Problem with step 3

[leopold1]

Doing the calculations, I'm having a hard time getting past step 3

I have 3 static log files that seem quite good. I allready used them with a quad setup without problems.
I did three dynamic logs and used all of them.

The first two calculation in step 3 are great. The MAE drops. When I try to do a third calculation,it never stops and keeps on going. It gets at an ever lower MAE but then it keeps increasing.

At a certain time I stop and restart but with the same problem.

I even started step 4 but then the MAE value goes up and down.

Any suggestions?

Dynamic log


Static logs

[leopold1]

Did the calcs with only the flat static and with a third dynamic file.

It keeps hanging each time when I do a third calc in step3.

Whats going on???

[LPR]

The only answer I have you wont like but it works very well. Don't use all the files. Use one or two static logs and the dynamic log. If you can follow the instruction and do the many step and redo the step looking for the lowest MAE you should be able to get through with all three but I was not able to so I use one or two static logs. You can use one static log and be flying in a day and then spend a few more days working out how to get the lowest MAE for all the logs to finish the calibration for the best parameter file.

Larry

[Max]

Hi Leopold,

I'm not really sure what you mean by the "third calculation [in step 3]." Do you mean running step 3 a 3rd time, after saving the generated parameters from the previous runs?

Anyway, step 3 is the first step to use the dynamic log. Step 4 doesn't use the MAG values calculated in step 3, but it does use the dynamic log. And if you used the 3 static files with success before, this would all seem to suggest a "bad" dynamic log. You said you tried 3 dynamics, but I'm not sure if you meant this time, or the previous time. If the latter, then I would suggest simply trying with a new dynamic. Make sure you're doing the dance correctly, not forgetting any of the sides/rotations, and be sure to trim any "noise" off the ends of the log.

You could also have data errors in your dynamic file... sometimes it's hard to spot the message when using QGC. You could try running cal or sim3 from a command line and check the output for checksum errors, or you could export the log using the function in QGC (in the chart viewer), which will also report any checksum errors in the log.

HTH,
-Max

[leopold1]

Larry, Max,

Just to be clear, its each time on the third run in step3, whilst trying to lower the MAE value, it just keeps on running.
The lowest value, last line during the calculation, stays at that value and the MAE keeps rising again. Calc never stops.

Things I tried:
1 static log (flat)
3 static logs
2 static logs (corner up & down)
3 different dynamic logs.

Still no joy

I'm going to get me a new sd card this morning.

Keep you posted.

[leopold1]

Got myself a new SD card and redid a dynamic log.
Now I got about 6 reruns of step 3 until the MAE began to increase again.
I stopped manually and continued with step 4 but again the values started to go up and down.

Checked for errors with the export function in the GCS but no errors were found, even in the previous files.

What the hell is wrong here!!!!

[LPR]

Let us see your logs.

Larry

[leopold1]

Larry,

They are in the first post.

[LPR]

It would be much better if you would post a screen shot of the log curves your using.

Larry

[JussiH]

Do you get any Checksum errors in the files. Usually, if step3 wont complete, it points to problems with the mag data or checksum errors in the static files. Have you tried to just run a single static and single dynamic file?