Jump to content

  •  

CNers have asked about a donation box for Cloudy Nights over the years, so here you go. Donation is not required by any means, so please enjoy your stay.

Photo

Building 10Micron mount models

  • This topic is locked This topic is locked
723 replies to this topic

#601 FredOS

FredOS

    Ranger 4

  • -----
  • Posts: 323
  • Joined: 16 Feb 2017

Posted 08 December 2020 - 10:11 AM

Thank you both. Very helpful.


 

#602 C8er

C8er

    Mariner 2

  • -----
  • Posts: 259
  • Joined: 15 Sep 2012

Posted 08 December 2020 - 03:38 PM

Focuser is a Moonlite nitecrawler, which I assume is unlikely to be slipping. 

Even the better brands of focusers can have internal movement problems sometimes, and in the 10Micron owner forums I have seen at least one report of a problem with a Moonlite. As Tonk says, focuser flex is more likely lateral movement than slippage, but you can't really assume any part of the chain isn't going to be the source of a tiny flex/flop/drag/settling problem (as Tonk found with sagging felt in his scope rings, and fixed), so you need to systematically check every possible physical link in the chain, from the stability of the ground/flooring under the tripod/pier, the tripod feet, etc, through every link from there to the camera, if there are clear problems with your modelling RMS, including checking and re-tightening the bits you least expect to be the source of the problem.

 

I found a totally unexpected problem in my APM 140mm frac, with one slightly loose set of objective cell collimation/tilt push/pull bolts of the three sets (each of three bolts) securing the cell, that I couldn't find until I took the dew shield right off (not easy as it screws on and the large rough thread was jammed/very stiff) and looked at the cell and found that one collimation bolt had worked slightly loose somehow during a recent many hundreds of miles long transport in a series of commercial "fragile items" freight vehicles, which then allowed the rest of that set of three bolts to also work themselves just slightly loose. With that glitch fixed I get consistent approx 3" RMS models with that OTA on my GM1000, on an Avalon brand T-Pod tripod, not a pier.

 

Good luck with the hunt for the problem.

 

Chris M. 


 

#603 mccomiskey

mccomiskey

    Explorer 1

  • -----
  • Posts: 57
  • Joined: 17 Apr 2017

Posted 08 December 2020 - 03:49 PM

Yes, the felt example gave me a sinking sense of what I had let myself in for with this hobby!  I will be spending, I suspect, a lot of time on this project...


 

#604 C8er

C8er

    Mariner 2

  • -----
  • Posts: 259
  • Joined: 15 Sep 2012

Posted 08 December 2020 - 03:57 PM

I had an interesting experience recently experimenting changing the chain of connectors between focuser and camera between an all threaded set of adapters/extenders, and a setup that included a Baader clicklock compression adapter in the chain. I was pleasantly surprised to find that the Baader clicklock setup gave me just as good 3" RMS models (53 points) as the all threaded setup, so my example at least of the clicklock seems well rigid enough for me. I should though add that I did that with a fairly lightweight ZWO cooled CMOS colour camera and maybe a heavier camera and filter wheel etc would give a different modelling result with the clicklock. That was with a140mm APM frac on a GM1000. 


 

#605 EFT

EFT

    Vendor - Deep Space Products

  • *****
  • Vendors
  • Posts: 6,234
  • Joined: 07 May 2007
  • Loc: Phoenix, AZ

Posted 08 December 2020 - 04:05 PM

Yes, the felt example gave me a sinking sense of what I had let myself in for with this hobby!  I will be spending, I suspect, a lot of time on this project...

Believe me, you would rather be chasing down these small issues than working with a mount where these shifting/flexure issues are hidden by all of the other bad things going on.  You've probably got 90% of the battle done now before you even get started.


Edited by EFT, 08 December 2020 - 06:31 PM.

 

#606 mccomiskey

mccomiskey

    Explorer 1

  • -----
  • Posts: 57
  • Joined: 17 Apr 2017

Posted 08 December 2020 - 04:33 PM

Believe me, you would rather be chasing down this small issues than working with a mount where these shifting/flexure issues are hidden by all of the other bad things going on.  You've probably got 90% of the battle done now before you even get started.

