David's Astronomy - Notes 2018-09-04 (S00626)

David's Astronomy Pages
Notes - Session 626 (2018-09-04)

Notes
(S625)

Notes
Main

Home
Page

Notes
(S627)


	Session Aims & Highlights
	- Observing Result - Night Summary Plot

	Operational Issues
	- Critical Issues (0), Major Issues (2), Minor Issues (6), Continuous Improvement (1)

	Fully Automated Operation
	Critical Focus Zone
	Autofocussing
	Sky Flats Routine

	Images from 2018-09-04 >>

Session Aims & Highlights (2018-09-03)

Main aims

Targets. Acquire images for a selection of variable stars, nearby stars and deep sky targets
Software. Test fixes & modifications made to the Observatory Control Program (CCDApp2) since the last session
Stability. Continue checking stability of the Observatory System

Equipment & Software

12" LX200 and ST-10XME
80mm APO and ZWO ASI 178MC
CCDApp2, TheSky6, CCDSoft5, PHD2, POTH Hub

Highlights

Targets. Images were successfully acquired from 13 targets (U Her failed due to airmass cutoff (3.97) exceeding cutoff)
Stability. Observatory operations and system were somewhat unstable during the first part of the session due to several instances where the Observatory Control Program locking up during manipulation and movement of overlying windows etc

Summary Plots & Logs

Observing Result (2018-09-04, S626)
Observatory Unlocked at 20:50, Operations started at 21:02,
Automated Mode from 21:05, Job Queue started at 21:26,
Software interruptions at 21:36, 22:10, 22:17, 22:30, 22:35, 22:50
Session Suspended 22:08-22:12, 23:24-23:37, 00:08-00:20, 01:18-01:22,
01:51-03:06 due to Cloud. Session Closed at 04:42,
Session crashed before Session Housekeeping & FInishing

(Observation Status : Green=Completed, Yellow= Partially Completed, Red= Failed)

Night Sky Summary Plot - 2018-09-04
Top axis: Sky Brightness at Zenith (in ADU/s)
Lefthand axis: Local Time (hh LT). Righthand axis: Sun Altitude (degs)

[ Prev | Next ]

Critical Issues

None

Major Issues

CCDApp2 locking-up/freezing. Observatory Control Program completely locked-up/hung on several occasions during the first part of the session. This seemed to occur when other windows passed across it main window and /or other window manipulation were being made. The Program was being run in debugging mode from VS2017 and the only solution was to stop debugging and restart debugging again. This problem occurred in several other recent sessions (e.g. 2018-08-30), but not the last (2018-09-03) .Possible association with poor LAN performance which was affecting computers on the LAN at the same time. Unknown reason but is resumed to be related to thread operations trying to manipulate components on the CCD2App2 window whilst it was busy. Issue Largely Resolved (Program is now run outside of VS Studio/VS debugger which has improved stability)
Closed
Autofocusing Failed (star lost due to cloud). An attempted autofocusing operation after slewing to CY Lyr, was manually aborted when it was seen that the focus star become lost / covered by cloud after around 6 step position out of 20. This was done so as to not waste unnecessary time, and because of uncertainty over how the routine will handle curve fitting with missing points (at worst the routine could have a crash and bring down the whole program). Code needs to be modified such that the autofocussing routine checks that data is going to be sufficient to define a curve and abort if not. Issue Partly Resolved (Code modified, but produced an unwanted side effect. Code modified again in Jan 2019, and performance with be checked over next few sessions)
Closed

Minor issues

LAN Performance very poor in first part of session . Unable to open latest allsky picture on AllSky Computer (Laptop7) for example. Difficult to use UltraVNC to control Observatory Computer (Laptop9) from IPad or Desktop Computer (HP_Main). Problem was seen before
Resolved or no deemed an issue.
PHD2 Guiding reporting Slew Movements when none are occurring. With the "Abort Guiding if Slew Starts" option active, it was noted that PHD2 showed one case when it stopped guiding and put up a message that it was stopping because the slew was / had taken place, even though no slew was actually taking place. Seems to be a one-off issue.
Closed

