Piezo control

[jsheppard95]

PR to begin discussion/obtain feedback on Pitch Control function block necessary for stepper/piezo combination

Summary of Changes

• Function block is a modification of: https://github.com/jsheppard95/HOMS_FEE/blob/master/HOMS_FEE/HOMS_FEE_PLC/Common/POUs/HOMS_FBs/FB_PitchControl.TcPOU that now uses DUT_MotionStage, FB_MotionRequest, and FB_MotionStage

• Currently focusing on basic pitch control state machine, i.e waiting for position requests, properly executing stepper motion to requested position, and then transferring to piezo control function block : https://github.com/jsheppard95/lcls-twincat-optics/blob/piezo-control/lcls-twincat-optics/lcls_twincat_optics_plc/POUs/FB_PiezoControl.TcPOU

• Current logic simply waits for changes to stMotionStage.fPosition and traverses the state machine.

Current Usage

• Test project defined in Main, this would eventually go into https://github.com/jsheppard95/MR1L0-MR2L0/tree/piezo-control which would instantiate the structure HOMS_PitchMechanism and function block FB_PitchControl along with run the serial communication tasks.

• Currently using simulated axis to test the coarse motion and transfer to PCM_FineMove but beyond this it is difficult to test without connecting to real hardware.

Current Issues

• Most obvious is that when a limit switch is hit, motion stops as expected but cannot move off limit. This is due to https://github.com/jsheppard95/lcls-twincat-optics/blob/piezo-control/lcls-twincat-optics/lcls_twincat_optics_plc/POUs/FB_PitchControl.TcPOU#L86:

IF (NOT Pitch.Stepper.bAllForwardEnable OR NOT Pitch.Stepper.bAllBackwardEnable) AND (Pitch.Stepper.Axis.Status.MotionState = MC_AXISSTATE_STANDSTILL) THEN
    // Not a good way of doing things - can't move off limit switch as setpoint will stick at limit switch position
    Pitch.Stepper.fPosition := lrActPos;
    lrCurrentSetpoint := Pitch.Stepper.fPosition;
    lrLastSetpoint := lrCurrentSetpoint;
END_IF

• Something like this is necessary to reset the setpoint to the current position when a limit switch is hit, but as written further changes to the setpoint will revert to limit switch position until stage gets FeedFw/Bw enable.

TODO

• It now seems necessary to connect to the physical serial controller to get a sense of the necessary changes, but it may also be worthwhile to go through a preliminary review first

[klauer]

(Closing/re-opening to kick travis + get a code summary)

[klauer]

Looking back, the linter error was caught by CI, but it didn't fail the build: https://travis-ci.com/pcdshub/lcls-twincat-optics/jobs/287061871 (known issue https://github.com/pcdshub/pcds-ci-helpers/issues/7). I'll get on that to try and save this from happening again...

[jsheppard95]

Anyone against merging this? I haven't noticed any major issues from working with the real hardware.

[klauer]

Most important thing is that it works on your end. I haven't managed to find time for a full review, sadly, but am very curious to see your results with the piezo.