Add HWP operations to scheduler

[kmharrington]

The HWPs work, we need the scheduler to be able to turn them on and off at the appropriate times. (do we know the appropriate times yet?)

[jlashner]

Just want to comment that even before HWP operations are added to sorunlib, it would be very useful right now to have the ability to add a wait buffer for HWP spinup in between detector setup and scans.

The method I've been taught is to delay the scan start time by ~20 min to account for spin-up time, however I think without building the schedule with this buffer built in, I think this might mess up the scan pointing.

Below is a snippet of a modified schedule with hwp spin-up to show how things are currently happening...

#################### Detector Setup Over ####################

HWPPrep()
forward = True
hwp_freq = 2.0
HWPSpinUp()

################# Scan #################################

now = datetime.datetime.now(tz=UTC)
scan_start = datetime.datetime(2024, 1, 26, 22, 27, 33, tzinfo=datetime.timezone.utc)
scan_start += datetime.timedelta(minutes=20) # Account for HWP Spinup
scan_stop = datetime.datetime(2024, 1, 27, 1, 37, 30, 429667, tzinfo=datetime.timezone.utc)
if now > scan_start:
    # adjust scan parameters
    az = 345.931 + -0.00513*(now-scan_start).total_seconds()
else: 
    az = 345.931
if now > scan_stop:
    # too late, don't scan
    pass
else:
    run.acu.move_to(az, 50)

    print('Waiting until 2024-01-26 22:27:33+00:00 to start scan')
    run.wait_until('2024-01-26T22:27:33+00:00')

    run.seq.scan(
        description='jupiter', 
        stop_time='2024-01-27T01:37:30.429667+00:00', 
        width=20.833, 
        az_drift=-0.00513, 
        subtype='cal',
        tag='jupiter,setting,middle_focal_plane',
    )
    print('Taking Bias Steps')
    run.smurf.bias_step(concurrent=True)
################# Scan Over #############################

Edit: Reading this over I'm realizing that scan_start is not even used in the scan, so I think this might not be doing anything

[kmharrington]

yeah that isn't doing what you want it to? Have we been running scans like that? worth making notes we're probably missing sources

[kmharrington]

I'm starting to implement these on the satp1-updates branch

[jlashner]

Yeah I'm not sure if anyones looked at the data from these scans, but looking at the code, changing scan_start wont actually delay the start time of the scan, and is just used for determining az-shift offset.... so I feel like we're probably still catching HWP spin up and also have messed up pointing.