Possible Missing Data - Skyview log to USB

[Danal]

I believe that my Skyview has lost some data in the "User Log".  I have NOT changed sensors, or reset anything (of which I know).

Here are the filenames and timestamps of the contents, from the files produced by the most recent two USB unloads. The values shown are Session Time, GPS date & Time, System Time from the first and last row in each file.

Copied to USB on 6/15:

2013-06-15-N2750-5.1.0.1457-A-USER_LOG_DATA
741.38 3/30/2013 13:00 13:00:15
Only one line in this file

2013-06-15-N2750-5.1.0.1457-A-USER_LOG_DATA-PART_02
4191.06 4/13/2013 15:50 15:50:44
Only one line in this file


2013-06-15-N2750-5.1.0.1457-A-USER_LOG_DATA-PART_03
321.38 4/20/2013 21:57 21:57:31
1605.69 6/15/2013 17:25 17:25:09
Approx 158758 lines in this file.




Copied to USB on 6/24:

2013-06-24-N2750-5.1.0.1457-USER_LOG_DATA
1600.44 6/15/2013 17:25 17:25:03
1627.38 6/15/2013 17:25 17:25:30
Approx 60 lines in this file.
Note the timestamps 'overlap' with the data USB copied on 6/15. 

2013-06-24-N2750-5.1.0.1457-USER_LOG_DATA-PART_02
6065.75 6/22/2013 0:12 0:12:38
Only one line in this file.

There is a gap in data either here, and/or in the prior file, vs. actual flights flown.  In fact, the data below in part 3 starts in the middle of a flight. At 06/22 13:16 Zulu, I was in the air over OK or AK or MO on my way from the DFW area to Northerm MO (KMYJ to be specific).  ALL legs of the return flight were properly captured; however, part(s) of the outbound flight(s) are missing.

Specifically, a leg was flown from T31 to KSLG departing about 17:20 local and landing about 20:15 local. This is entirely missing.  I stayed the night in Siloam Springs.

A leg was flown from KSLG to KMYJ departing about 7:15 local and landing at 09:24 local.  About 2:15 total in the air. The last 1:08 of this flight is contained in the part 3, 4, and 5 files below. The data just suddenly starts in the middle of this flight.  Note that the Dynon was NOT power cycled, reset, configured, or anything else other than the overnight power down of the Dynon in Siloam. 

2013-06-24-N2750-5.1.0.1457-USER_LOG_DATA-PART_03
5808.5 6/22/2013 13:16 13:16:30
6501 6/22/2013 13:28 13:28:02
Approx 2772 lines in this file.

2013-06-24-N2750-5.1.0.1457-USER_LOG_DATA-PART_04
6501.19 6/22/2013 13:28 13:28:02
Only one line in this file.

2013-06-24-N2750-5.1.0.1457-USER_LOG_DATA-PART_05
6501.38 6/22/2013 13:28 13:28:02
132.06 6/24/2013 0:28 0:28:35
Approx. 141757 lines in this file.

Part 5 contains part(s) of the data for the outbound flight, and contains complete data for all my return flights.  Because of headwinds, I stopped several times.  The Dynon was power cycled at each stop (for fuel).  There were 5 power cycles (or more).  Yet all the data is here, and in one file.



This raises numerous questions.  Why are the part 3, 4, 5 so different in size?  4 has only one line... yet it was right in the middle of a flight.  the end of 3, the one line in 4, and the first line in 5 form a continuous data set... good stuff... then 5 has numerous flights... so many questions about how this is split up.  What causes the Dynon to end one part and begin another?  Obviously not a power cycle.  Obviously not file size.  What does?

But... the real questions... WHERE IS THE  MISSING DATA?  Why did it go missing, when there is data 'before' and 'after' it? 

Thanks,

Danal Estes
CH750 N2750
Dynon Skyview D10

 

[Dynon]

So the best way for us to see what you're seeing is to get a special backup from you. Can you PM your email address, phone number, and mailing address (we'll need to send you a special usb stick)? We'll give you a call to explain what we'll be doing.

 

[Danal]

PM sent. Thanks for your support!

Danal

 

[Danal]

Special "Dump the system" memory stick received, executed, and returned. Tracking shows delivery at Dynon on July 15.

First, THANKS for taking the time to prepare and send me a system dump diagnostic.

Second, any word as to what happened?

Danal

 

[Dynon]

I'll have to check with engineering to see if they've had a chance to analyze it yet. A lot of us are at Oshkosh this week, so we may not have an update right away.

 

[johnsteichen]

I have one USB port on my panel. I use one of those very short memory sticks in that port. It remains there and does not get in the way like the supplied dynon stick

 

[Danal]

Just checking, now that Oshkosh is over for a while, if there has been any findings from the data dump.

THANKS for taking a look at this.

 

[Dynon]

Yep, you've run into a couple of bugs.

The bottom line is that what you're getting as an export is kinda weird looking, but we don't see any real evidence of data loss.

The details:

It looks like there's a bug in our throwing away of the oldest log data that's causing you to get short weird fragments of old log data. This is obvious from the first two one-line parts from your export on 6/15 and from the first two parts from your export on 6/22. If we throw those away, then the picture is much less confusing.

What you're left with is the last > 8 hours of flight data that you should expect from the SYSTEM SETUP -> DATA LOG SETUP -> RECORD RATE configuration of "4 PER SEC: EST. 8 HOURS". Actually by my calculations, the first export containing 158758 records represents ~11 hours at 4 Hz, and the second export of 141757+2772 records represents ~10 hours. That's actually the default setting, so you may not have been aware of that correlation, but feel free to adjust it to get what you think you want. If you're going to be doing long cross-country flights like that and don't need 4 records per second, you might want to get more time coverage by going to 1 second or even slower records.

Which brings us to the second bug... There is something causing a momentary splitting of the export to give you 3 "contiguous" log parts instead of one big lump. We've logged this bug previously, but haven't tracked it down yet.

Thanks for bring these bugs to our attention!