David's Astronomy Pages Notes - Session 638 (2018-11-30)

Notes (S637)

Notes Main

Home Page

Notes (S639)

	Session Aims & Highlights
	- Observing Result  - Night Summary Plot
	Operational Issues
	- Critical Issues (0),  Major Issues (2),  Minor Issues (5),  Continuous Improvement (1)
	Images from 2018-11-30  >>
	2018-12-03
	Unexpected Computer Restarts
	Disconnecting Telescope from TheSky6

Session Aims & Highlights (2018-11-30)

Main aims

1. Software. Test operation of Observatory Control Program (recently renamed as AstroMain)  following several code changes since last session.
2. Focusing. Test corrected routines for autofocusing the LX200 main scope with TFC-S focuser/SBIG Camera and autofocusing the TS 80mm secondary scope with Pegasus Universal Stepper Motor/ZWO camera images
3. Targets.Acquire images of a range of targets (comets, variable stars, nearby stars and deep sky objects)
4. Stability.  Check stability of the Observatory System (hardware & software) with regard to  SBIG Camera, TCF-S focuser, TheSky6, LX200 and Pulsar Dome which have each given problems in recent sessions
5. AstroGuard. Test operation of AstroGuard program (a program providing a second level of defence for the observatory)

Equipment & Software

•  12" LX200 and ST-10XME
•  80mm APO and ZWO ASI 178MC
•  CCDApp2, TheSky6, CCDSoft5, PHD2,  POTH Hub
•  Pegasus UPB & Motor Focus, TCF-S

Highlights

• First use of the AstroSuite Software set under their new umbrella.
• First use of AstroPlan to quickly install programs, and later, at end of session,  to downloaded collected data
• First use of AstroLaunch to launch observatory programs at start of session
   (CCDSoft5, TheSky6, DeviceManager, AstroPlan, PHD2, AstroMain, AstroGuard)
• AstroMain (formerly CCDApp2) worked on except for an unhandled  exception that required the program to be restarted.
• Fixes to Automated Focusing using the Foc1 System (TCF-S & SBIG) were successful
• AstroPlan (formerly ObservingTool) was working again and used the newly added GA Convergence Plot
  (the previous session S637 had accidently used a test version in which the AstroDatabase was opened as Read-Only, preventing new plans from being saved, a showstopper! )
• First time AllSky Pictures with Key Scope Movements captured to a separate "_Scope" folder.
• Images.   Images from a number of targets were successfully acquired (hrough cloud ended the session early).

Summary Plots & Logs

Observing Result (2018-11-30, S638) Observatory opened at 21:47 .  Foc1 Focusing (TCF-S) at 22:04 Foc2 Focusing Attempts (TS80) at 22:08 & 22:16 Observatory Manager set to Auto at 22:49. Job Queue at 22:22 AstroMain crashed at 22:36 (ShowInventory commanded on an unsaved File).  Job Queue restarted at 22:49 Session ended at 01:47 (scope/dome parked by manual command)  .

Night Sky Summary Plot - 2018-11-30 Top axis: Sky Brightness at Zenith (in ADU/s) Lefthand axis: Local Time (hh LT). Righthand axis: Sun Altitude (degs) Shutter/Telescope active from ~ 22:00 to ~ 01:30

Back to Top

Operational Issues (2018-11-30, S638)

No significant or generic issues arose during the session, but there were some issues either side of the session that should be noted.

• Critical Issues (0)
• Major Issues (2)
• Minor issues (5)
• Continuous Improvement (1)

[ Prev | Next ]

Critical Issues

• None

Major Issues

• Unexpected Computer Restarts.  For no obvious reason the Observatory Computer has begun to restart by itself
   ( 3 restarts in 48 hour period though none during the Session itself.
  Fixed around 2018-12-11 (see Update on Computer Crash Issue)
   
• COM Port Switching.    Although is didn't cause a problem during the S638 session the Pulsar Dome has continued to show random switching between COM9 and COM4 in the week leading up to the Session with a frequency of around once every 24-36 hours. 
  ( This issue can lead to 'System.InvalidOperationsException. The port is closed" exception and the failure or at least a delay in the session in situations where POTH.Dome connects to an invalid port or the Dome's COM port switches during the course of a session itself, which can stop a session. New code has been introduced from 2018-11-30/2018-12-01 to get around the Port Switching issue, and prevent future operations from being stopped by this problem. Fix still being tested at 2018-12-16. Eventually deemed to have largely resolved the effects of the issue.
  Closed 2019.

Minor issues

• AstroMain Crash (1).   Untrapped exception occured in 'AnalyseStarsInImage' routine relating to the use of Image.ShowInventory on an image that hadn't been first saved to disk.  Issue still  open at 2018-12-16 (despite attempted fixes). Issue deemed closed in 2020.
    Issue resolved/closed 2020
   
• Foc2 Autofocusing not working due to ShowInventory errors whilst analysing Focus Box images (presumably this was due to the box  images not being saved to disk at the time the Show Inventory operations was requested. Review of Log files or code couldn't identify what the problem was. 
  Fixed on 2018-12-05
   
• Unwanted 'Moving Focuser' messages being written to Console & ReportFile.  This is a consequence of recent code modifications involving the introduction of Foc1 / Foc2 handling. 
  Fix attempted on 2018-12-01
   
