

 

 

5. Incorrect ARTCC Stations

 

Units affected: all models having ARTTC data

Cycles affected: all cycles since the introduction of ARTCC data; fixed in 0311.

 

Some stations have a completely incorrect list of frequencies, and in most of these cases, a center is shown with only one (incorrect) station when it should have more. This is an error made by the Garmin database compiler, and appears to happen when the last (or only) station of a center is co-located with the first (or only) station of the next center. All the correct data is present in the image, it just is not referenced.

 

Examples are:

 

The SCOTTISH center shows one station from SHANWICK OCEANIC center.

The VARNA center shows one station from the SOFIA center, when it should actually have two stations.

The MURMANSK center shows one station from the MURMANSK OCEANIC FIR center, when it should actually have two stations.

Finally, and most dramatically, the SALT LAKE CITY centre shows just one station (the last station from the previous LOS ANGELES center) when it should actually have forty stations.

 

 

20th September 2005