10-Micron GM2000HPS - installed and waiting :-)

[CCD Imager]

I believe it does, the only missing function was dithering, but there is a fix in the pipeline. I don't think there are many 10u owners using SGP. There are 2 ASCOM drivers available, one from Per and one from 10 micron. I use the former which is compatible with Pers Model Maker - a must if you have a 10u mount.

Adrian

[CCD Imager]

As I said, I think it does, but check with SGP writers. Also there is a Youtube video for an hour about SGP, go have a look. BTW, you probably wont need guiding with a 10u mount

Adrian

[ChrisLX200]

I was told anything over 20 minutes requires guiding! What is 10U?

10-Micron... the mount doesn't need a plate solver, and tracks unguided (30mins is the longest sub I've used so far, others have gone out to an hour). All this comes at a high price unfortunately.

ChrisH

[CCD Imager]

The plate solver is really used when building a model in Model Maker. The resulting pointing errors are typically less than 30 arc secs, so no need to plate solve after a meridian flip and also no need to guide.

Chris, the price is 1/2 that of competitor mounts with absolute encoders.....

Adrian

[ChrisLX200]

Sound impressive but 30 min is the shortest I do.  10 micron advise guiding on exposures over 20 minutes so I intend to guide.  SGP don't seem to have the answers on compatibility. Not many 10 micron users using that software I guess. Must all be using maxim. It is amazing how little information there is on these high end mounts. I am thinking of buying one but the lack of information out there is not encouraging.

I'm using SGP with the GM2000HPS-II, it does meridian flips OK without the need to plate solve afterwards. Only issue I have is SGP sits there with a dialog box waring me it is about to flip instead of getting on with it :-)  Probably some setting I;ve missing.

ChrisH

[ChrisLX200]

The plate solver is really used when building a model in Model Maker. The resulting pointing errors are typically less than 30 arc secs, so no need to plate solve after a meridian flip and also no need to guide.

Chris, the price is 1/2 that of competitor mounts with absolute encoders.....

Adrian

Yes it's a 'bargain' price at the moment :-) Or it was until the Euro went up again...  The new Mesu300 is going to be interesting, has Renishaw absolute encoders.

My sky model is usually around 3 arcsec RMS, and the only 2 occasions I;ve tested the pointing accuracy it has been a good deal better than 30arcsec!

ChrisH

[ChrisLX200]

Sound impressive but 30 min is the shortest I do.  10 micron advise guiding on exposures over 20 minutes so I intend to guide.  SGP don't seem to have the answers on compatibility. Not many 10 micron users using that software I guess. Must all be using maxim. It is amazing how little information there is on these high end mounts. I am thinking of buying one but the lack of information out there is not encouraging.

There's no harm in guiding if you want to, you don't get bonus points for imaging entirely unguided. How accurate the mount tracks is dependant on how good your sky model is - good model = good tracking, and of course the amount of error which is acceptable depends on what focal length you image at.

ChrisH

[ChrisLX200]

Thank you Chris that  is what I was waiting to hear!  When  you say "no plate solve necessary" would you trust that  when shooting a large mosaic sequence over the period  month? 

Yes, but a plate solve is such a trivial thing to do you could check if you like. Mine puts the scope on target to within 5 arcsec - just how accurate do you need to be with a mosaic? Most people leave quite a large overlap anyway. I should add, mine is on a permanent pier - not a portable ring, so I can afford the extra time to develop a good model. Mind you, even a 95-point model is finished during astronomical twilight and before I can start imaging so it represents no loss of time, and it's all automatic so I just have a cup of tea while watching the mount workout :-) 

ChrisH

[ChrisLX200]

Sounds good Chris. 

Once you have the model locked can't you just re-use the same model for duration of the mosaic, since your mount is stable  on a pier ?  With the mosaic Its obviously good to keep the pixel error as low as possible . I try to keep it down to within 10 pixels with a 20% overlap. plate solve gives me that . I don't  like to change anything during the mosaic i.e. alignment etc

Yes for sure, you can save several models in the controler's memory and the last used is retained for immediate use next time it's powered up. It is normal to have to re-synch your model if you have removed the OTA in between times (just plate-solve and it shifts the whole model to suit), and the first time I used my mount I checked this. It turned out my dovetail sits accurately in the saddle and a re-synch was not required, but I will continue to check for a while in case that was a fluke...

ChrisH

[CCD Imager]

Moonrocks, successful unguided imaging for your minimum of 30 minutes depends on a few important factors. 1) The scope you put on the mount and how rigid is its structure? A moving primary or wobbly focuser will harm your chances considerably 2) The focal length of your scope - you are much more likely to see round stars with say a 500mm scope, but if you are imaging at 3M FL, then its much tougher, although I know one 10u user who has this FL and achieves 40 minutes unguided, but he does have a top class scope with everything fixed, a 100 point model and permanently mounted. So as you can see there is no easy answer to your question.

