As of mid 2018 I am keeping a log to track general observatory goings on. Entries are in reverse chronological order.
9/16 -20/2018 – Internet down in Torrey. TeamViewer cannot connect. Evidence points to bad lamp timer meaning the router does not get a power reset. I expect that the lamp timer has a little slider switch on its side that sets it to be a timer or to stay on. If it is switched to “on” it would not reset.
9/15/2018 – Back in SLC and internet connection already lost. I don’t understand.
9/14/2018– Drove to Torrey for one night to see what is up with the internet connection. Found a router with internet light on but not connecting. Not clear why the lamp timer power offs and ons did not reset it. I doubled the number of daily resets from two to four.
Results: Everything worked first time. Guiding was tight and excellent for the first six 10 minute frames. Then, while I watched, the guider looked liked it had a few drinks. I let it go thinking maybe a little bad seeing drifted by. But it never got better.
I suppose I need to experiment. It will waste a few nights. Get a night of 10 minute frames without guiding and see if the mount picks up a jiggle? Or five minute frames all night that might actually be usable. If those work maybe I can mix them in with the 10s that worked. If five minute frames all night show no jiggle then it is safe to say autoguider. Then try with PHD2 either with the batch program and CCDAutoPilot or with TSX and PHD2 running at same time for some long series that won’t be dithered. Try the batch program first.
If PHD2 works then the question might be for Software Bisque about why their autoguider isn’t cutting it.
9/13/2018 – Still no internet connection. By now the router has powered off and on with the lamp timer several times and the PC has rebooted. Yet Something else is amiss.
It is becoming clear that it is hard to get the equipment in the observatory to work and even harder to get it all to work remotely. It is so difficult that it is not realistic to expect success. The only time it will work is when everything in the observatory is working (rare), it is not cloudy (sometimes), I am in Torrey without guests (less than a quarter of the time), and a small moon phase for everything but the Ha narrow band filter (half the time). What does that Venn diagram look like? How often will those things overlap? Not every month. It has been 5 months it has not all come together now. Right now all the ducks are in a row except I am in SLC and cannot get an internet connection.
I do not believe I have ever seen a forecast like that. Severe clear. And the TeamViewer page shows that the router in Torrey has gone offline. I won’t be connecting to the observatory PC until the lamp timer recycles it. Which will be about 5AM. A perfect night goes under the what-can-fail-will-fail bus.
Reasonable success. A couple of late images have stuttered stars but most are usable. I got the RGB’s for the Eagle Nebula. Need some more greens for the Crescent. Then at least 3 nights (I only have until about midnight before the Eagle flies to close to Boulder Mountain) of Ha for the Eagle. I will gather more Ha for the Crescent along the way. I will have a lot of Crescent.
9/10/2018 – Good news from John Smith at CCDAutoPilot:
What may have happened is this. A session abort [due to weather] was in process which, among other things, parked the telescope. The rest of the abort process continued properly. Then, there was a clearing which may have attempted to restart the already aborted session. The non connection… Error 200 may have referred to the mount.
While this is messy, I don’t think there is a problem with your system.
I will try to start getting some Ha of Eagle Nebula until it gets too low around 12:15 and then back to the Crescent Nebula until about 4:30. Then some dawn flats.
Result: Ran an entire night without system failure! @focus2 did not return an error although it missed the focus first time and wasted four or five frames. Autoguider calibrated okay. BUT, more than half of the frames had double stars. John Smith at CCDAutoPilot thinks the mount is not working smoothly. Makes sense although very surprising. Today I will check the RA and DEC tightness and re-balance. More likely too loose than too tight. Would explain disappointing T-Point/slew accuracy.
Got some Ha frames for Crescent Nebula and Eagle Nebula. Tonight I need to front run the moon and get some color 2X2 frames for each.
September 9, 2018 – CCDAutopilot crash might have been caused by connecting to the wrong ASCOM Pyxis driver. There is a universal Pyxis driver that connects but seems to cause the crash. There is another Pyxis driver that needed to be clicked down into properties and then Setup and then told to use COM3. That may be working . . .
