autofocus – IMT3b Observatory

Viewing Report 13th June 2020 – IMT3

23:31 – 04:02

So I opened the dome late this evening as it was not due to be clear. However an opening in the cloud meant I could test guiding again on the 12″, especially whilst it was light in the late Spring weeks.

The first job was as always to focus which brought me to a reading of 61944 at 19.83℃.

Another small job was to sort the guider FoV out. I went ahead and used M92 to align the guider.

The final FoV settings are here for completness.

Set AS1600 to Gain and Offset 10 due to cluster being very bright and I needed to set a standard of 60 seconds minimum exposure. Gain 139 and Offset 21 gas saturated unless I selected 15 seconds, Gain 75 and Offset 12 saturated at 30 seconds so hence 10 and 10 which came in about 58k ADU.

I then performed a slew to a nearby star so I could centre the scope, there platsolve completed successfully and I updated TSX and the FoV for the 12″ with the new angle.

The first image of 60 seconds came down and was out of focus, I then realised changing the profile SGPro forgot the autofocus setting, so I had to stop the run, delete the images and set the original focus point then rerun.

Next I ran a few images but then to my horror I had the same guiding issue, where the star moves being dragged up and down in a periodic way. I slewed elsewhere and tried again and the problem did not occur. I was near M92 and just East of the Meridian and quite high up. Not sure why that is a problem.

I could not resolve, I waited a while then performed a meridian flip and low and behold the problem went away, again not sure why. I still have this terrible noise coming from the RA motor/gear area. I decided to bite the bullet and take off various caps on the scope listening and looking inside. I decided it was not after all, the through the mount cabling but coming from the RA gear itself, so I looked for the MEII guide for removing the worm block and then followed the instructions to take off the RA cover.

This gave me instant feedback on what the issue was, the belts driving the axis were making a noise. On looking through forums on Bisque.com I found a few people with similar issues and needing to grease the belts, they were told Lubriplate was a good grease. This is an American grease so I will find a similar here and then apply, I will ask Bob first for his suggestion.

So the night wore on and the LRGB frames of M92 I thought I would take whilst testing guiding progressed. At one point the imaging stopped due to cloud. I just caught the dome before it closed to change the safety sensor due to cloud. When it cleared it never really cleared, with the sky temperature reading about -14℃.

Nearing the end of the imaging session, I had caught about 15 frames of each of the filters.

The guider was behaving mostly with he odd funny jolt. By 3:30 am the sky was lightening very quickly.

By this time I had stopped guiding and imaging. I closed the dome, slewed the scope to the flat panel and proceeded to take a set of LRGB flats for Gain 10 Offset 0 and also Gain 139 and Offset 21 as request from the previous nights imaging.

Viewing Report 12th June 2020 – IMT3

22:09 – 02:02

It was unexpectedly clear this evening so I opened the dome late, so no cooling down.

Focus 60,279 Lum filter at 19.93℃

Slewed to Arp 286 as I had seen on Flickr and wanted to see what it looked like in 12″. I noted I needed a new set of Darks for the Lodestar off axis guider after we had changed the driver recently, so I set about taking those with PHD2.

The problem I then saw was no guide star in the FoV! I really need a rotator!!!!!

So I will image without the guider. The first image jumped as I realised the auto-guider was still on and trying to track nothing sending the mount this way and that. So I disabled and set about running 3 x LRGB for 300 seconds a piece

The first image was Luminance and looked ok, although bright due to high level cloud and no astronomical darkness this time of year. I also noted the mark on the filter caused by the LensPen! I won’t use that again. However I do expect that to come out with the flat so not too bothered. Also I could go clean it I just don’t want more dust on the filter so will leave until I have a reason to take the camera off again.

For reference here is the luminance flat.

Next up was Red filter for 300 seconds

Then Green filter

Then finally Blue filter.

So I left it to run for 3 x 300 seconds each filter. Meanwhile Mil Dave came online and opened his dome and we decided to go for a joint target to see how they compare. Given Arp 286 was below Dave’s hedge I mentioned the Coddington Nebula IC2574 which is actually a galaxy, being the object in my latest image from the travel rig with the Esprit120 and the comet passing. It has some interesting star clusters in it. We agreed on luminance and 300 seconds. Here is Dave’s result ….

