Sigh - cannot get past static calibration

[moellenb]

Thanks a lot!
It feels good to hear that at least the statics are alright so far. Now then, I re-uploaded the zip with the dynamic. Please try again. If the archive (produced with Win 7 built-in zip) is still corrupt, I'll try to upload the files separately.

[JussiH]

still screwed - just upload the full file - dont zip it! I dont need the screendumps, just the raw dynamic file straight off the SD card

[moellenb]

The dynamics should be available now. These are the raw files from the SD card, i.e. without cutting off anything. The one I later used for calibration is number 004.

[JussiH]

If you want me to help you with this, you will have to put the files somewhere accessible. This "mediencenter" is about the worst online file storage I have ever seen. Packs things into zips when downloading. And it wont allow me to DL the new files.

Use a dropbox, pls

[JussiH]

Ok,

I managed to DL the files now.

004 is not usable, the gyro data is currupted.

005 looks strange, there are offsets in the gyros:

Bottom line is that none of these dynamic files is gonna produce a valid result. I am afraid you will have to redo them

[moellenb]

Okay, thanks a lot for your effort! I'll redo the Calibso, create a dropbox account, and post again when new files are ready for download.
In the meantime, I'd like to understand how the gyro data in file 004 could possibly get corrupted. Does this just sometimes happen when recording to the SD card - or is it likely that I did something wrong when I prepared the Calibso? I mean I inserted the card first, then put the copter on the ground, then plugged in the battery, then waited for the GPS fix plus a few seconds, then carefully picked up the copter, then stood still for another 5 seconds or so, then started the dance.

[JussiH]

The problem here is that you have an offset between the X and Y zeropoint. X and Y is supposed to be 1.38V when board is still.

The fact that one gyro did not initialize in the earlier file could point to a problem with the AIMU on your board. Most likely an intemittent connection. If thats the case, its an RMA case and you need to contact Flyduino to get it worked out.

[moellenb]

Yes, I understand your explanation of the screenshot (file 005). Which is a problem because now I feel I can't trust the board any more, even before the very first flight experiments.
But what about 004? In what way is that corrupted? And how can it happen?
Anyway, I'll redo the measurements on Wednesday this week. Let's hope for the best...
Oh, and BTW, I'm looking forward to meeting you in Hannover.

[JussiH]

Listen, I am very sure this is a HW problem.

If you had gotten the board from Viacopter, I would have offered you a DIMU replacement as the best possible solution. But you need to contact Flyduino and I am sure they will work out an equally reasonable solution for you in a hurry.

The most likely failure here is that one of the small solder joints on the gyros did not flow properly and a "cold" joint was made, resulting in an intermittent connection. This is the type of failure that might not show in initial board testing after production, but may become evident after the board has gone throuhg a freezing/thawing cycle due to contraction and expansion during the freezing/thawing cycle.

When this happens, it is a warranty issue and you can expect to get it repaired or replaced by your dealer.

[moellenb]

Yes, you must be right. I redid the measurements and the x & y gyro voltages show a persistent offset.
Anyway, I opened a DropBox account and uploaded two new dynamic files.

https://www.dropbox.com/sh/0mj48g2mrxkv ... VCEFGWHQJa

I've also had a look at the other voltages (is there a document about which voltage represents which sensor?) and particularly voltages 3, 4, and 5 look strange to me. Zooming in I can see the voltages jumping between consecutive samples. Whatever that means...

I'll contact Flyduino. Nevertheless I'd appreciate it if you could have look also at the new data.