I gave up on the PHD2 with the batch file interface as guider and reinstalled the remote head guider cam. Found out Orion Starshoot Autoguider (SSAG) will not work with TheSkyX in 64 bit mode, period. Wrestled with new settings for the autoguider and autofocuser. Seemed to be getting closer. Getting a good V curve on focus graph and using autoguider in both axis working best. Need to note the settings for both here.
I am skeptical that CCDAP can handle the flip. Had to be recalibrated and started over after the flip tonight. The dither is giving me a lesson. The scope moves then the autoguider reacquires the guide star and move it back into apparent center/same place but the scope has moved by the dither amount. I hope. Either that or the scope moves and the autoguider moves it back, which would be bad.
. . .
It clouded up but at 1:48AM CCDAP was going to try again and then apparently crashed. I got 6 frames out of 36. Progress of sorts but back to the forum again.
September 8, 2018 – It is good I am keeping this log or I might not be keeping my sanity. Another night wasted. Without the log I would not be able to recall what in the world went wrong that I could not acquire an image since May.
CCDAutoPilot initialized with PHD2 and PHD_broker closed. But it crashed after focusing whether those programs were open or closed and even after I returned CCDAutoPilot to its initial state. More on forum thread here. Focusing with @focus2 quit working, too. Way too damned finicky.
Another wasted night.
September 7, 2018 – CCDAutopilot will not initialize. Sending the error codes to CCDAutoPilot forum.
Then . . .
The whole camera train fell out of the telescope!
All the wiring caught it. I had it as tight as I could by hand but rotator wiring previously getting caught must have loosened it. Could have been disaster but everything seems to work except for software above. Used wrench tonight to tighten camera on telescope. Wiring seems to be arranged and stable. Sheesh. I don’t think this is related to CCDAP problem.
Stymied yet again, I am taking an hour of 10 min Ha subs without CCDAutoPilot through TSX and guided with PHD2. There is no ability to dither since PHD2 and TSX don’t know the other one is running and can’t communicate. I could take more but I can’t afford to stay up late tonight. Plus, I am not sure why I am taking these shots.
September 5-6 – Cloudy. It even rained a trace. Clear Dark Sky forecast wrong again.
September 4, 2018 – CCDAutoPilot renewed my father’s annual software license ($80) in my name. Good of them. Now to retry PHD2 broker package with all of the new features, including pre-slew, in CCDAP working.
No testing tonight. Cloudy.
|A private observatory near Capitol Reef park and the town of Torrey.|
September 3, 2018 – Ran a new T-Point automated mount calibration model with about 100 points. Seemed to work okay.
Success too with the hard drive clone. I did not replace the drive in the PC but I am glad to have a backup. I downloaded Macrium Reflect Free software on the observatory PC and used the instructions here.
The clone was the end of my short success streak. I installed the PHD2 broker package to work with CCDAutoPilot but no luck. There is a pre-slew function in CCDAutoPilot that seems critical but is greyed out. Waiting for advice from the support forum here. — Answer is there is a much buried–and separate process to execute–requirement to renew the CCDWare license annually. I knew about that but it was not asked for in the upgrade and there was no apparent place to renew. Plus the existing license and password is in my father’s name. I am now waiting on advice for that . . .
I downloaded PHD2 manual and saved it in Observatory Docs. The manual suggests using an ASCOM mount driver and if you have an on-camera ST-4 mount connection, as I have, then use the ASCOM mount driver in Aux Mount. I downloaded the appropriate ASCOM from Software Bisque as directed by ASCOM standards page. It seems to connect. In PHD2 the ASCOM driver in Aux Mount only functions to receive declination info so a calibration is not required for every new declination. Will see.
The new CCDAutoPilot upgrade did not keep track of the link needed on the Settings/Control Settings/Cloud Sensor/File Path. Somehow it is in the log but not in the software (if it is not there, how did the log find it??):
C:\Users\David\Downloads\AAG_CloudWatcher(v720)\AAG_CCDAP4.dat (I found the closest file to this name and type and used it)
I also could not find the log to send to the CCDAP forum. Hint was in the read this first thing: See Help/Data Organization for the default location of the log file. Answer =
2 AM. PHD2 wouldn’t calibrate. That is a first. Re-ran the brain/first time wizard and then it worked. Also had a sudden and long struggle with @focus2 in TheSkyX with a focus failed to converge error. I don’t know why that happens. Maybe something to do with not liking particular stars. Wrestled with these finicky components until 3AM.