Agreed!  I switched to the 10Micron from another mount, and I have never had reason to regret that.


 

#607 airscottdenning

airscottdenning

    Viking 1

  • *****
  • Posts: 560
  • Joined: 22 Aug 2008
  • Loc: Colorado

Posted 12 December 2020 - 11:48 PM

I am trying to import a csv file with a mask for my local horizon into the ModelCreator profile. I see the option for this and I see it mentioned in the manual, but I can't find any documentation for the file format.

 

My file contains two comma-separated values per row: azimuth, altitude and proceeds from north through east, s, w, and back to north.

 

I select the file in the profile editor but nothing changes in the main window and no mask is displayed when I check the little box that says "Mask."

 

Can anybody help me understand how to construct this mask csv file?

 

Many thanks,

Scott

Attached Thumbnails

  • Screen Shot 2020-12-12 at 9.44.01 PM.png
  • Screen Shot 2020-12-12 at 9.46.08 PM.png

 

#608 TestnDoc

TestnDoc

    Viking 1

  • *****
  • Posts: 714
  • Joined: 23 Jul 2009
  • Loc: USandA

Posted 13 December 2020 - 07:43 PM

I don’t have my scope PC in front of me. As I recall, the mask uses 2 numbers per entry. For instance if you have an obstruction due south that extends 25 degrees above the horizon, the entry would be 180:25. I saved the model on notepad and named it mask. I don’t recall if it’s .csv or .txt. Good luck. 


 

#609 RazvanUnderStars

RazvanUnderStars

    Apollo

  • *****
  • Posts: 1,140
  • Joined: 15 Jul 2014
  • Loc: Toronto, Canada

Posted 13 December 2020 - 08:37 PM

BTW, there's a new version, 2.6.1.0.

 

I just created an arbitrary mask and saved it, it created a semicolon-separated file of the following form:

 

9.81646565285789;38.2534775888717;
29.4493969585737;46.3214837712519;
55.1232302045097;45.7650695517774;
102.31777661248;48.5471406491499;
179.339276350288;54.3894899536322;
246.166754063975;68.8562596599691;
317.524908232826;50.2163833075734;


 

#610 mccomiskey

mccomiskey

    Explorer 1

  • -----
  • Posts: 57
  • Joined: 17 Apr 2017

Posted 14 December 2020 - 08:32 PM

I am hoping for some help from the collective braintrust here.  I appear to have plateaued at about 8"-10" RMS in my model building.

 

As a practical matter, that doesn't seem to be having a massive negative impact on my images, likely due to a short focal length (530) and large pixels (9u). 

 

However, it is driving me crazy, and I will shortly me moving to higher focal lengths and smaller pixels, so would like to get to the bottom of this, to the extent possible.

 

I have a rigid system, with cables all clamped down.  I am running a GM2000 with a Tak FSQ106, with an Eagle3 on a dovetail plate mounted on top of the scope rings.  So all rigid, only one cable coming off the mount, which is attached to the counterweight bar (using a variation of the 3D printed part that MOrion of MW fame has recently made available - Thanks!).  I am operating from a stone patio using the Centaurus tripod.

 

My understanding is that if I am stuck at 10" RMS, using a mount with absolute encoders, then that is because some aspect (or multiple aspects) of my system must have have some degree of mechanical play in them.  That is to say, that the mount knows where it is pointing, and presumably is getting that right, but the scope ends up pointing somewhere slightly different because something is not rigid/is shifting.  Is that correct? 

 

Based on prior posts, this weekend, I disassembled everything, checked every bolt and screw and retightened.  That includes tightening up the focuser to OTA grip, and all the threaded pieces linking the focuser to the filterwheel and camera.

 

I use both MC and MW4, but this evening used MC.  I did a 35 point model, chose Deneb from the polar align routine, centered it at 400% magnification, then ran a 58  point model and got 8.6" RMS.  Result below.

 

I am not sure how to interpret these graphs as an aid to identifying what might be causing the problem.  I am open to any ideas at all as to how to improve system performance, but if someone can give me a sense for how to read and understand the graphs, that might give me a lead in tracking things down.