Meanwhile the humidity kept rising

At just gone midnight I finished the short run on Arp 286 and slewed to Coddington to catch up with Mil Dave. First I did a quick refocus as the temperature had dropped about 2℃ since I started.

The refocus put me near but not near enough so I changed to 62,056 for a better HFR which worked. Then I went on to do Coddington and here is my result.

I had problems with unguided exposure, then had a problem finding a guide star, then the object was behind a tree!

So next I slewed to Mil Dave’s choice of object, here is Mil Dave’s image

I managed to find a guide star straight away and grabbed this

Notice my screen stretch is different hence the bright background. It is just a quick screen grab off the NUC. My turn to pick now, so went for several objects but all behind offending hedges at Mil Dave’s house. So I sent Dave off to choose one. He came back shortly with Arp 278. So off we set. Here is mine.

and of course Mil Daves….but no, he forgot to save it 🙁

So my object next, I selected one nearby to save the rather long dome rotation I just did, and the resulting loud noise when it jammed! I need to look at that. Meanwhile Mil Dave trundled his round manually. So I choose NGC 7331, also known as C30 and Herschel H53-1, so also on my list for the Herschel 400. This is what I got.

Meanwhile the cloud from the South East started to creep in and my daughter, who just came in even at this hour of 1:36 said it was foggy outside.

Dave’s grabbed the NGC 7331 below.

I was about to suggest a Sharpless object when the clouds rolled in enough for me to shut the dome. It was reading -4.8℃ sky temperature and the limit was less than 30 for overcast, so I manual overrode.

Just before it shut this is the image I got of SH2-126 which is impossible to see since it probably needs the Ha filter rather than luminance.

Well a good night all round, given we thought it was going to be cloudy it was nice to come out and play with Mil Dave and go hunting faint fuzzies, so a goodnight from me and a goodnight from him ?

Viewing Report 1st June 2020 – IMT3

20:25 – 00:15

Dome once again open to cool down

Solved FoV for Lodestar guider on 12″

Now guiding on star to West side of mount around 0 degrees Dec and near the Meridian to see if problem reoccurs, graph looks ok but there is a slight regular pattern of error.

Slewed to M92 which is a glob currently at Alt +74, Dec 43 and on the East side of the mount where I had issues last night. I plate solved the image.

The auto focus run looked good

Found guide star by moving M92 slightly off centre. Now guiding. All looked good for a short while. So each section of PHD2 graph is 25 points, so when you have 100 points selected there are 4 sections of the graph. This mean each section/column represents own my case 250 seconds as I expose for 10 seconds, so just over 4 minutes per section. So the errors I saw last night were about 12mins apart. This cannot be the work gear as it has a cycle of 2min 29sec. This is from the latest Paramount manual for the MEII.

Tracking at the sidereal rate, one revolution of the worm takes 149.6 seconds (2 minutes 29 seconds).
The right ascension gear has 576 teeth.
The declination gear has 475 teeth.

The error I am seeing is about 4mins. I cannot continue imaging as every few frames are trailed.

Viewing Report 15th May 2020 – IMT3

21:58 – 02:02

Just setting up for an imaging run and to test imaging without temp compensation to see if the 12″ keeps focus without it. I started by myself then was joining by GingerGeek and then Bob.

Performed a SGPro autofocus run on Mag 7 star produced focus position of 71,828 @ 4.6 HFR at 14.47℃.

Resulting M98 image from 1st autofocus run

The resulting image was good with good star shapes. Although I suspected at this point the seeing was not excellent.

I let the sequence run for a bit imaging M98 through LRGB and then decided the HFR was gradually getting larger so I performed a 2nd Autofocus run which came in at position 72,215 HFR 5.7 at 13.97℃.

Again I let the sequence continue for at least 4 images and the performed another Autofocus run, note all the time this was on M98 and not slewing away to another star. This came in at focus position 72,697 @HFR 5.7 at 13.82℃.

I continued this routine again and performed another Autofocus run on M98 focus position 73,441 HFR 5.4 at 12.98℃.