I am at the Kelling star party atm and had a problem with my router not connecting to the mount, I had to buy a new one today! Any way I did a manual 8 point model and ran 15 minute subs with an FSQ and ALL images over a 3 hour session had nice round stars.

Adrian

[ollypenrice]

Sound impressive but 30 min is the shortest I do.  10 micron advise guiding on exposures over 20 minutes so I intend to guide.  SGP don't seem to have the answers on compatibility. Not many 10 micron users using that software I guess. Must all be using maxim. It is amazing how little information there is on these high end mounts. I am thinking of buying one but the lack of information out there is not encouraging.

I'm impressed! Narrowband, I guess?

Olly

[CCD Imager]

Hi Adrian;  I would be interested to see you work you don't seem to have any links here!

Yeah, I've been bad recently and would you believe I have at least a couple of years data to process. I took down may old web page about a year ago when I switched providers and haven't decided on what web host I should use. I also don't know how to upload here, but I'll look into it!

You could have a look at the 10 micron forum for discussions about the mount, you can't contribute unless you are an owner, but there is a wealth of information there. Have you searched this forum and also cloudy nights?

Adrian

Adrian

[perfrej]

I start each night with a single align sync and then do nothing in terms of pointing updates. This works well as does the 5 arc-sec dither. Pointing is within 2-10 arc-seconds.

/per

[ChrisLX200]

Well I now have an ODK12 on order (nods at Adrian... :-) )  so it will be interesting to see what quality model this will generate on the GM2000HPS. I'm quite prepared to drill holes in it and stabilise the mirror further if needed, I tend to do that sort of thing.

ChrisH

[CCD Imager]

Whooaa Chris, I hope it works for you!

I think you will find the mirror is solidly mounted with I suspect, minimal movement, no more than a refractor lens in its cell.

Clear here at Kelling Heath and the mount is behaving very nicely

Adrian

[ChrisLX200]

Glad it's clear and you should have at least tonight and tomorrow night mostly clear, anyone that can forecast further ahead than that should contact the Met Office for a job...

The mount is working really well here too, the last 4 full models I've run (99 points) have all been 3" RMS, pointing is superb with no need to plate solve, I've twice done a meridian flip tonight and just continued imaging with nothing further needed - just relying on the accuracy of the pointing model without bothering to check with a plate solve. All subs 30min unguided of course with pin-point stars now.

I talked to Barry about how the ODK was put together, my only concern was (and remains) the stock focuser may not hack it. I don't like Crayfords anyway so may end up with an Atlas.

ChrisH

[CCD Imager]

Similar here Chris, the pointing accuracy is eerily accurate after a meridian flip! No plate solves needed here either. I do have a problem with my Tak FSQ with the dedicated reducer leading to tilt, shall try and debug tomorrow daytime.

The stock ODK focuser is okay, I had it for a short time, but as I mentioned, I chose the Starlight feathertouch focuser.

Looking at your kit inventory, I reckon you need to offload a few scopes

Adrian

[ChrisLX200]

Similar here Chris, the pointing accuracy is eerily accurate after a meridian flip! No plate solves needed here either. I do have a problem with my Tak FSQ with the dedicated reducer leading to tilt, shall try and debug tomorrow daytime.

The stock ODK focuser is okay, I had it for a short time, but as I mentioned, I chose the Starlight feathertouch focuser.

Looking at your kit inventory, I reckon you need to offload a few scopes

Adrian

Well, I think the 12" Newt will have to find a new home, I don't do visual observing any more and that is a visual scope. A fair few premium eyepieces should go for the same reason.

ChrisH

[Big Bang!]

Afternoon Chris,

I have a 1000HPS but I am still learning the basics of set up and alignment - Per has been helping me along.  Please forgive my lack of understanding, and, at the risk of making you experts laugh I would like to ask about your alignment procedure. What is the best technique for accurate results?

I have an Altair RC250TT (f8) and have been using a Celstron illuminated reticule eyepiece on a diagonal to centre the stars.  I have been utterly meticulous and patient in getting the stars right in the middle of the square but my RMS results have varied.  Usually, I would need a finder for the first two or three stars as they would be quite a way off; after about twelve or so the stars would be reasonably close to the square - the odd one might be inside the square but all would require some adjustment.  I have been working on this for a couple of months. Finally, last week, having removed some of the inaccurate stars I managed to get an RMS down to just under 4" and I was hopeful of great things.  However, the first 15 minute test exposure showed oval stars and knowing what is possible from these mounts I was disappointed. 