 

Thanks in advance for any help offered.

 

 

 

   

  

Attached Thumbnails

  • Model 2.jpg

 

#611 RazvanUnderStars

RazvanUnderStars

    Apollo

  • *****
  • Posts: 1,140
  • Joined: 15 Jul 2014
  • Loc: Toronto, Canada

Posted 14 December 2020 - 11:02 PM

I compared your graph with what you posted in message #589. The trends are similar (i.e. higher errors along a line) but the orientation of that line is different (about 90 degrees in azimuth). This might provide a clue. What was different between the two times, apart from tightening you did which may have played a part? Was anything rotated?

 

Can you take pictures of the whole setup while it's building the model and then post one for one of the worst points? If you take one photo for each point, however boring, it's easy to correlate them with the points later. Maybe we see something. 

 

Also, what is the distribution of the errors in two models run in quick succession (like the 35-point and then the 58-point ones you did)? Are they oriented along the same line?

 

I also thought momentarily to ask about larger sources of heat nearby that would disturb the seeing (eg. cars with hot engines) but that wouldn't explain the positioning of the errors along roughly a line. 


 

#612 mccomiskey

mccomiskey

    Explorer 1

  • -----
  • Posts: 57
  • Joined: 17 Apr 2017

Posted 15 December 2020 - 10:48 AM

RazvanUnderStars,

 

The only thing I changed that I can think if that would cause a change in the orientation of a problem was to rotate the base mounting plate on the tripod, so that i could orient the tripod with one of the legs facing north to facilitate polar alignment (strangely, and contrary to the assembly instructions in the manual, it had come from 10Micron installed in a manner that properly oriented, north was between two tripod legs).  As a consequence of my change, the base plate and the tripod legs are now facing a different direction than they were when I ran the model in post #589.

 

Will do on the pictures, and will run two models the next clear evening and post the results.  Very much appreciate the ideas and the willingness to help.

 

No heat nearby.  Telescope is on a stone patio in the back of the house, about 15 feet from the building and there are no vents of any kind on that side.  Telescope was left outside all day, and was at ambient temperature when modelling started.

 

This is going to be a profoundly ignorant question, but how does one read the two graphs that MC produces?  I assume that the first (polar) graph plots the compass direction of the point and the degree of error? 

 

What are the axes on the second graph?  Apologies if this is explained in the manual, but I could not find it there.


 

#613 Whichwayisnorth

Whichwayisnorth

    Gemini

  • *****
  • Posts: 3,331
  • Joined: 04 Jul 2011
  • Loc: Southern California

Posted 15 December 2020 - 11:43 AM

What is your tripod sitting on?

 

Also I will occasionally get bad RMS when I run a model.  Usually it is because of one of two things.  A) I didn't manage my cables well enough and B) My focus was not perfect.  

 

So if your tripod is not on very solid material like a cement slab, that could account for some model error.  And if your cables are rubbing against anything even the ground, that could also introduce some error. And lastly if your focus is not perfect, that will also raise the RMS.  Use a focus mask if you are not positive it is perfect.


 

#614 mccomiskey

mccomiskey

    Explorer 1

  • -----
  • Posts: 57
  • Joined: 17 Apr 2017

Posted 15 December 2020 - 11:59 AM

Gemini,

 

Thanks for the questions.  By way of response:

 

The tripod is placed on a stone patio.

 

The cables for the camera and focuser are all short (as they connect to a computer mounted on the OTA and are bundled/tied down).  The only cable that comes off the OTA is the power cable for the computer, which is secured to the CW bar using a clamp, and then again to the mount (with a loop at each stage to provide enough play.  I will post a couple of pictures.

 

I refocus before every model run using SGP focusing and get very tight HFRs and FWHMs, so feel reasonably confident that focus is very good.


 

#615 RazvanUnderStars

RazvanUnderStars

    Apollo

  • *****
  • Posts: 1,140
  • Joined: 15 Jul 2014
  • Loc: Toronto, Canada

Posted 15 December 2020 - 12:33 PM

OK, this is useful: it tells us the problem rotated with the tripod and equipment, so it's not fixed relative to the ground (such as an area where it's sagging more, for instance). 

