XBPM feedback

[DominicOram]

• Set transmission to 100%
• Check that feedback is in position (with some settle time)
• Stop feedback
• Move to transimission for data collection
• Do data collection
• Start feedback
• Set back to 100% transmission

[DominicOram]

This is what's used to tell if we have beam so we can just wait on it.

For testing when we have no beam it might be useful to turn this on/off but don't want this accidentally

[DominicOram]

The checking with settle time relies on bluesky/ophyd#1142

[DominicOram]

The code used in GDA currently to do the waiting part is:

def waitWhileBusy():
	logger.debug("waitWhileBusy()")

	if is_unavailable():
		logger.debug("waitWhileBusy() skipped as feedback unavailable")
		return

	if is_enabled() and not is_paused():
		if isI03():
			if expt_shutter() == 'Closed':
				return
			# beam already in position
			if not beam_position_feedback.isBusy():
				return
			# wait for beam to be in position
			beam_position_feedback.waitWhileBusy()
			beam_in_position = False
			time_to_wait_for_beam = 3 # beam must be in position for this many consecutive seconds
			logger.info("Waiting for beam in position for "+str(time_to_wait_for_beam)+" seconds")
			while not beam_in_position:
				for n in range(time_to_wait_for_beam):
					if not beam_position_feedback.isBusy():
						sleep(1)
						beam_in_position = True
					else:
						sleep(1)
						beam_in_position = False
						break
		if isI04():
			beam_position_feedback.waitWhileBusy()

[DominicOram]

As discussed with @neil-i03, the reason we want to run with the XBPM feedbak on and transmission at 100% for most of the time and turn it off for collections are:

1. Objects crossing the beam that illuminates the feedback xbpm (e.g. attenuators) risk a feedback event while crossing - this leads to the feedback motor heading to an extreme and it may not recover ( seen many times)
2. At higher energies the signal reduces. So low transmission/high energy risks hitting a point where feedback on the xbpms is unreliable. There is a gain change feature, but it can cause the XBPM readback to hang and then it just sits there forever.

XBPM feedback is mainly for correcting slow thermal drift so safe to assume beam is stable when feedback off for collection