So last night I tried a different approach: I used my Atik imaging camera to centre the stars, accepting that this would probably be very time consuming.  However... Whilst the first one required use of the finder scope the second was half way into the FOV (2000mm f/l); the next 3 or 4 were within 6 about star widths of the centre and the following 60 were all just about smack in the middle - very minor adjustment, if any, required.  This level of pointing accuracy was WAY beyond anything I had seen before and I was elated.  Yet, I still only got a final RMS of 7", which seems quite poor to me, given that I barely had to nudge the vast majority of stars.  Unfortunately, the cloud then rolled in and I was unable to test the tracking. You will understand I can barely wait for the next clear night!  

You clearly have very quickly got to grips with the 2000 and your experience will have helped you along. I am at the other end of the spectrum; I don't yet have Pinpoint and am therefore unable to use Per's Model Maker.  May I ask you if you see an obvious error in my technique.  How do you go about manual alignment?

Very grateful for any help you can offer me.

Gus

[Xplode]

I can't use Per's model maker either so i slew with Stellarium and plate solve and sync with Astrotortilla.

I would guess it takes around 30 sec for a slew and solve this way with a pretty fast computer, Astrotortilla usually solves in 9 sec or fails at 10-15 sec. If if fails just move a little and try again.

[ChrisLX200]

Afternoon Chris,

I have a 1000HPS but I am still learning the basics of set up and alignment - Per has been helping me along.  Please forgive my lack of understanding, and, at the risk of making you experts laugh I would like to ask about your alignment procedure. What is the best technique for accurate results?

I have an Altair RC250TT (f8) and have been using a Celstron illuminated reticule eyepiece on a diagonal to centre the stars.  I have been utterly meticulous and patient in getting the stars right in the middle of the square but my RMS results have varied.  Usually, I would need a finder for the first two or three stars as they would be quite a way off; after about twelve or so the stars would be reasonably close to the square - the odd one might be inside the square but all would require some adjustment.  I have been working on this for a couple of months. Finally, last week, having removed some of the inaccurate stars I managed to get an RMS down to just under 4" and I was hopeful of great things.  However, the first 15 minute test exposure showed oval stars and knowing what is possible from these mounts I was disappointed. 

So last night I tried a different approach: I used my Atik imaging camera to centre the stars, accepting that this would probably be very time consuming.  However... Whilst the first one required use of the finder scope the second was half way into the FOV (2000mm f/l); the next 3 or 4 were within 6 about star widths of the centre and the following 60 were all just about smack in the middle - very minor adjustment, if any, required.  This level of pointing accuracy was WAY beyond anything I had seen before and I was elated.  Yet, I still only got a final RMS of 7", which seems quite poor to me, given that I barely had to nudge the vast majority of stars.  Unfortunately, the cloud then rolled in and I was unable to test the tracking. You will understand I can barely wait for the next clear night!  

You clearly have very quickly got to grips with the 2000 and your experience will have helped you along. I am at the other end of the spectrum; I don't yet have Pinpoint and am therefore unable to use Per's Model Maker.  May I ask you if you see an obvious error in my technique.  How do you go about manual alignment?

Very grateful for any help you can offer me.

Gus

Hi Gus,

The key to accuracy is having a plate-solver, the key to speed and automation is Per's ModelMaker utility. I do use PinPoint right now but there are other plate-solvers that don't cost an arm+leg. ModelMaker does rely on MaximDL + PinPoint for the capture of images but Per is working on a new version that will not be, I believe he mentioned October sometime for release. However, someone else wrote a script that enables using SGP to capture images (see: http://forum.mainsequencesoftware.com/t/using-sgp-for-modeling-run-image-captures-on-10micron-mounts/2035) but unless you're deeply into playing with software at this time I would just wait for Per to release his new version.

You don't need to centre stars, the mount only needs to know it's exact pointing position (which is easily obtained via a plate-solved image) and the data from that is then compared with its encoder readings to create a sync datum point. The mount really doesn't care if it is pointing at a star or not :-)  A plate-solved image will have a centre position calculated to sub-arcsecond accuracy, there's no way you can match that sort of accuracy with an eyepiece plus star image. Plus it's a lot less hassle!

So if you are doing this manually, rather than using ModelMaker, just point the scope anywhere in the sky and take a short exposure (preferably using a red filter to darken the background - because you'll be doing this just after dusk probably and before it's properly dark). I use SGPro + PinPoint like I said, but it would also work equally well (if a little more slowly) using the Free Elbrus solver, or PlateSolve2 or Astrometry.net.  When SGPro plate-solves an image it will sync the mount - you would need to set the option "Syncs Refine Model" on the mount so these data points are added to the model.