 

Regarding the charts, it would have been helpful indeed if they had labels or titles with the same info.

 

The one on the left shows the magnitude of the error for each point relative to what MC calls "polar angle". The error is displayed in two ways: the distance from the centre (the circles mark 5, 10 and 15 arcsec in this case but they vary depending on the range) and the colour (whose scale is above the charts).  I'm a bit confused by the use "polar angle" - AFAIK, usually it refers to the angle from the ground to the celestial pole, i.e. the altitude. I'm not sure if MC refers to the azimuth, rather, or to an angle on the sky sphere relative to the celestial pole - I hope others will chime in.

 

The one on the right appears to be fundamentally the same as the one on the left, but flipped (look at the position for the red dots). I'm not quite sure what the sign in the labels means however and what is the purpose of this second chart.

 

Hopefully those who used MC more can help further.

 

 

The only thing I changed that I can think if that would cause a change in the orientation of a problem was to rotate the base mounting plate on the tripod, so that i could orient the tripod with one of the legs facing north to facilitate polar alignment (strangely, and contrary to the assembly instructions in the manual, it had come from 10Micron installed in a manner that properly oriented, north was between two tripod legs).  As a consequence of my change, the base plate and the tripod legs are now facing a different direction than they were when I ran the model in post #589.

 

[...]

 

This is going to be a profoundly ignorant question, but how does one read the two graphs that MC produces?  I assume that the first (polar) graph plots the compass direction of the point and the degree of error? 

 

What are the axes on the second graph?  Apologies if this is explained in the manual, but I could not find it there.


 

#616 Whichwayisnorth

Whichwayisnorth

    Gemini

  • *****
  • Posts: 3,331
  • Joined: 04 Jul 2011
  • Loc: Southern California

Posted 15 December 2020 - 12:36 PM

Gemini,

 

Thanks for the questions.  By way of response:

 

The tripod is placed on a stone patio.

 

The cables for the camera and focuser are all short (as they connect to a computer mounted on the OTA and are bundled/tied down).  The only cable that comes off the OTA is the power cable for the computer, which is secured to the CW bar using a clamp, and then again to the mount (with a loop at each stage to provide enough play.  I will post a couple of pictures.

 

I refocus before every model run using SGP focusing and get very tight HFRs and FWHMs, so feel reasonably confident that focus is very good.

I have a stone floor in my kitchen which is where I sometimes set up my telescope when I am changing a setup or testing something. Mainly doing some sort of collimation.  What I have found is that I can have something centered in the crosshairs but if I walk up to my mount the target moves off center and then back again when I walk away.  There is no way I'd ever suspect that my stone floor would move if I didn't have live video showing it.   

Attached Thumbnails

  • IMG_4208.JPG

 

#617 mccomiskey

mccomiskey

    Explorer 1

  • -----
  • Posts: 57
  • Joined: 17 Apr 2017

Posted 15 December 2020 - 01:43 PM

I agree that given the tiny degree of movement required, everything should be suspect. 

 

I have tried to see if movement of the stone slabs is a factor, by centering a star in cross hairs at very high magnification using a Ha filter, and then walking to different positions around the mount while having someone monitor whether the star moves.  If there is any movement, is not detectable.  Might still be an issue, and I will move the mount to a different position on the patio to see if that gives a different result (which will be inconclusive if not, obviously) but I have run down any other potential causes as well...


 

#618 mccomiskey

mccomiskey

    Explorer 1

  • -----
  • Posts: 57
  • Joined: 17 Apr 2017

Posted 15 December 2020 - 01:48 PM

Posting pictures of the rig.  When I can run another model, will post photos associated with poor points.

 

Rig1.jpg Rig2.jpg


 

#619 Whichwayisnorth

Whichwayisnorth

    Gemini

  • *****
  • Posts: 3,331
  • Joined: 04 Jul 2011
  • Loc: Southern California

Posted 15 December 2020 - 02:09 PM

Looks fine to me.   

