Volkslogger Data Problem explained in detail.

I got this explanation from Garrecht Avioncs the manufacture of Volkslogger..

The GPS date problem shows up since end of October. Currently only from SouthAfrica and Namibia, where folks are flying now.

Not so many complaints currently, but enough to know that it will become more ...


This seems to be a new variant of the GPS week-rollover problem nobody has expected.

The cause is the Firmware on the internal Garmin GPS OEM receiver board. This board is installed in all Volksloggers, but I guess also into many other IGC logger brands (at least one of them we know has the same problem).

A quick analysis of the IGC flight recorder approval documents shows that the same GPS boards are built into
- all Volksloggers
- all Cambridge 302
- all Cambridge 10-20-25 models from 1997 onwards
older Cambridge models have still older 1-channel receivers, don't know the status of these

The firmware is, against all statements from Garmin some 15 years ago (when the issue was first discussed in the GPS community), not able to handle the week rollover properly, although there are solutions to this (most other GPS manufacturers know how to do it properly).

From all we could find out, Garmin have introduced a temporary solution at in the middle of 2000, causing all GPS OEM modules not to have a problem from 1996 to End of October 2014. Exactly, they function correctly from 1.January 2005 +- 512 weeks.


Outside this date range, all GPS25 OEM receiver boards with an init time of 1.1.2005 will have the same problem, as soon as they loose their internal time/date due to battery drain after some days/weeks (depending on the battery health).

Starting up without knowledge of last received date, they think they are 1024 weeks ago = currently, this is April 1995.

We had the problem 5 years ago with few receivers, namely the ones with an internal Init-Date of 1.1.2000. We've updated them to a newer (official) firmware, which seemed to have solved the problem, but see above.



We're working on two fixes.

1st solution
Is an (inofficial) firmware update, which (the installer software) as a "product" must not leave our house due to very possible license trouble with the Garmin company (they most probably don't like the idea that someone makes their products live longer than planned). Garmin does not support the GPS25 receiver boards in any way anymore, all software / documentation links on the Garmin website are dead.

Therefore, all affected units (or at least the GPS board) must be shipped to us, where we will fix the GPS receivers with our own firmware modification for a nominal fee covering the costs. Due to the amount of loggers sold, and expected return volumes, the processing time might become very long ...

Currently, we are calculating the costs and planning a solution for coping with the high volumes.


2nd solution
To be be negotiated with IGC within the next weeks, and can most probably postpone the problem to 2018 (first date of any Volkslogger sold + 1024 weeks), but requires a lot of programming work on very old software for many companies.

This would require a software update of all tools used for readout of the Volkslogger (Strepla, SeeYou, Garrecht DATA-GCS and VALI-GCS.exe files, IGC DLL file from Garrecht, all open-source tools, all PDA software, etc.). This software would recalculate all wrong IGC file content and names into the right, and making them.

Alternately, allowing all tools to show the wrong date at first place (on download from VL), we'd only publish a FIXGPS-GCS.EXE and a new VALI-GCS.EXE and DLL.



3rd solution (not very comfortable to the user, but least cost):
All Volksloggers with healthy GPS battery, still knowing the right date (as can be seen in the NMEA stream) should stay connected to a power source until used, and again after usage, to avoid a GPS battery outage after some days / weeks.


I hope this helps.

Thanks for this information, Richard. Please do post new information on this topic to the group, would you?

-John, Q3

On Wednesday, November 26, 2014 7:56:10 AM UTC-5, Richard wrote:
I got this explanation from Garrecht Avioncs the manufacture of Volkslogger.

The GPS date problem shows up since end of October. Currently only from SouthAfrica and Namibia, where folks are flying now.

Not so many complaints currently, but enough to know that it will become more ...


This seems to be a new variant of the GPS week-rollover problem nobody has expected.

The cause is the Firmware on the internal Garmin GPS OEM receiver board. This board is installed in all Volksloggers, but I guess also into many other IGC logger brands (at least one of them we know has the same problem).

A quick analysis of the IGC flight recorder approval documents shows that the same GPS boards are built into
- all Volksloggers
- all Cambridge 302
- all Cambridge 10-20-25 models from 1997 onwards
older Cambridge models have still older 1-channel receivers, don't know the status of these

The firmware is, against all statements from Garmin some 15 years ago (when the issue was first discussed in the GPS community), not able to handle the week rollover properly, although there are solutions to this (most other GPS manufacturers know how to do it properly).

