David's Astronomy - Notes 2022-01-17 (S955)

Issue: For some Guide Runs the time gap between the final focusing movement and the start of the first image frame is still unneccessarily long. For some guide runs (eg Guide Run 8, 16, ) there is still a gap of around 30s.

Review: Reviewing timings for CN Ori (Target 9/48, Guide Run 3, Focus Profile 4)

20:11:38 ScopeMode.Guiding 20:11:39 Start Guide 20:11:40 Set Dec Guide Mode / PreJog 20:11:57 Finish/Save Cam2 Reference Frame (10s) 20:12:22 Finish Linking Cam2 Reference Frame (9.2s + 9.5s + 0.9s) 20:12:23 ConvertFitsToJpg (Cam2 Ref) 20:12:25 Send GuideRequest (timeout 15s) 20:12:54 Start PHD2Guiding Monitor (28s) 20:12:54 ConvertFitsToJpg (FocusFrame GuideRef) 20:12:55 Sleeping for 3.1s to throttle BatchFile executions. 20:12:54 ConvertFitsToJpg (FocusFrame Guide03_F1) 20:13:00 Start FocusMainScope (Foc1) 20:13:02 Take Focus Locate Frame (3s) 20:13:10 Take Focus Frame (3s) 20:13:15 Start Focus Profile 20:13:25 Finish Focus Move to Start of Focus Range (10s) 20:15:14 Finish Profile Steps 20:15:23 Finish Focus Move to Best Focus Position (9s) 20:15:23 Take 3 focus frames at best focus (11s) 20:15:39 Finish 12s delay (delay after making a large (>250 steps) focus change . This delay was 20s at one stage) 20:15:51 Set ScopeMode Imaging. 20:15:58 Begin TakeMainScopeFrame 20:16:06 Finish Focus Move to Best Focus, V Filter (8s) 20:16:16 Finish 10s delay (delay after making a large (>250 steps) focus change) 20:16:16 Adjust Lock Position 20:16:16 Starting 20s delay after resetting lock position (ReleaseFromWaitTime = Now().AddSeconds(20) )20:16:32 Adjust Lock Position 20:16:43 Check Focus Position (0s, but could be 4-10s) 20:16:43 Take First Frame

Example Charts

Begin Guiding Chart for CN Ori (Guide Run 3)

Operations (mins) CN Ori U And RR Tau

Planned Duration 12.70 14.10 11.10

Waiting 0.00 0.00 0.00

Slewing 0.56 0.50 0.40

Centering 0.95 1.01 0.80

Focusing 2.65 2.50 2.53

Imaging 7.32 8.95 6.64

SubTotal 11.48 12.96 10.37

unallocated 1.96 ? 1.55 ? 1.55 ?

Total 13.44 14.51 11.92

Conclusion
Several opportunities to reduce time.

1) 10-25s of time could be saved by carrying out Cam2 Reference Frame Locate and Start Guiding operations in parallel.
- the issue blocking this is the management of the two Report Messages streams which are designed to be serial tasks at present.
- prehaps the Start Guiding messages could be temporaily 'saved' and then reported in one go at the end of Locate Cam2 Frame operations.

2) Time can be saved by eliminating 3 frame focus check at end of a focus profile>

3) Can the 12s delays be reduced or eliminated ?

Actions

Check Focus (3 frames). Remove 3 frame Focus Check at end of Profile in cases where Focus Mode is auto And target isn't a Near Zenith Focus Field And target isn't well ahead of schedule. How to determine whether there is sufficient time for the Check Focus or not ?
Focus Profile is performed if Session.DeltaMinsFromPlan < -3.5

Deferred Focus. Reinstate deferral of the end of profile move to best focus when first filter is not C and Check Focus isn't being performed.

Delay after Focus change. Reduce delay after large focus change from 12s to a standard 10s.

Parallel Activity. Consider possibility of making Locate Cam2 Reference and Request Guiding a parallal activity.
On Hold

Update 2022-01-21
Session S956 (2021-01-20) was run using new default timings: Slew(80s, unchanged), Locate&Centre (120s), BeginGuiding (90s), Focusing (150s), BeginImaging (60s).

Wait time at the start of many targets was frequently around 2 minutes, but sometimes up to 4 minutes. Total wait time may have been as high as 43mins equivalent to 2 or 3 whole targets. These waits are partly due to 1) overall impact of changes to default times (the explicit inclusion of 150s for focusing and 60s to Begin Imaging, and extra extra time to Begin Guiding with reduction in time allowed for centering ) and 2) some targets ending 2 mins earlier than planned in cases where focusing is cancelled early due to focus star being determined as 'too dim' or 'lost'.