So basically you would select random points covering the whole sky, or as much of it as you can see from your position, adding refinement points as you go.

My philosophy dealing with a technically advanced mount like this is that it really needs the software tools to make best use of it, there's little point skimping here and thereby not realising the full potential of the (expensive) mount you have bought. Had I not already purchased PinPoint for other reasons I would have bought it for its superior speed and performance. I would not have bought MaximDL though - far too expensive to be used for this job alone (I prefer to use SGPro as a sequencer). I already had an ancient copy of MaximDL from over 10 years ago which I tried and amazingly it worked with ModelMaker :-) 

So your options right now are:

a) Use SGPro (not free software) plus PinPoint (not free either!) then use the script method to use SGPro to interface with ModelMaker.

Use SGPro (still not free!) but use manually captured images plate solved internally by SGPro.

c) Wait until Per releases a new version of ModelMaker (which I think will still rely on PinPoint).

I should add... I have never tried manually creating a model for the 10-Micron mount! But if I didn't have ModelMaker working for me the above is the way I would do it. Anything to avoid using one of the eye-peace things :-)

ChrisH

[Big Bang!]

Well, that's why this Forum is so useful.  My problem has been that I don't know what I don't know and I am therefore fumbling around in the dark.  You are saying that the Mark 1 eyeball is not good enough to extract the tracking quality that 10 Micron mounts offer, no matter how careful you are.  That being that being the case, having invested for the best I have to go that one step further.  Yes, Per has told me to be patient, the new ModelMaker should be ready in October.  

Just to be clear: Having set the mount to Syncs Refine Model, each photo will  send a pointing message to the mount which will essentially compare that to the encoders and adjust the model accordingly.  Do you ever need to Delete Stars with this method and what final RMS would you expect?  To get 30 minutes unguided tracking what RMS do you need?

This shortcoming is likely the cause of my drift; roll on October.

Thank you very much indeed.

Gus

[ChrisLX200]

Well, that's why this Forum is so useful.  My problem has been that I don't know what I don't know and I am therefore fumbling around in the dark.  You are saying that the Mark 1 eyeball is not good enough to extract the tracking quality that 10 Micron mounts offer, no matter how careful you are.  That being that being the case, having invested for the best I have to go that one step further.  Yes, Per has told me to be patient, the new ModelMaker should be ready in October.  

Just to be clear: Having set the mount to Syncs Refine Model, each photo will  send a pointing message to the mount which will essentially compare that to the encoders and adjust the model accordingly.  Do you ever need to Delete Stars with this method and what final RMS would you expect?  To get 30 minutes unguided tracking what RMS do you need?

This shortcoming is likely the cause of my drift; roll on October.

Thank you very much indeed.

Gus

Well I think Tonks (on CloudyNights Forum) would give me an argument about that, he exclusively uses the eyeball and eyepiece method and claims good success :-)

Yes, you have 'Syncs Refine Model' set to ON while building the model, but remember to switch it to OFF when finished or any further syncs will still be added as refinements (which might not be what you want to do). The Mount can hold 100 points maximum and the best (i.e., most representative) model is the largest dataset that you can have. Note that may not give you the lowest RMS error! But it will give you the best performance. ModelMaker allows you to simply delete the worst data point. How good your model is will depend a lot on how rigid your OTA/focuser etc are.

Note also two other things the mount needs for best modelling - barometric and temperature data from which it can calculate the refraction parameter. It also needs very accurate time. For the former I use Per's BlueAstro Stickstation, for the latter I have installed an NTP Time Server Daemon (Meinberg) on my Obs PC which continuously updates the mount keeping it accurate to within 2-3 milliseconds. There will be a 15arcsec pointing error for every 1 second the time is out.

Including last night, I'm up to 5 full model runs now (99 points) and all were 3"RMS. I really can't say what accuracy you need - but the smaller the error the better I think.

ChrisH

[CCD Imager]

Gus

6 arc secs pointing accuracy is very good, to put that in perspective you are pointing to within 1/7 the diameter of Jupiter!, not many can boast that! As Chris suggests, you are so much better off having automatic plate solves for building the model. Pinpoint is a great and fast program, well worth the money (about £100), considering how much you have spent on the mount. I did a couple of models whilst at Kelling Heath this last week and a 40 point model took 25 minutes.

I'm mulling over getting an automated program and am trying out CCDAutopilot atm, but still undecided, each seems to lack compatibility with some software.

Adrian

[Big Bang!]

Thanks chaps,

Am I right that the Pinpoint/ModelMaker combination basically does the job automatically? 

Gus