 

 

Edit:  I would also consider checking things like your filters to make sure not only are they mounted correctly but also that the threaded retaining ring that actually holds the glass in place is tight.  I have found that on some filters I can shake them and hear a rattle.  I can't help but wonder if the glass can move will that translate to movement of the sky?  I actually don't know either way.  Also I suppose it is possible that the lenses in the main objective can move.  I think I had that issue with an old TMB130 I had several years ago.  I had to remove the whole front assembly and take part of it apart to tighten something.  I vaguely recall what I did.   Then after all that I still was not really excited at the unguided performance so I ordered a pair of rings and added a third leaving one extra to put away (still have it). 

 

That third ring dropped my RMS from 6-8 down to 3-5

 

I am not suggesting that any of the things I just mentioned are your issue but instead I am suggesting that the problem could be something you dismissed as a possibility.  

 

Second Edit:  You may have mentioned what filter you are using.  I find that using an Ha filter and exposing longer helps quite a bit.  As well as I threw in a 2 second settle time.  An IR filter would work great too.


Edited by Whichwayisnorth, 15 December 2020 - 02:18 PM.

 

#620 mccomiskey

mccomiskey

    Explorer 1

  • -----
  • Posts: 57
  • Joined: 17 Apr 2017

Posted 15 December 2020 - 02:47 PM

Gemini,

 

I hadn't considered a third ring, but may give that a try.  I was thinking about switching to the Moonlite rings to see if switching rings would make a difference...

 

I am using 50MM unmounted filters which are held in place through pressure by a face plate.  Nothing seems loose, but then I am unlikely to physically notice something that causes such a small change.

 

I am using 4 second exposures and 3 second settle times during model building. 

 

I will try building the model with Ha to see if it makes a difference.


 

#621 RazvanUnderStars

RazvanUnderStars

    Apollo

  • *****
  • Posts: 1,140
  • Joined: 15 Jul 2014
  • Loc: Toronto, Canada

Posted 15 December 2020 - 03:16 PM

Two more suggestions:

  • try without the filter wheel - for the purpose of the experiment, as long as you can focus, not being at the ideal backfocus distance is ok. I'm thinking of it because that, and the focuser, are the only movable parts. The case itself is rigid but the rotating wheel inside might have a little bit of looseness. 
  • while you're building the model, MC outputs the error after each point. Watch the numbers and if you see one high, take a photo of the setup right then. It might help identify the issue when you know exactly in what positions things move.

 

#622 C8er

C8er

    Mariner 2

  • -----
  • Posts: 259
  • Joined: 15 Sep 2012

Posted 15 December 2020 - 03:17 PM

Hi mccomiskey. As whichwayisnorth commented, I also reported recently that I had one of the three sets of collimation/tilt push/pull bolts in an APM refractor objective cell work itself slightly loose and that let the objective cell move slightly. IF that hasn't already been part of your troubleshooting it might be worth taking off your the dew shield and checking that those are all still snug, assuming the FSQ has an adjustable objective cell. I would also definitely explore more on how non-flexing the patio really is under your tripod.

 

You say there are no heat emitting vents on that side of the house but I presume you are you imaging over the house roof during part of your modelling runs. Any house roof will often be warmer than the ground around the house at night and may be giving off heat that could potentially be badly disturbing the air above the house roof. In my models, while they are good (approx 8-10" with a C11 Edge HD SCT with mirror locked, and consistent 3" with a 140mm APM refractor), the worst points in them are usually where the model images were being captured with the scope pointing across the top of our house roof. You could investigate that by doing some modelling runs away from home, or just with the mount situated in a different part of your backyard (or from your frontyard if you have one) so the bearing from the mount to the house is quite different, and see if the location of the bad points move in relation to that. 

 

Those robust handles on it make me think that is probably a fairly heavy camera? Are you quite sure that there is no lateral flop/flex in your focuser? Even the best brands of focusers can have occasional problems.

 

Cheers,

Chris M


 

#623 nathanm

nathanm

    Vostok 1

  • -----
  • Posts: 186
  • Joined: 28 Jun 2020

Posted 15 December 2020 - 03:19 PM

As per previous posts here, I don't have a 10micron yet but I am getting ready to order one. 

 

Here are some other question about the 10micron models.   