The average task times over the S956 session were : slewing 33s (vs 80s plan), centering 77s (vs 120s plan), begin guiding 56s (vs 90s plan), aborted focusing 30s (vs 150s plan), completed focusing 163s (vs 150s plan), begin imaging 19s (vs 60s plan).

Looking at longest task instances : longest slew was 81s (vs 80s plan), longest centering was 180s (vs 120s plan), longest begin guiding 65s (vs. 90s plan), longest focusing 184s (vs 150s plan), longest begin imaging 54s (vs 60s plan).

On the other hand U And had an actual duration of 15.5 mins (excl. waiting) vs a scheduled duration of only 12.6 mins. The overrun is due to the frequent focus changes associated with each filter change ( 3 x BVRI, so a total of 12 filter/focus changes).   When focus change is large one (including taking out focuser backlash) an extra 10s is added to allow guiding to resettle & be checked.   Current timings assume only 2s per filter change.

The order that filters are changed affects the timings significantly   Decreasing filter step position ( 2 large focus change + 10s settle) takes 17s but increasing focus step position (a single small focus change) takes just 3s. A BVRI set carries an overhead of 54s (3 x 17s + 1 x 3s), but optimising the filter order would carry an overhead of only 26s (3 x 3s + 1 * 17s) a saving of 28s (or 84s for a 3 x BVRI sequence).       A fully optimised sequence (say III-RRR-VVV-BBB) might carry an overhead of only 9s vs a 162s overhead for BVRI-BVRI-BVRI sequence.

For the timing being filter operations/order will be left unchanged as most targets involved only C Filter imaging, however extra time clearly need to be incorporated for making the filter changes that are currently programmed.

Proposal new timing allowances for planning (2022-01-21 onwards) are :

Task New Time (s) Previous Time (s)

Slewing 70s 80s

Centering 100s 120s

BeginGuiding 60s 90s

Focusing 150s unchanged

BeginImaging 30s 60s

Filter Change 10s 2s

Caveats : Things are not allowed for in calculated expected target durations are:

- impact of dome slewing slippages/holdups that may delay target slewing
(plan assumes 70s average time of slewing)

- impact of cloud that will number of centering attempts and centering exposures that may delay target centering
(plan assumes a 100s average time for centering)

- guiding issues that may require extra adjustments between frames
(plus assumes 'perfect' guiding with no extra delays between frames)

- carry over of wide search focusing on to 1st/2nd targets if focusing is not achieved for near zenith focus field
(plan assumes 150s average time for focusing which is the time for a single focus profile only)


	Session Aims & Highlights
	- Observing Result - Night Summary Plot - Session Event Log

	Operational Issues
	- Critical Issues (0), Major Issues (1), Minor Issues (1), Small Defects (3), Continuous Improvement (8)

	Images from 2022-01-17 >> [ Local Files >> ]

	Review - Optimising Guiding & Focusing Times

This Web Page:	Notes - Session 955 (2022-01-17)
Last Updated :	2024-09-25
Site Owner :	David Richards
Home Page :	David's Astronomy Web Site

David's Astronomy Pages
Notes - Session 955 (2022-01-17)

Session Aims & Highlights (2022-01-17)

Operational Issues (2022-01-17, S955)

Review - Optimising Guiding & Focusing Times

Operations (mins)	CN Ori	U And	RR Tau
Planned Duration	12.70	14.10	11.10
Waiting	0.00	0.00	0.00
Slewing	0.56	0.50	0.40
Centering	0.95	1.01	0.80
Focusing	2.65	2.50	2.53
Imaging	7.32	8.95	6.64
SubTotal	11.48	12.96	10.37
unallocated	1.96 ?	1.55 ?	1.55 ?
Total	13.44	14.51	11.92

Task	New Time (s)	Previous Time (s)
Slewing	70s	80s
Centering	100s	120s
BeginGuiding	60s	90s
Focusing	150s	unchanged
BeginImaging	30s	60s
Filter Change	10s	2s

David's Astronomy Pages Notes - Session 955 (2022-01-17)

Session Aims & Highlights (2022-01-17)

Operational Issues (2022-01-17, S955)

Review - Optimising Guiding & Focusing Times

David's Astronomy Pages
Notes - Session 955 (2022-01-17)