CL Calibration

[bn999]

funth1ngs: It was recently found that the esc32.h header file in the ground software source was outdated and missing some parameter defines. If you compiled from this source, there is a chance that this may have caused or contributed to your problem. The repo has been updated, you might try again.

[funth1ngs]

Hi Bill!

With current limited to 15A or 10A the motor still stalls, but the calibration works now with 5A. With the old header file it didn't, maybe a coincidence?

In the zip file is a small patch to catch a SIGINT and disarm the motor.

Cheers,
Heiko

[funth1ngs]

Ok!

For tests with different motors I had reset the ESC to default values. With ADVANCE=25 and MIN_PERIOD=100 the calibration works even with 10A and 15A.

It's much better with the new header file!

Thanks,
Heiko

[sandmen]

Heiko,
did you try the qgcs for calibrate the esc32?

[bn999]

Heiko,

That's very good news. Thanks for the shutdown code, I've been meaning to add something like that for safety. I might go a little further and just accept any key to disarm.

[funth1ngs]

[bn999]

The current limit set on the command line of esc32Cal is only used while collecting the data. After the CL terms have been calculated, they are using in conjunction with the MAX_CURRENT onboard parameter. In the configuration show above, it is set to 20A. If you are trying to stay under 10A, then it must be reduced.

[kinderkram]

[Wingspinner]

SimonK may work on the bench with Simonk software but it is a well known problem that under load they get out of sync and perform terribly. After much effort and experimenting I gave up on Simonk software and was able to get wii-esc to work very well. I've posted about this both in the Simonk forum on RCGroups and also on the multiwii forum. "High" timing is advanced timing but the real problem with Simonk software is the way he does zero crossing detection. I think the reason it hasn't been fixed is he didn't actually write the original algorythms in his software as it its a re-write of someone elses software where Simon has added features but not really done much, if anything with the actual motor algorithyms. I worked with him and got the impression that he didn't know how to fix it and had really given up.

I'm not an expert on motor control either but I did find that with the wii-esc firmware using the option of advanced timing by one increment (I forget how my microseconds that is) the MT3506's work beautifull. It's good to know that they work fine on MT3506 however I would not be confident in that unless it was tested in actually flying conditions where RPM's and load are changing rapidly. They can be made to work fine on the bench with Simonk as well but not in flight.

Cheers,
Ron

[funth1ngs]