Cloudy nights now forecast. I am hoping to get the autoguider working and to get some 10 minute or even some 20 minutes sub-frames this lunar cycle. Been awhile since I got anything usable.
September 2, 2018 The Torrey House Press staff is coming today so some of my night will be as a star guide. Fun enough. I only note it here to remind myself later of what actually happens to the time where I am not able to get images.
At this point before imaging I need to:
- clone the observatory hard drive
- set up to use PHD2 with CCDAutoPilot: √
- install Orion Starshoot Autoguider √
- install the respective batch code and set up. The instructions are in observatory docs and link to download here.
- Run a new T-Point calibration. This is not absolutely necessary but I think the calibration could be a lot more accurate than it is currently. Read up on and review how first.
- Entertain the THP staff with Torrey’s dark skies.
September 1, 2018 In Torrey. Spent about five hours trouble shooting the issue that Windows showed I had WiFi and internet connection but I still could not browse. There are apparently some common problems with a “winsock” stack that can be cleared but that was not it. It turned out to be the AVG antivirus software was itself corrupted and was blocking the connection, even with a “clean boot.” One clue was that I could browse in safe mode. Another random way to waste a night.
August 31, 2018 Lost web connection from SLC. TeamViewer worked but could not connect to web power switch, the webcam and Weather Underground weather station was offline. Another lost night. Previous two nights my uncle was in town and staying with us so no time to fiddle here.
August 28, 2018 Back in SLC, I tried to get 24 ten minute guided Ha frames of the Crescent Nebula in Cygnus. I got maybe 2 good frames. Three more have oblong stars. The rest were star streaks. The autoguider is unimpressive when it works at all and then it goes completely to hell after the first re-focus. CCDAutoPilot can’t handle it without frequent intervention. A guy might still have to stay up all night, or not refocus, or both. The fact that guiding was lost after about six frames makes refocus suspect. I am also concerned that the scopes are not always centered enough in the observatory opening and that the guide scope might get eclipsed by the dome.
- Set up TSX to save the autoguider frames to see if getting eclipsed by dome.
- Monitor CCDAutopilot to see what happens after re-focus
- temporarily set refocus for 20 minute intervals
- If inexplicably fails send log to forum for help
- See about using PHD2 with CCDAutoPilot
- John Smith at CCDWare says, “Yep! See the links under Help/Version History for version 5.09.8.”
It took 3 hours of wrestling to get the oh-so-finicky TheSkyX autoguider to calibrate. I did finally get one good ten minute frame with better stars than unguided. With that I went ahead with a CCDAutoPilot session but it failed. Do I go back to five minute frames unguided? I have a dark sky and can take long exposures but not without autoguiding. PHD might work but apparently not with CCDAutoPilot. Daily serving of frustration with a side of sleep deprivation.
Zero streak since May continues.
August 27, 2018 Mostly wrestled, until 1 AM with TheSkyX autoguider. Calibration was almost impossible and even when calibrated guiding made things much worse, not better. Morals of story:
- Select “Imager’s built-in autoguider” and with camera connected choose external guide head in Camera Setup/Settings.
- Needed T-adapter to get focused with remote guide head on ST-80.
- There is no place in CCDAutoPilot to choose PHD that I can see so working with finicky TSX settings is only choice.
- See what 10 minutes unguided looks like. 20 too. Then try with Y axis corrections only. Try only correcting every 60 seconds using autoguider exposure delay.
- If use internal guider, not remote guide head, then probably need to change telescope focal length in settings.
- Try calibration again with backlash set to 0, and
- DEC/x-axis(?) relay turned off,
- switch to SB’s “Direct Guide” instead of relay,
- calibrate in optimal place with one dominate star before moving to target,
- turn down aggressiveness from 10 to 6
- use camera image short exposure delay (2X autoguide image time) to give autoguider chance to recenter after image download.
- Trouble shoot tips TSX pp. 486
- I may still have a problem with too small an opening on the observatory shutters. Keep in mind if guiding is lost.
See notebook page 81 for other notes.
