CL Calibration

[funth1ngs]

[bn999]

Heiko,

That does not sound good at all. I have not tried the MT3506, perhaps others have. When the motor stalls, what does it sound like? If the ESC did not disarm, was the motor rotor locked with full current going through it?

I don't want you to toast a motor, but could you perhaps post a video of this happening? You should be able to get the same effect from either the CLI or via PWM input.

[bn999]

Thinking about this a bit more:

Do you have another MT3506 and does it happen on it as well?
Do you have another ESC32 - same thing?
Does this happen on any other motor you have?
Note the conditions of the LEDs on the ESC when this happens.

[funth1ngs]

[bn999]

The ESC has definitely lost sync. I have not seen it where the ESC does not disarm when this happens. A quick search shows people recommending that "high" timing be used for this motor. I don't know what they mean, but if you've already tried changing the ADVANCE parameter, you might also increase the MIN_PERIOD parameter to make sure the ESC does not treat very short "ghost" ZC detects as possibly real.

The value is in microseconds and it is the minimum time a commutation can take. If for instance, the highest RPM you could possibly achieve is 10,000 then your MIN_PERIOD could be set to 150us.

period = (3,000,000 / (2 * RPM))

If none of that works and the problem is not that particular ESC and not that particular motor, I would need a motor here to investigate further.

[funth1ngs]

Hi!

I tested different values for ADVANCE and MIN_PERIOD without great success. It's better with ADVANCE at 35-30°, but the stalls are still there.

Everything works perfect with a RCTimer A2830/14 (3S/1045, no sync porblems at all), but a HK NTM 28-30 (3S/1245) sometimes stalls without disarming (not as often as the MT3506).

With telemetry enabled I noticed that BAD_DETECTS incresed to ~600 during a stall. BAD_DETECTS_DISARM is set to 48. Is there a time span in which BAD_DETECTS_DISARM must be reached?

Simon has a MT3506 and I think he's trying to fix the lost sync problem for his tgy firmware.

If you would like to investigate I can send you a MT3506.

Cheers,
Heiko

[bn999]

The BAD_DETECTS must be sequential in order to trip a disarm. If you had enough good detects in between the bad ones, it could think it has recovered. Although this should be sorted out, the more important thing is to figure out why sync is being lost in the first place. If you want to lend me a motor, I will try to investigate further. It might be good to first determine if anyone else can recreate this problem as it might just be a bad ESC.

[JussiH]

Bill, I will send you a MT3506 NIB tomorrow. I will try and test CL calibration tonight with this motor and see if I can replicate the problem.

Funthings, can you try another ESC?

[kinderkram]

He only has 1
But he can try another one if someone sends him replacement.

[Flyduino_Paul]

Can only comment that basically there is no problem with ESC32 and MT3506, didnt tried the CL callibration till now but a quick setup on the testbench showed a good result,i guess the failure is in the soft, Peter tested only with small Motors so far.
I am also wondering a bit for the SimonK, we have a copter with 3506 and SimonK TGY ESC, works, how did this combination behaved for you,would be nice if you can describe the failure a bit more, mean its was only a quick build, we didnt tuned it till perfection, but it flew.
I cant say there where any timing problems, those can be heard easily as far i know.
If Peter or Menno wants i can send a 3506 to them, but dont know if they have time for it now.
Motor bigger then 28XX are harder to drive we are aware of this, as the Hobbywing Fw doesnt work with Motors above 28XX, at least the versions we know.