From all we could find out, Garmin have introduced a temporary solution at in the middle of 2000, causing all GPS OEM modules not to have a problem from 1996 to End of October 2014. Exactly, they function correctly from 1.January 2005 +- 512 weeks.


Outside this date range, all GPS25 OEM receiver boards with an init time of 1.1.2005 will have the same problem, as soon as they loose their internal time/date due to battery drain after some days/weeks (depending on the battery health).

Starting up without knowledge of last received date, they think they are 1024 weeks ago = currently, this is April 1995.

We had the problem 5 years ago with few receivers, namely the ones with an internal Init-Date of 1.1.2000. We've updated them to a newer (official) firmware, which seemed to have solved the problem, but see above.



We're working on two fixes.

1st solution
Is an (inofficial) firmware update, which (the installer software) as a "product" must not leave our house due to very possible license trouble with the Garmin company (they most probably don't like the idea that someone makes their products live longer than planned). Garmin does not support the GPS25 receiver boards in any way anymore, all software / documentation links on the Garmin website are dead.

Therefore, all affected units (or at least the GPS board) must be shipped to us, where we will fix the GPS receivers with our own firmware modification for a nominal fee covering the costs. Due to the amount of loggers sold, and expected return volumes, the processing time might become very long ....

Currently, we are calculating the costs and planning a solution for coping with the high volumes.


2nd solution
To be be negotiated with IGC within the next weeks, and can most probably postpone the problem to 2018 (first date of any Volkslogger sold + 1024 weeks), but requires a lot of programming work on very old software for many companies.

This would require a software update of all tools used for readout of the Volkslogger (Strepla, SeeYou, Garrecht DATA-GCS and VALI-GCS.exe files, IGC DLL file from Garrecht, all open-source tools, all PDA software, etc.). This software would recalculate all wrong IGC file content and names into the right, and making them.

Alternately, allowing all tools to show the wrong date at first place (on download from VL), we'd only publish a FIXGPS-GCS.EXE and a new VALI-GCS.EXE and DLL.



3rd solution (not very comfortable to the user, but least cost):
All Volksloggers with healthy GPS battery, still knowing the right date (as can be seen in the NMEA stream) should stay connected to a power source until used, and again after usage, to avoid a GPS battery outage after some days / weeks.


I hope this helps.

I posted this on the UK site, I'm wondering if you can send the $PGRMI
sentence to the VL and initialise the epoch correctly.

=================

Did some research on this and found the manual for the GPS 25 board in the
Volkslogger.

On page 16 para 4.1.1 it describes the almanac resetting procedure.
Sentence $GPALM

Page 17 para 4.1.2 describes $PGRMI which can be used to set the UTC date
and time.

So ... how do we hook up a terminal emulator to the GPS board, when the VL
goes into PC mode is it possible to send the NMEA sentences to the GPS.

I don't have a VL any more, but it seems like a problem worth solving.

John

http://www.google.co.uk/url?sa=t&rct...M7Fg&cad=r ja

On Wednesday, November 26, 2014 6:56:10 AM UTC-6, Richard wrote:

3rd solution (not very comfortable to the user, but least cost):
All Volksloggers with healthy GPS battery, still knowing the right date (as can be seen in the NMEA stream) should stay connected to a power source until used, and again after usage, to avoid a GPS battery outage after some days / weeks.



So, I suspect I best keep my 4 good ones powered up to keep that hidden internal battery charged.

And wait and see if there is another way to get the others back on track, short of spending nearly what they are worth to make them useable again. Would be nice if the "repair" also put the battery where it was owner replaceable when it does die. Just sayin....

On Wednesday, November 26, 2014 11:39:15 AM UTC-8, Steve Leonard wrote:
On Wednesday, November 26, 2014 6:56:10 AM UTC-6, Richard wrote:

3rd solution (not very comfortable to the user, but least cost):
All Volksloggers with healthy GPS battery, still knowing the right date (as can be seen in the NMEA stream) should stay connected to a power source until used, and again after usage, to avoid a GPS battery outage after some days / weeks.