I then decided, due to struggling to get a good HFR on focus runs to see if the autofocus was introducing an issue so I changed the autofocus setting from 9 data points to 11 data points to try and get fuller deeper curve. The resulting curve was better and more complete on both sides of the U shape. I then imaged further and then attempted an autofocus with the settings change for the step size from 2500 to 1500 and data points from 11 to 15. This was because I felt we always have a flatfish bottom to the autofocus which at this focal length of 2.5m shows the quality of the seeing with a narrower flat bottom being better seeing. The new autofocus came in at position 73,534 HFR 5.1 at 12.66℃. Meanwhile we kept noticing satellites going across the ASC which I now believe are potentially StarLink so very annoying.

The new autofocus settings seem to work better. Anything less than 1500 step size would be less than the seeing, as proved tonight so I may find that 2000 is ideal, a test for another night. Also noted that Red filter was showing the worst HFR changes due to seeing and humidity was around 75%, again worthy of note to see how good the seeing is. The guiding was all over the place tonight, again another indicator of poor seeing. So all these things are not poor setup or poor software but poor seeing!

This I believe was the ISS going over captured in the ASC.

I was really pleased GingerGeek and I had spent time a week or so ago when the Moon was around working out the location of the Field of View (FoV) indicator on TSX, it makes it much easier to find a guide star, although tonight M98 had a couple strategically placed which was great.

FoV for Off Axis Guider and the OS main camera

By 2pm the cloud had started to appear, first at South Winston with Steve’s setup, then at Mil Dave’s at Tadley and finally here some 15 minutes later. The guide star was lost by PHD and SGPro in a well ordered fashion did what it is really good at and stopped imaging.

SGPro can’t continue imaging due to guide star loss 🙂

Here is where I got up to so LRGB on M98 for the night with 15 x Luminance and 12 x Red, Green and Blue was the original first image for each was there wrong exposure time, so RGB at 2mins and Luminance at 5mins. Very happy for an evening testing and gathering data at the same time.

Here is a set of image statistics charts for each filter for the HFR changing over the evening whilst I refocused. Next time I will focus once and not refocus and see what happens with the temperature drop.

Final look at the AAG weather station as the cloud sensor which is Infrared makes the dome unsafe and shuts it.

Here is the final view from the ASC

and of course to finish the evening off another satellite!

Viewing Report – 26th April 2020 – IMT3

21:30 Slew to M85

Camera -23°C, focus point 74534, temperature 15.81℃

Frame & Focus / Plate Solved / Centred

21:38 Autofocus Run – Failed.

21:50 Integrating M85 Event 1 Frame 1 for 300 seconds
Lost guide Star

22:13 Switch Guider to SW Lodestar

Integrating M85 Event 1 frame 1 for 300 seconds. Aborted run as M85 approaching the meridian and guiding graph was not looking good. Guider not calibrated.

Meridian flip to Chertan in Leo, Tak FS-102 now on top of configuration, Esprit below the OS 12″ so will be the better to guide with the QHY5 MiniGuideScope attached to the FS-102.

22:51 Slewed to a star field near Chertan for Auto Focus run. Start focuser position 72885, final focuser position 72178 ….. nice graph.

22:56 Calibrated QHY5 MiniGuideScope guider

Review calibration

23:08 Calibration suceeded, guiding.

Stoped guider and slewed to M85 which had just crossed the meridian.

Frame & Focus / Plate Solved / Centred

23:21 Integrating M85 Event 1 Frame 1

23:44 Telescope connection lost due to poor communication

23:48 SGPro reported USB error – lost FLI Focuser. FLI connects Ok in The Sky X.

Fix is to click on the spanner in SGPro for the Focuser and Rescan, Focuser now reconnects.

00:01 Re-centered M85

PHD2 restarted as it appeared to have hung.

00:12 Integration new Luminance sequence Event 1: 180s, Event 2: 300 seconds.

Viewing Report 25th March 2020 – IMT3 12″

Viewing time period – 19:15 – 23:05

As I found last night I need to get Temperature Compensation working for the focuser if I am to produce any decent images. SGPro has a Temperature Compensation Trainer which I will follow.

Temperature Compensation Trainer in SGPro under Tools menu

I started @19:35 by taking an image of a star field to make sure there were no bright stars and then ran the autofocus routine. This came back with a focus position of 76974 at 11.5℃.

Next I started the Temperature Compensation wizard which seems to measure the difference in focus position over a 5℃ decrease in temperature. It takes the initial reading above and then you wait until the temperature has dropped by 5℃.

