Remove power limits after initial run phase is done

[damosvil]

This pr moves the release of startup power limits until the end of the initial run phase:

[damosvil]

Tested in the workbench jamming the motor.
In previous commit if throttle is raised during startup the last kick was done at throttle value, not limited.
In this last commit if the motor is jammed max startup power is respected unless the motor starts spinning. In that case applies power rampup implemented in set_pwm_limit_function.

Please, when possible could you check on your side?

[damosvil]

Tested in workbench. Power supply over current protection never triggers.
Tested also crashing with a drone (motor 1 shaft was damaged so it didn't spin well):
https://youtu.be/HBaYtLEnpSM
https://youtu.be/6kz2UXlnqHg
https://youtu.be/VXC0j5dCzlE
https://youtu.be/WyG72G9rX_g

No magic smoke, in any motor neither in any esc.

[stylesuxx]

Alright, painful to watch, but looking good in general. I did some testing with whoops, although not with the latest push. Did you do any turtle mode tests with 5" by any chance? Also was this with dyn idle on?

[damosvil]

Tbh no dyn idle testing yet and no turtling. I will make more testing this week

[damosvil]

More testing done. This time turtling:
https://youtu.be/WhxjvZtFrDQ
https://youtu.be/StgaBe4b51Q
https://youtu.be/8_GBsrCebOw
https://youtu.be/euY0VAXHNz4
https://youtu.be/07bvxammWyY

Dynamic idle also tested. Min/max startup power are 1010 and 1020 respectively. They are the bare minimum to start the motors. Above that the likelyhood of burning anything increases.
I realized that when dynamic idle is active the motors fight to keep the rpm when crashing and not in a proportional way, and the likelihood of burning the motors/esc increases.

[stylesuxx]

Perfect, I added a note to the release notes for people to be careful when setting up min/max startup power.

I am wondering if we should adjust the min/max startup power defaults to a "save minimum". If not, we should probably also do a test run with the defaults, just to be on the save side.

[damosvil]

In my opinion defaults are very high for 5 inch kwads (normally 1010/1020), even for 3 inch they may be too much on some ocasions. Unfortunatelly they should be adjusted to the absolute minimum for each type of kwad, depending on battery S, motor type and esc deadtime. Some sort of preset system or calculator may help, but I am not sure how far of scope may it be.
In addition some pilots may want to intentionally keep min/max startup power high, because they would want more power to turtle their kwads at the expense of burning them

[stylesuxx]

Yeah, we could add a calculation to the wiki, I think that'd be out of scope for this issue. What I'd feel comfortable with is reducing min/max defaults, keeping 5" in mind. Worst case we would have people complaining about motors not spinning up or turtle mode not working properly - those requests would be easier to handle than people blowing up their shit because they did not read the release notes.

[stylesuxx]

OK, looks fine to me. We should still consider if we want to adjust the min/max startup power defaults - I would also need to adjust that in the confiurator then.

[damosvil]

I think that we could set it to 1010/1020 for 5 inch kwads. I know that this is the most conservative option, but many people already know that they have to move sliders up to make it work on tinywhoops.

[stylesuxx]

Yes, better to err on the side of caution. Let's adjust min/max defaults then.

[damosvil]

Assets for 0.20.1 RC1 (testing)
v0.20.1_RC1.hex.zip

Remember to keep min/max startup power to the minimal possible value for maximum protection on crashes and set motor idle to keep motors spinning after startup.

[nerdCopter]

O_H_5_48_g93e662f.hex compiled by @hypOdermik Aug 31
1080 startup
1120 max
(pretty much lowest i could set for consistent startups / turtle)

BETAFPVF4SX1280
0802SE 19500kv, 1S, 9% motor idle, no rpm filter, no dynamic idle.

0.21.1 startup testing https://youtu.be/YXQBVlivhEk (thumbnail is settings)
0.19.2 1S flight https://youtu.be/Afy_1ygsy8Y
0.21.1 1S flight https://youtu.be/oaZx7F-ed3Q

[damosvil]

BETAFPVF4SX1280 deadtime 5, 0702 28500kv motors & Bluejay 0.20.1
All motors start at 24khz, 48khz, 96khz: idle = 8% min/max startup power = 1040/1060

TMOTOR F411 AIO 1S 6A ELRS, deadtime 10, 0702 32000kv motors and Bluejay 0.20.1
at 24khz: idle 8%, min/max startup power = 1050/1070

[hyp0dermik-code]

GEPRC-AIO-F722 G-H-30 1804 2450kV @ 6S Bluejay 0.20.1 from assets above
All motors consistently start at 1010/1020 min max. 8% static idle

[damosvil]

Assets for 0.20.1 RC2 with power rampup during initial run phase (testing)

0.20.1-rc2.zip

[stylesuxx]

Alright, looks fine to me.

I suggest we merge and move on to make version "v0.21.1 RC2 (for testing)" available via configurator.

Maybe it would make sense to move the test results regarding min/max startup power and idle to the Startup power wiki page as a reference?

[damosvil]

Yes, please merge it.
I will update the wiki