Query over target overrun logic. Due to a session suspension due to cloud, target 61 Cyg was running so far behind schedule that it cancelled all or nearly all of its schedule set of 7 x 2s exposure logic to notionally allow the next target to begin on time. However the next target had already passed in schedule start time, so that the queue moved onto the target after that (BL Lac) . This targets still had 11.5 minutes scheduled to wait, though in practice the wait would have been 5 minutes less that this, i.e. 6.5 minutes). During this period of wait, 61 Cyg could easily have completed it's set of images in under 1 minute. Issue Largely Resolved (Targets stopped if they will exceed there end date, unless they are a ToO target, these continue execution)
Closed
SS Cyg duration given as 12.1 minutes in database. How well does this equate to actual duration taken. The target had 5 sub schedueled for each of the 9 images scheduled for acquistion. Issue Largely Resolved.
Resolved (AstroPlan / AstroMain)
Observing Database Tool (AstroPlan/SQLForm2) was showing just the first target in each plan, and showing no target observations at all in session at all (reason is that the TheSky6 that was connected to when Sqlform2 was opened, has since been shut and reopened. The reopened copy of Thesky6 can not be seen by the program.)
Fixed (AstroPlan)
Observatory Control Program Locked Up or Otherwise Failed after closing down the observatory.
Whilst taking images of final target in the scheduled observing plan (Hickson 30) observatory manager intervened at 04:42:05 to start closing the dome shutter and suspend the job queue due to cloud conditions. (this was just before the job executor was going to take its 3rd 180s image (out of 5) and with sun alt at -12.5 deg. Almost immediately at 04:42:08 the Observatory Manager began Closing Session ; Parking Dome (04:43:14). Parking Scope (04:43:24) and disconnecting the Camera (04:43:24). At 04:44:08, Job Queue issued an "Error during TakeImage" error to logfile and a "Failed()" message to the Report File. This is presumed to be residual job queue activity associated with the 3rd image from Hickson 30. Certain program threads (such as Monitor) continued running after the main program thread stopped (Why was Session Closed when it did, and why is Job Queue not allowed to be suspended or aborted more gracefully. Whilst cloud conditions deteriorated quickly around the time the Dome was closed, the 2nd image of the target was unaffected by cloud, and the Observatory Manager could probably have let the 3rd image finish before suspending and then closing the session).
Fixed (AstroMain)

Continuous Improvement

Can targets be finished before suspending unless the condition are absolutely critical. Unclear if an issue still remains
Closed (AstroMain)

[ Prev | Next ]

Critical Focus Zone

The traditional Critical Focus Zone is the range of travel on the focuser draw tube within which the defocus of a point source of light is smaller than the first dark diffraction ring of the Airy disk. The traditional Critical Focus Zone (denoted CFZ) is given by the formula

    CFZ = 4.88 * Lambda * f ²

where
    CFZ - Traditional Critical Focus Zone (micrometers)
    4.88 - constant (unitless)
    Lambda - wavelength of light (micrometers)
    f - effective imaging system f/ratio (unitless)

For my 12” LX200GPS, operating at f/10.4, and using wavelength of 0.55 micrometers (550 nm, green) the critical focus zone is calculated to be 290 micrometers which is equivalent to 134 steps for my TCF-S Focuser (0.000085” per step), ie a zone that is +/-67 steps wide.

The basic foundation of the traditional Critical Focus Zone (CFZ) is that when at perfect focus, very small changes in focus do not degrade image quality in any significant way.

