Improved tuning sliders calculation #10509
Conversation
src/main/config/simplified_tuning.c
Outdated
| @@ -34,59 +34,25 @@ static void calculateNewPidValues(pidProfile_t *pidProfile) | |||
| { | |||
| const pidf_t pidDefaults[FLIGHT_DYNAMICS_INDEX_COUNT] = { | |||
| [PID_ROLL] = PID_ROLL_DEFAULT, | |||
| [PID_PITCH] = PID_ROLL_DEFAULT, // using the same defaults as roll | |||
| [PID_PITCH] = PID_PITCH_DEFAULT, // using the same defaults as roll | |||
There was a problem hiding this comment.
Switching from Roll to Pitch Default here. How about changing the comment too. For reference and understanding what is the difference in this change?
There was a problem hiding this comment.
Ops, forgot to remove the comment, in the old code calculated roll and pitch pids were the same(based on the same roll defaults)
There was a problem hiding this comment.
Also this seems to include the commit / changes from 10508 but I guess it's to test this PR until squashed / rebased?
Any idea about the disabled filter sliders (cannot reset / enable) in configurator? (Not aware about ongoing work / discussion - already looked at solving but couldn't find any other than there is a new simple* setting not involved in filtering but pid settings only - have to check yet the dterm / gyro filters)
There was a problem hiding this comment.
Yes, I've done this upon the other branch
|
@IllusionFpv: Let me know when this has been tested enough. |
|
Will this solve the issue of the Configurator calced PIDs and the FW calced PIDs not aligning? Seems like the Configurator calced PIDs are a little off and when it saves and refreshed the PID grid from MSP, it updates to be what the FW calced???
|
Yes, in current code pids misaligned after some slider iterations |
|
@mikeller I fixed applying tuning, for me is good to go now |
roll pitch ratio now increase pitch constrain dmin ratio, disable dmin if multiplier is 200 fix apply simplified tuning
IllusionFpv
force-pushed
the
fix/tuning-sliders
branch
from
10e4d0b
to
3f3c143
Compare
|
@IllusionFpv please provide test firmware using |
|
@IllusionFpv , can you set Testing this build on ["Reorder fw sliders" #2402], it looks like all is in order. The only thing that is off yet is it seems like the PID Grid is always one step behind?
Also, fast movement doesn't register correctly. Seems like in the configurator, we need to create another call/update (call Refresh) of the PID grid in the firmware so the PID forms isn't always one step behind with fast movement or small tweaks. Let me look at that. |
Not sure that aligning the simplified settings with the legacy configurator sliders should be the goal here - if the new settings and their defaults are better then we should go for them even if they do not match the legacy configurator sliders. |
Refactored
calculateNewPidValuesfunction, now save pids once, need testing