[whifof100ll]

My 396 was flawless for the first year, but lately I've had a few strange problems. Two months ago, it began to have problems locking onto sattelites. I called Garmin tech support and they said that I should upload new firmware. I did not see why it shoud need new firmware since it worked perfectly for the previous year, and why should firmware go bad? I did it anyway, and sure enough, no more lockup problems.

On the first few flights, the 396 worked fine. Today I was seeing calculated wind speeds from on my EFIS of 100 knots. I knew this was not right, and first suspected that the flux gate sensor had come loose. I landed and lined up with several runways, and the magnetic heading was spot on. I then began to fly ground paths where I knew the direction. The EFIS read the proper magnetic heading, but the ground track reported to the EFIS, Tru Trak, and displayed on the 396 was off by about 40 degrees. Winds were less than 20 knots. No matter what heading I flew, the ground track was consistenly off by about 40 degrees to the right.

I drove home with it in the car on road where I know the heading. Sure enough, the groud track was off by about 40 degrees. It will draw the proper line on the map that you can follow and get where you need to go, but it will not give the proper ground track.

Must be another firmware bug, but this is certainly not a failure that I would have expected could happen.

 

[whifof100ll]

PIREP, update on the mystery

A PIREP for anyone using a Garmin 396.

While we are not supposed to depend on handhelds for IFR, I'll bet many use 396 and 496 data to augment situational awareness during an approach. I?m writing this as an advisory, because if your Garmin handheld does what mine has been doing, it could cause some in flight confusion just when you don?t need the distraction. Ask me how I know.

I have traced the problem down to the 396 setting it?s parameters for magnetic variation to user defined, and then setting the user value to some random number. Today on a flight from Houston Southwest to Hammond Louisiana, (1.5 hours) the Garmin reset this parameter 4 times. Since I had finally figured this out and was ready for it, each time I would go the settings menu and change the variation back to true course. It would fly along problem free for a while then change. I usually picked this up as a disagreement in the magnetic heading and true course. On one occasion, I had the Tru Trac in heading mode flying along on my intended course, then after 20 minutes or so the plane took a right turn all on it's own. Reminded me a little bit of the movie Christine. Sure enough, the 396 had reset itself again.

Since I have the serial link in to/from the AFS 3500, I still have to rule out some anomaly in the data link that could be causing the problem. I?ll bring this up with both Garmin and AFS. I will say however that this feature did not ever appear in 300 hours of use before I loaded new 396 firmware.

I also did find out that all WAAS Garmin handhelds require new software as the FAA has put up a new WAAS satellite that does not meet the original spec.

 

[ronschreck]

A PIREP for anyone using a Garmin 396.

(SNIP)

I also did find out that all WAAS Garmin handhelds require new software as the FAA has put up a new WAAS satellite that does not meet the original spec.

The L5 third civil signal from the IIR(M)-20 satellite launched March 24, 2009 successfully began transmitting on April 24, 2009. If you have not updated your software after that date you are not getting full use of all the satellite signals available.

 

[mburch]

I have traced the problem down to the 396 setting it?s parameters for magnetic variation to user defined, and then setting the user value to some random number. [...] Since I have the serial link in to/from the AFS 3500, I still have to rule out some anomaly in the data link that could be causing the problem. I?ll bring this up with both Garmin and AFS.

Are you using the "Aviation" serial format on your 396? If so, be aware that the 396 will obey whatever magnetic variation value it receives when used in this mode. If the other device is sending bad data, that could be your problem.

mcb

 

[Jamie]

...or more likely just a bad block of flash memory.

It's interesting to me that the GPS would output the calculated magnetic heading over NMEA as opposed to the GPS heading -- regardless of any setting. Really, really weird.

 

[mburch]

It's interesting to me that the GPS would output the calculated magnetic heading over NMEA as opposed to the GPS heading -- regardless of any setting. Really, really weird.

The 396 does not know anything about heading, but the NMEA 0183 data that it transmits includes GPS ground track and local magnetic variation. When using the Aviation serial input format, the unit will also update its idea of local magnetic variation from the input data it receives.

mcb

 

[Jamie]

The 396 does not know anything about heading, but the NMEA 0183 data that it transmits includes GPS ground track and local magnetic variation. When using the Aviation serial input format, the unit will also update its idea of local magnetic variation from the input data it receives.

mcb

When I say "GPS heading" I of course mean GPS ground track.

 

[TThurston]

Avweb Explanation

The following is a copy of today's Avweb headline:

URGENT: BENDIX/KING DATABASE PROBLEM
Certain Bendix/King KLN and KLX products have been compromised by their latest database updates and the company says those products should not be used until the problem is corrected. Bendix/King adds that "it is imperative" that the units not be used for arrival, departure or approach operations until corrected. Jeppesen apparently delivered some bad data to Bendix/King that contained incorrect Dynamic Magnetic Variations for all terminal and en route waypoint records. Affected databases include Cycle 1001 databases either downloaded from Wingman Services or delivered on programmed media and received prior to Jan. 12, 2010. Affected units are KLN 35A, 88, 89, 89B, 90, 90A, 90B, 94 and 900 models; plus KLX 100, 135, and 135A models. To fix the problem, Bendix/King customers just need to acquire a new database that's already available.

 

[mburch]

The following is a copy of today's Avweb headline:

URGENT: BENDIX/KING DATABASE PROBLEM
Certain Bendix/King KLN and KLX products have been compromised by their latest database updates and the company says those products should not be used until the problem is corrected. Bendix/King adds that "it is imperative" that the units not be used for arrival, departure or approach operations until corrected. Jeppesen apparently delivered some bad data to Bendix/King that contained incorrect Dynamic Magnetic Variations for all terminal and en route waypoint records. Affected databases include Cycle 1001 databases either downloaded from Wingman Services or delivered on programmed media and received prior to Jan. 12, 2010. Affected units are KLN 35A, 88, 89, 89B, 90, 90A, 90B, 94 and 900 models; plus KLX 100, 135, and 135A models. To fix the problem, Bendix/King customers just need to acquire a new database that's already available.

Not a problem for Brand G products, fortunately.

mcb

 

[whifof100ll]

The mystery may be solved. 296/396/496 users please read

This is Rob Hickman's response to my post on the AFS support forum.


*************
"The 296/396/496 have a new bug/feature that always set the magnetic variation from the external data. You need to disconnect the RXD line on the GPS from the EFIS or set the EFIS serial port type to "AFS GPS". This will keep the EFIS from sending data to the GPS."
**************

This may also be an issue with other EFIS systems x96 combinations.

 

[mburch]

"The 296/396/496 have a new bug/feature that always set the magnetic variation from the external data.

I hate to say "it's a feature, not a bug" but it's a feature, not a bug.

All Garmin handheld units that have ever been able to receive Aviation-format serial data have always had this same behavior, going all the way back to the GPSMap 195, circa 1997. It's intentional that they obey the magnetic variation received from the remote unit, which is assumed to be a certified panel-mount navigator (something like a GNS 430). If you think about it, it should be apparent why this is so.

In the case of the original poster, I'm fairly confident that the third-party EFIS system is responsible for the problem; the 396 is simply using the magnetic variation data it is receiving, as it is designed to do.

good luck,
mcb