 

1. The various posts on this thread discuss RMS error.    Which I assume is root mean square error in tracking.   How is that determined?  Is it from actual plate solves?     PHD2 measures error by movement of the guide star, which it knows from image analysis.  

 

2.  I understand from the posts here that flexure of things in the imaging system is a major source of residual tracking error after building a model.    However, I also think that some level of flexure is accounted for in the model, is that correct?   

 

3.    How sensitive is a 10micron to balance?   I use some complicated imaging set ups - like two imaging systems piggyback, or even an array of 3 or 4 imaging systems.   There is sometimes a problem with my existing mounts that in certain positions the center of mass of the complicated set of things put more or less torque on the mount motor, and change its performance - i.e. the motor struggles in some positions and that impacts tracking.

 

A different way to ask #3 is that in principle, if you wanted to perfectly counterbalance an OTA, you would have the counterweight have a 3D center of mass in the same position as the OTA.  Instead, we tend to approximate this with a weight on a stick.   Which means that a complicated set up on a mount experiences often has different balance for some positions.  Would this be a problem for 10micron models?

 

Nathan


 

#624 mworion

mworion

    Explorer 1

  • -----
  • Posts: 53
  • Joined: 08 Mar 2016
  • Loc: Hohenbrunn / Munich

Posted 15 December 2020 - 03:32 PM

Hi,

 

there is a good description from Tonk about rings (please search in this thread), and here is my solution to get the best out of it in my setup. I setup the rings as wide from each other as possible.

Sorry I got trouble with images attached for some reasons: https://github.com/m...r/setup_gear.md : I used the full distance of the plate which is about 380mm.

 

Michel


Edited by mworion, 15 December 2020 - 03:34 PM.

 

#625 RazvanUnderStars

RazvanUnderStars

    Apollo

  • *****
  • Posts: 1,140
  • Joined: 15 Jul 2014
  • Loc: Toronto, Canada

Posted 15 December 2020 - 03:57 PM

Hi Nathan,

 

For #1: it's the RMS of the pointing error, calculated during model building by comparing the calculated vs real position where the mount is pointing. The tracking accuracy is much better (I have the smaller model, GM1000 and its specs read: "Average tracking accuracy < +/− 1” typical for 15 minutes (< 0.7” RMS)")

 

For #2: correct, the reason for using multiple stars during model building (vs only 3) is for the mount to model flexure. 

 

For #3: I don't have direct experience with that kind of setup, but given that the mount has absolute encoders, it knows if something is pulling or resisting. To image without guiding still requires a rigid system (this thread is about debugging a particulare case of rigidity problems) and a good model, that covered those positions that you refer to in the last sentence.

 

 

As per previous posts here, I don't have a 10micron yet but I am getting ready to order one. 

 

Here are some other question about the 10micron models.   

 

1. The various posts on this thread discuss RMS error.    Which I assume is root mean square error in tracking.   How is that determined?  Is it from actual plate solves?     PHD2 measures error by movement of the guide star, which it knows from image analysis.  

 

2.  I understand from the posts here that flexure of things in the imaging system is a major source of residual tracking error after building a model.    However, I also think that some level of flexure is accounted for in the model, is that correct?   

 

3.    How sensitive is a 10micron to balance?   I use some complicated imaging set ups - like two imaging systems piggyback, or even an array of 3 or 4 imaging systems.   There is sometimes a problem with my existing mounts that in certain positions the center of mass of the complicated set of things put more or less torque on the mount motor, and change its performance - i.e. the motor struggles in some positions and that impacts tracking.

 

A different way to ask #3 is that in principle, if you wanted to perfectly counterbalance an OTA, you would have the counterweight have a 3D center of mass in the same position as the OTA.  Instead, we tend to approximate this with a weight on a stick.   Which means that a complicated set up on a mount experiences often has different balance for some positions.  Would this be a problem for 10micron models?

 

Nathan


 


CNers have asked about a donation box for Cloudy Nights over the years, so here you go. Donation is not required by any means, so please enjoy your stay.


Recent Topics






Cloudy Nights LLC
Cloudy Nights Sponsor: Astronomics