Temperature Compensation Trainer Starting figures

It took 3.5 hours for the temperature to drop by 5℃. So @23:08 I then ran the autofocus routine and got the next image and result. This resulted in 531 steps per 1℃ of temperature change.

Viewing Report 10th/11th August 2019 – IMT3 Observatory

Viewing time period – 21:04 – 04:28

So whilst the wind is blowing a gale and branches have come down off of 300 year old oaks where I live, the weather decided to ease as we went into the evening. There were still gusts of high wind but nothing really to be too concerned about for the dome.

Due to more changes on the mount and the polar alignment changes I needed to do a new TPoint run. I first tried to complete a recalibration run that would add additional data, however after a 30 point run the pointing got worse to the point where I was not landing up on the object but several fields of view. So I bit the bullet and deleted the recalibration data and the original model and started again.

It took me about 1.5 hours to run 30 points on the East side of the mount and another 1.5 hours to run the next 30 on the West. I was happy with the results and I turned on some new Terms as I went through. As you select the term you can see the resulting change to the position of the telescope. So rightly or wrongly I used this process a few times when the pointing was either not improving or it just needed to improve a little.

I also created a Supermodel of the data and once again enabled Protrack. I then went to my usual target of the Elephant Trunk to try and get use to pointing to the right object and then imaging it from SGPro. I decided the easiest thing was to use The SkyX to move to the object as I knew I wanted to be centred on SAO 33570 a star in the trunk. I did this and with my new pointing capability since this evenings TPointing, the scope landed up pointing at pretty much the right area. So instead of Sync and Solve I left it at this location for tonight.

The Polar Alignment report produced by TPoint on this new 60 point model showed very little error in either RA or DEC which is a testament to the long hours I put in drift aligning the mount.

At 2:04 am I then tried to cool the camera but it was non responsive……..this threw me for a while then I remembered the other evening having this same problem which probably meant the camera power was not on.I went into the dome to find that was the case and the reason once again was the fuse on the Mount Hub Pro had melted. I cut this off and put a chocky block in for this evening to bypass the fuse, but I did unplug everything else from the mount hub pro whilst there was no fuse there.

Whilst doing this I was reintroduced to nature with a Hornet the size of my thumb bouncing around in the dome. After quickly removing myself from the dome I cam back armed with an insecticide and sprayed the offending hornet. It kealed over and died quickly.

Next I focused and this worked very well, a nice V curve on the Luminance and then I switched to Ha for the imaging, This would be slightly out but I need to find a strong HA source of stars to be able to focus with Ha.

Once again I ran the Image Sequencer to see if this would work given I had made some changes suggested by my good friend Mil Dave to the guider settings in PHD2 and SGPro. However once again I was foiled with some new error messages, I am either getting use to this or possibly very fed up, SGPro may be a great piece of software from a functionality perspective, but it is complex, unintuitive and a pain. The error complained about the PHD2 profile ‘OS 12 Lodestar Guider’ is not valid.

This is indeed my profile and is valid so not sure about this, another thing to investigate when I am not so tired. The follow on message was Could not start the autoguider and connect to the equipment so aborting. This is to be expected.

So I went back to Frame and Focus and too a single 5min shot, guiding on a good star that was in the FoV of the Lodestar off-axis guider. I took a 10min image and then a 20min image.

Looking at the TPoint model there was a nice improvement for where I started with a with an RMS, Root Mean Square of 100 so when pointing the object I am targeting will be within 100 arc seconds of the centre of the CMOS chip, so 1.7 arc minutes, whereas now it is leas than 1 arc minutes out at 57.9.

I then went back to SGPro to try and fix this error as I don’t like giving up. I changed more settings within SGPRo and PHD2 around the error size for the guider to settle, however SGPro was still waiting for PHD to report it had settled even though it was now guiding.

Error message

I turned off both the Settle At and the Settle Auto Guider check boxes. This then allowed me to bypass the whole settling thing which wis is really not that important to me as I manually setup guiding first and now the sequence has started at last!

Finally the guider looks very smooth and the only thing now stopping me from taking some more images is the fact it is 4:28am and I am very tired and it is getting light. So I will disconnect and shut down until the next clear night at a weekend. All in all a very productive night.