I was also having the usual problems with the rotator. I got an image link and set the position angle and FOVI to the image. CCDAutoPilot could not now find the ASCOM driver (WTF?). There was some sort of conflict suggested back when by John Smith at CCDWare that had me using the ASCOM driver instead of TheSkyX for the rotator. Common sense suggests that if everything else is run by TSX, the rotator should work from there too. I turned off a setting in TSX that homed the rotator every time it connected. Once on power up should be enough. Seemed to work with TSX in CCDAP last night. Question is how to set the rotator position angle. Is the CCDAP initialize enough or is an image link/sync PA required in TSX every time?
√ Check for ability to open shutters further and tighten drive wheel on dome.
Still no successful image acquisition since May.
August 26, 2018 With help of Daniel Bisque of Software Bisque, I troubleshot the failing COM connection for the rotator to a failed port on the 8 port serial to USB hub (equipment failure). The trouble-shooting was a brute, straight forward process of elimination. The support forum thread is here. The filter wheel was not connecting because the IFW controller box had been shut off (pilot error).
I also got the blue screen of death on the observatory PC for the first time. I PM’d Christopher Erickson, one of the moderators at Cloudy Nights, to see what particular software he thought I should be backing up. He said he probably had the Meridian Controls file structure stored away somewhere. He also recommended cloning the observatory PC hard drive asap with a new hybrid hard drive. In addition he sent some links to new dome controller sources here. His links show just a bare printed circuit board as the new solution. It promises to be quite involved. I hope to nurse the Meridian Controls along for a good while longer.
August 25, 2018– On site in Torrey. I have been trying to figure out if the C14 is worth using. Wave theory has it that resolution = wavelength/aperture. I seem to understand that this equation would hold except that light is both a wave and a particle/photon and so is a quantum thing. This introduces uncertainty where noise, as Poisson figured out, is the square root of the number of events (photons). More intuitively one has to ask about the quality of the optics in the old C14 versus that of the higher quality TEC-140. The C14 has the big aperture and so has the main element for resolution. But the TEC-140 is better glass. The question is then how to measure actual results. If it is the measured FWHM of the stars (found in CCDStack and DSS), then knowing the camera pixel size and focal length of the scope it is possible to measure the size of the resulting stars in an image. Some of my prior results seem to indicate the C14 was getting about 3.2 arcsecs and the TEC 140 is 4.3 arcsecs. Hmmm. Also it is possible to say what angle of space a pixel in each camera sees. The TEC 140 with the SBIG 10/XME is 1 pixel = 1.4 arcsecs. (see page 77 in my Moleskine log for equations) The C14 with the Canon 500D is 0.246 arcsecs and 0.431 arcsecs with reducer. If true, it is worth working with the C14 again next summer during galaxy season. On the other hand I understand that seeing blur is often greater than one arcsec so the C14 might be overkill. But then why would any telescope need to be bigger? Bigger sensors? As in sensor arrays?
If I do work with the C14 again I should plan on focusing often. If remote, focus every hour. If on site and ambitious I should focus and lock the mirror until the meridian flip, then unlock, refocus, relock. Compare how that works with unlocked frequent focusing. And get a lot of data. Also found a site that says Canon 500D optimal ISO is 1600. So go to that. I might also get a expandable saddle for the G-11 so can more easily side mount and try the C14 outside. Might be cool for viewing.
To get a lot of data and take advantage of Dad’s observatory capability, and thereby also get some sleep, use CCDAutoPilot (AP). AP only delivers FITS files, not camera RAW somehow. So start a new library of dark files for the Canon 500D. DSS can stack FITS files so I can use Dark Master to match temperatures of the dark files.
- Today I will switch scopes:
- check tightness of all fittings on SBIG camera train √ There is still the tiniest bit of play coming from somewhere. Seems intrinsic and probably okay.
- change the observatory “dome geometry” back (from Xt = 17 to Xt = 13) √ In Dome Setup
- take the Robofocus off the C14 and reinstall on TEC 140 √
- C14 parts are in a marked baggie in observatory. Should leave the Robofocus drive axis set screws loose until manually focused, first for eyepiece, then with camera. Otherwise manually moving focus knob strips the set screws. Once any set screws are set only move focuser with Robofocus. The Robofocus step motor cannot be moved manually. Don’t try to do it with focus knob.