However research and testing by GoldAstro has shown that the assertion that certain focus errors are non-measurable must be revised. The "New Critical Focus Zone" (denoted NCFZ) recognizes that all focus error is measurable. Instead of non-measurable focus errors as in Critical Focus Zone, the New Critical Focus Zone uses a criteria of whether focus has a negligible impact on image quality. NCFZ uses a "tolerance" for focus error designated by the astroimager that is appropriate to his/her needs. The New Critical Focus Zone is given by the formula

        NCFZ = 0.00225 * Theta_FWHM * sqrt (Tau) * A * f²

where
    NCFZ - New Critical Focus Zone (micrometers)
    0.00225 - constant (micrometers per arc second per millimeter)
    Theta_FWHM - total seeing (arc seconds)
    Tau - focus tolerance as a percentage of total seeing (unitless)
    A - telescope aperture (millimeters)
    f - effective imaging system f/ratio (unitless)

[ see New Critical Focus Zone at http://www.goldastro.com/goldfocus/ncfz.php ]

For my 12” LX200GPS, operating at f/10.4, and with 1.5” FWHM total seeing and a focus tolerance of 10% the critical focus zone is calculated to be 352 micrometers (0.35mm) which is equivalent to 163 steps for my TCF-S Focuser (0.000085” per step), i.e. a zone that is +/-81 focuser steps wide.

This CFZ changes to +/- 54 steps for 1" seeing, +/-109 steps for 2" seeing, +/-136 steps for 2.5" seeing & +/- 163 steps for 3" seeing.

Unattended Autofocussing

Reliable autofocussing is key component of successful unattended operation of an observatory. Whilst my currently written autofocusing works well when my eyes (and brain) are in attendance to either acknowledge the computer's best focus position based on the data collected or quickly intervene to correct the situation if an invalid/inappropriate focus position is selected or to cancel the focusing operation if cloud develops and the star is lost.

Entry routine for performing autofocus is "PerformAutoFocusRun"
RunAutoPickFocusStar takes a full frame image and automatically select a star   )
objConsole.ShowFocusGraph shows the Focus Graph
objConsole.MeasureFocusProfile measures the focus profile and
   objConsole.DrawGraphTab draws the Focus Graph

With a good dataset the existing Focus Algorithm works well (e.g.. 1st Focus Profile below), but examples of where it doesn't work are shown in 2nd & 3rd profiles. In these cases only 1 or 2 'bad' data points are enough to completely invalidate the calculation of Best Focus. A revised algorithm has been developed (2018-09-06) which checks the result for consistency with 3 other measures of Best Focus position (based on 'peak' FWHM position, and the centre position & P50 position of the points lying within 1.5 arc sec of lowest FWHM position).

1) Example of 'Good' Focus Profile
Curve Fitting has worked well, 2018-09-01 )

2) Examples of 'Bad' Focus Profiles
Curve Fitting has been thrown significantly off by two 'bad' points (2018-08-30)

Curve Fitting has been thrown off by a single 'bad' point (2018-09-04)

Focus Profile using new Focusing Algorithm
Using same data as above profile, the new algorithm (2018-09-06) recognises
the discrepancy between the Curve Fitted Result and an Alternate Result
that uses the just the highest FWHM values and elects to use
the Alternate Value at Best Focus Position in this case

Focus Profile using new Focusing Algorithm
(annotated with selected focus star images and CRZ)

Sky Flats Routine

Working on a new routine for taking Sky Flats

- Button 'Take Dusk Flats' (Take Dawn Flats will be added at a later date)
- File defining the filter/bin sets to be acquired by sky flats routine is hard coded as AutoSkyFlats.txt (this will be changeable at a later date) . The order of rows in the file defines the order in which flats are to be acquired if taken at dusk. (The order will be automatically reversed for Dawn Flats). Filters for wavelengths where the ST-10 CCD is less sensitive (eg. B) or narrow band filters should be given first, C filter should be given last.

- No. of frames for each filter/bin set is hard-coded as 15 (this will be changeable at a later date)

- Routine objconsole.RunTakeDuskFlats

- This calls the routine TakeAutoSkyFlatsNew(1, 0) ' where 1 (Dusk), 0 (Not Auto)
- After reading the AutoSkyFlats.txt a Loop calls TakeAutoSkyFlatsSetNew routine for each filter/bin set

CalcFlatSpot calculates Flat Spot Az/Alt based on current Sun Location
SlewToFlatSpot slews the scope to current Flat Spot