So, I suspect I best keep my 4 good ones powered up to keep that hidden internal battery charged.

And wait and see if there is another way to get the others back on track, short of spending nearly what they are worth to make them useable again. Would be nice if the "repair" also put the battery where it was owner replaceable when it does die. Just sayin....

We have some news on the Volksloggers.

Hello Richard,


we're still testing our GPS firmware update fix. Next week, we'll get a GPS simulator to try things out, so that GPS modules will work for the next 9 or 19 years.

In order to completely remove the need to open the VL's at all, which is the ultimate goal, we're also preparing a firmware change together with a change in the validation EXE file for FAI/IGC / OLC.
This will take some more weeks. Consider that we have not touched the VL firmware the last 12 years. It's difficult to think into this again and find a quick solution for a change.


Anyway, if you are able to open a VL (you are), you're able to do both kind of service fixes in the US. The only thing we have to do (in both cases) is to send you the firmware files for resealing the VL's. I hope you have access to good anti-tamper seal stickers or the like.


best regards
Georg

One more confirmed date corruption on Volkslogger A1ZY.

05/15/1995

Walter Rogers WX

PS - Fortunately, this logger is a spare as I've upgraded to a Nano and PowerFlarm for IGC files.

Reason #42 to get a Nano! But for a less emotional explanation than Garrecht's of the how's and whys, this from the FAI:

To: Owners and Operators of IGC-approved GPS Flight Recorders
From: Chairman IGC GNSS Flight Recorder Approval Committee (GFAC) Date: 29 November 2014
Subject: Date Errors in IGC files from Flight Recorders

Summary: Owners and operators of GPS Flight Recorders (FRs), particularly of older designs, are advised to switch on the FR at regular intervals during periods when it is not in use for flights, in order to keep the FR and GPS receiver batteries from running down. This will minimise the possibility that when the FR is used again, a wrong date may be recorded in the IGC file.

Background.
Posts on Newsgroup Rec.Aviation.Soaring (r.a.s.) have shown that in some early models of IGC-approved Flight Recorders, the date recorded in the IGC file has suddenly started to be in error by several years. This has been noticed with older models of Cambridge and Garrecht Volkslogger FRs, and modifications have been offered to correct it. This error may also be present in other types of FR.
The error occurs because the small internal battery inside the GPS receiver has run down and the Real Time Clock (RTC) in the GPS engine has stopped. When the GPS is re-powered, the RTC is unable to reset to the correct date, for the reasons given below. It should be noted that this battery is not the same one that is used to back up the flight recorder's memory, but is a smaller battery inside the GPS engine itself.
The IGC GNSS Flight Recorder Approval Committee (GFAC) is investigating. It is contacting FR manufacturers so that they are aware of the anomaly, can inform GFAC whether their FRs could also be affected, and have the opportunity to take remedial action such as through replacing the GPS battery, changing the GPS receiver, or developing a Firmware update that corrects the date.

Technical Detail.
GPS dates are expressed as a week number and a day-of-week number after the date when the GPS system first came fully on line on 6 January 1980. The date system used was designed to cope with a maximum of 1024 weeks, after which it starts to count from zero again. There was no problem for 1024 weeks from 6 January 1980 until 19 years and 8 months later, when in August 1999 the week-count rolled back to zero. It will roll over again in May 2019.
Each period of 1024 weeks is called an "Epoch" and we are currently in the second date Epoch. A GPS engine with a functioning RTC can identify the Epoch rollover without problem. However, if the RTC sustaining battery fails, when the FR is powered up again it has no knowledge of which Epoch it is in and reverts to the first Epoch. The problem was made greater when some GPS engine manufacturers decided to replace relatively large-capacity memory batteries with smaller, rechargeable versions. These rechargeable batteries had a much smaller capacity and relied on being regularly recharged when the FR was powered up. They performed well for many years, but like all rechargeable batteries, their capacity reduces with time. After some 10 years or more, their capacity is reduced and some now fail after a few weeks unless they are re-charged by powering up the FR.
Some FR manufacturers solve this problem by replacing the GPS memory batteries and resetting the RTC, others replace the GPS engine itself, and it is also theoretically possible to update the FR firmware to a version that allows for this problem.
Finally, it should be noted that many modern GPS engines do not suffer from this problem and can retain the correct date for over 100 years.