• AstroAllSky program has stopped saving AllSky Fits Images in the correct Night folders since 2018-11-24. Problem was identifed as being caused by a bug introduced during the removal of Legacy 'CCDApp2'
  Fixed on 2018-12-02
   
• PHD2 Guiding issues when using the ZWO ASCOM Profile. Although Calibration details was manually entered for this profile invalid guide corrections where getting made to the mount. Guiding was correct /ok after returning the profile to 'ZWO ASI 178MC'.
  Fixed on 2018-12-04 through PHD2 Calibration (see details )

Continuous Improvement

• Obs.Viewer window could benefit from some tweaks and additions, including aliging PHD2 & SharpCap Flag, adding Telescope operation/target and adding flag/text when Telescope is Parked.  
  Changes Implemented on ~ 2018-12-01 

[ Prev | Next ]

Back to Top

2018-12-03

Unexpected Computer Restarts

For no clear reason the Observatory Computer has begun to suffer from random crashes and restarts in the last 5 days.
( 3 crashes/restarts in 48 hour period around the time of the last session and another crash/restart happened this morning 08:14.  The issue is a  showstopper for unattended operation, since if it occurs during a session it can leave the dome shutter open and telescope running, with potential damage from rain or cable wrap.

Examing Windows Event Logs show the occurance of Kernal-Power Events (Event ID 41), which say that the 'system has rebooted without cleanly shutting down first. This error could be caused if the system stopped responding, crashed or lost power unexpectedly'.  Loss of power can be ruled out as the laptop reverts to its batteries if AC Power drops out.

Critical Windows Error -  Event ID  41, Kernal-Power

The Event Log history shows Kernal-Power (ID 41) events. Besides the 4 recent events between 2018-11-30 & 2018-12-03, there were 2 earlier events around 2018-11-01 and 4 earlier events around 2018-11-15. 

History of Kernal-Power (ID 41) Events

Kernal-Power (ID 41) events are difficult to diagnose as there is typically little information to go.   Examing the Details tab for the most recent event shows it has a BugcheckCode of 209  (=  0xD1 hexidecimal). This code corresponds to DRIVER_IRQL_NOT_LESS_OR_EQUAL, which indicates that a kernal mode driver attenpted to access pageable memory IRQL that was too high.   
Parameters indicate the problem related to a write operation on IRQL 2.  but that does help identify the specific cause of the crash or a solution.

All that can be done is to monitor the situation for any pattern.

History of Kernal-Power (ID 41) Events

Update 2018-12-04 (S640)

The Observatory Computer, which has begun to crash at seemingly random moments during the past week, crashed
on two occasions during the S640 session. The first one at 18:58  was whilst the observatory was still being attended, but the second at 00:46 occured whilst the observatory was left running unattended, losing 4-5 hours of data and leaving the observatory unable to close itself at dawn (or indeed close itself if rain had unexpectedly occurred).  On this occasion there was no risk due to the confidence in the weather forecast, but the crashes are a serous showstopper for unattended operation of the observatory under normal, very changeable weather conditions.

Log

 

 

What's Changed ?

Windows Update
Windows Update was run  2018-12-01 19:15 to 20:36. This is after the start of the current sequence of computer crashes and specifically can't be associated with the computer crashes at 18:35  on 2018-12-30  and at 02:30 & 18:24 on  2018-12-01.

AstroSuite
AstroSuite programs (AstroLaunch, AstroMain, AstroGuard) have been undergoing regular updates and revisions during this period, but the nature of these programs and the changes being made shouldn't cause the operating system to crash.

The one possible change has been that AstroGuard and AstroMain programs have been running concurrently (as they are designed to do). Within a 12 hour or 24 hour period it is conceivable that both program read the same port at the precisely same time or request a property from POTH.Dome at precisely the same time.  To explicit test this AstroMain was left running (without AstroGurard) and the computer crashed within 12 hours suggesting some other cause.

ASCOM.dome

Back to Top

Disconnecting Telescope from TheSky6

For a while now my code to make disconnect the Telescope from TheSky6 hasn't been working, requiring manual interventation to click the Red Disconnect Scope button in the or and if needing to restart the Poth.Hub. Not disconnecting teh Telescope in TheSky6  can also lead to TheSky dialog boxes appearing when the Telescope is Parked, via POTH.Scope as TheSky6 tries to continue to commuincate with the scope (these require clicking ok in order to dismiss)

TheSky6 - Telescope Disconnect Button

Investigating show that my original code for disconnecting the scope in TheSky6 used two commands RASCOMTele.Disconnect() and RASCOMTheSky.DisconnectTelescope(), but as there were thought to perform the same task the  RASCOMTheSky.DisconnectTelescope command was commented out.  Whilst the description of two commands in TheSky6's help system is similar the help system's remarks suggest that there is a subtle difference between the two commads.

RASCOMTele.Disconnect()
Description : Terminates communication between the telescope object (RASCOMTele) and the TheSky6.
Remarks : Use this command to end communications with the telescope object and TheSky6. Note that this command does not terminate the telescope link in TheSky6. This must be done manually

RASCOMTheSky.DisconnectTelescope()
Description : Terminates communications to the telescope..
Remarks : Use this command to terminate TheSky’s telescope link.

Reinstating calls to RASCOMTheSky.DisconnectTelescope() resolved the problem.

Back to Top