- take down the C14 and stow in garage √
- put on TEC 140 √
- start with eyepiece to:
- see how T-Point calibration is and start new calibration if necessary. Big moon will help get established if necessary.√
- install camera train and balance√
- @focus2 with Robofocus√
- write down what hash marks show on the TEC-140 _____________
- get a few closed loop slews and start a new T-Point√
- SBIG remote head with ST80 √
- autoguiding setup in TSX (focus on Moon? Vega?)
- get some 10 minute darks at -12
- As of night of 8/25 not successful in getting filter wheel and rotator to connect.
Had idea to get a pier extension to see over the wall for planets. Already ordered before I check to see if will work. Having it delivered to Torrey post office. Stay tuned.
August 20, 2018– The NGC 6946 images from the 19th were all out of focus. Now I am running out of summer Milky Way season in which I want to get some HaRGB. Moon is out now but I will be taking off the C14 until next year without ever getting a usable capture. The 36 frames of NGC 6946 that I have are crummy, probably because not tight enough focus. Conclusion is that with the current C14 setup (BackyardEOS running the Canon DSLR on the C14) I need to be onsite to manually do the meridian flip and to unlock mirror, refocus, and re-lock the mirror. Next year for galaxy season I will consider creating a dark library of FITS files so that I can run CCDAutoPilot. With autopilot I can refocus often (and thereby go without locking the mirror?), automatically flip, shut down if it clouds up, and close down when the session is over.
August 2018- Early August a private star party with Lynn and Patrick DeFreitas. Lovely guests. C14 was a beast to observe through in the observatory. All ladder work right over the desk. And discovered Venus and Jupiter too low to see over the wall in the outside pier in the summer. Saturn and Mars also up. Later, because I was still getting oblong stars, I spent about three hours collimating the C14. Good two ladder workout. It made the stars a little better, I hope. Looked better visually and in pics. Painted the trim on the observatory. Now wishing I had used a zippier color like midnight blue. Kelsie Moore from NPR affiliate KUER came down with Aaron McMurray to get some video for the NPR program Science Friday. Radio West at KUER has started a Radio West Films where Kelsie is the film maker. Results expected next month. Kelsie was tough about staying up and getting the shots, mostly time lapse. I also did a time lapse. I expected to have a 30 second clip but ended up with three seconds. In addition to her dark sky work, Kelsie and Aaron set up elaborate lights and mics to interview me. Very smokey all month and often cloudy. Still virtually no rain. On 8/6 and 8/12 and 8/19 I tried to get some more data on NGC 6946, the Fireworks Galaxy.
July 2018- worked on M106. Short nights! Changed the dome’s geometry to better fit the C14 so that it could see out the shutter opening. I bought a new PC in part so I could leave my rugged, reliable Windows 7 HP laptop in Torrey. Removed the GFI in the garage that controlled the observatory power line. Both somewhat helpful in making remote internet connection more reliable. M106 results included oblong stars. I found the images unusable. Our week after new moon on July 12 was clouded out over with no rain. Achingly dry in Torrey.
June 2018– Kirsten and I traveled some in the Banjo in June including a trip to Idaho for the board meeting in Boise for Western Watersheds Project. At the end of the month we held the Torrey House Press board and staff retreat in Torrey. Moon, clouds, people. No photos.
May 27, 2018- With the C14 focused I discovered a still mysterious vibration problem. Photos were terrible with dancing stars and large objects like planets and the moon could be seen dancing around in the viewfinder. At Software Bisque I posted my problem on the Paramount ME support forum here. Daniel Bisque there was more than helpful, literally. Some of what he asked me to do to trouble shoot set me back substantially. What seemed to finally do the trick was lubricating the mount as Software Bisque’s video describes how here. Stars in photos are still slightly oblong. This troubleshoot took a long time. I had to leave Torrey and then lost my internet connection to do some of the troubleshooting requests I had from Bisque.
May 7, 2018 – Put the C14 on for the first time in Torrey. I could not get the focuser to work and had a extraordinary exchange on Cloudy Nights here to figure out what was wrong. I posted here about the Zen-like lesson of sitting with a problem and letting all the clues come to bear, from wherever they may.
May 15, put Canon Rebel on the C14
April 14, 2018 – Star party with friends/donors of The Nature Conservancy of Utah. Used the TEC 140 with 20mm wide angle eyepiece.