#1  
Old 16-02-2013, 09:55 AM
pvelez's Avatar
pvelez (Pete)
Registered User

pvelez is offline
 
Join Date: Aug 2006
Location: Sydney
Posts: 1,250
Latest Dumb question - TPoint

I understand that there are no stupid questions - just stupid people. So I may reveal myself as the latter...

If I have a good TPoint model but so-so polar alignment, I will have good pointing accuracy but will still have trailing stars

Is this correct?

No prizes for answering correctly - apart from admiration for your brilliance

Pete
Reply With Quote
  #2  
Old 16-02-2013, 10:04 AM
Mighty_oz (Marcus)
Registered User

Mighty_oz is offline
 
Join Date: May 2006
Location: Atm somewhere in Perth
Posts: 575
Looks like your not so dumb after all Yep trailing stars if unguided. I find that to get good unguided images ( 5 mins + ) i have to be spot on almost perfect with the PMX for PA.
Reply With Quote
  #3  
Old 16-02-2013, 10:27 AM
pvelez's Avatar
pvelez (Pete)
Registered User

pvelez is offline
 
Join Date: Aug 2006
Location: Sydney
Posts: 1,250
Thanks for that. I had suspected as much.

I compiled a good 140 point model earlier in the week, applied a supermodel to it and found I had 19 arcsecond accuracy (as reported by TSX).

Yet I still had trailing stars after 2 minutes. Bummer.

Looking at the PA report when assembling the model, it advised me to raise the mount 6 or so tics. But when I ran the supermodel, it told me my PA was good enough (az) and excellent (alt). So I thought I was mistaken.

I take it that the PA report changes after running supermodel and that I need to check the PA report before applying the supermodel.

I think.... I'd best go and read the manual again.

Pete
Reply With Quote
  #4  
Old 16-02-2013, 05:31 PM
PRejto's Avatar
PRejto (Peter)
Registered User

PRejto is offline
 
Join Date: Jan 2011
Location: Rylstone, NSW, Australia
Posts: 1,397
I have been "wrong" in two of my answers to you so far....But, I think the super model wins when it comes to polar alignment. I think you probably need to get azi spot on!

Peter
Reply With Quote
  #5  
Old 16-02-2013, 05:36 PM
gregbradley's Avatar
gregbradley
Registered User

gregbradley is offline
 
Join Date: Feb 2006
Location: Sydney
Posts: 17,877
Best to do several iterations of T-point of say 26 points and run the supermodel and apply the corrections to polar alignment from the PA report. Then delete model and do a new one say 26 and keep correcting until The Sky X says its close enough.

Then do a long model and use it for pointing accuracy or even Protrack guide corrections. I think the manual states an ideal t-point model size before using Protrack.

A big t-point model does not affect star trails - polar alignment does. It only helps with go-to accuracy which with a widefield setup is not critical.

Greg.
Reply With Quote
  #6  
Old 16-02-2013, 07:49 PM
Mighty_oz (Marcus)
Registered User

Mighty_oz is offline
 
Join Date: May 2006
Location: Atm somewhere in Perth
Posts: 575
I've found that with smaller T-point models under 40 that the PA recommendations bounced a bit up then down etc, even with 100+ models i do a couple to get it spot on if i can. Depends on whether u are permanent
( my case ) or setting up each day.

Marcus.
Reply With Quote
  #7  
Old 20-02-2013, 12:48 PM
acotiga (Andrei)
Registered User

acotiga is offline
 
Join Date: Nov 2012
Location: Christchurch
Posts: 32
There is however another posibility. You may have field rotation rather than wrong PA. Simply put, if your guiding star is too far away from the centre of the image, even with the autoguider the stars will drift due to the difference in angular velocity.

Could you be more specific and describe the setup because I think that 19 arcsec shoudn't give you any headaches.
Reply With Quote
  #8  
Old 23-02-2013, 09:43 PM
Logieberra's Avatar
Logieberra (Logan)
Registered User

Logieberra is offline
 
Join Date: Mar 2010
Location: Canberra
Posts: 1,635
Still learning Tpoint as well, but another sneaky test for PA (other than what a Tpoint SuperModel reports) is provided during PEC capture.

After loading and fitting a PEC log within TheSkyX, I see a definite drift (slope down to the right) in RA over time. So, my PA is off as well. Confirmed during unguided work, I can't go for too long. Further confirmed by the big trackbox size that I must use to actually collect PEC data!

Other users stay right around that PEC curve centre line with no drift. I look forward to that
Reply With Quote
  #9  
Old 23-02-2013, 10:51 PM
PTW (Patrick Wallace)
Registered User

PTW is offline
 
Join Date: Feb 2013
Location: Abingdon, Oxon UK
Posts: 5
There is a risk of confusion here between two distinct cases:
  • If your mount and control system does TPoint pointing corrections, but tracking rates are fixed, you can get spot-on slews but then unguided trailing, even if the polar axis is well aligned.
  • If your mount and control system does both pointing and tracking corrections, such as ProTrack on Paramount+TSX, then you should get both accurate slewing and minimal trailing, at least in the middle of the FoV.

Both of these apply even if the polar alignment is way off. But you will then get field rotation, and if you're guiding with a star at the edge of the field that means there will be trailing at the centre of the FoV.

Once your polar axis is within about an arcminute of the refracted pole, that's good enough unless you're planning to do very long exposures on a very wide field. And there are several choices, depending on whether you're more interested in minimizing unguided trailing or field rotation, and where in the sky you want the best results.
Reply With Quote
  #10  
Old 24-02-2013, 11:37 AM
pvelez's Avatar
pvelez (Pete)
Registered User

pvelez is offline
 
Join Date: Aug 2006
Location: Sydney
Posts: 1,250
Thanks to everyone for their comments.

My question was prompted by eggy stars in unguided images. I was surprised that I was unable to take round stars using my PMX. I suspected that it was poor PA. I had a PA report suggesting that I was 8 arcminutes too low. But as I had a 150 point TPoint model, I could slew quite accurately despite this. However, with poor PA, I would still get eggy stars.

Turning on guiding, I had a much better result.

So Peter's first point in his post is what I had anticipated.

To illustrate, attached are screenshots from the central portion of 2 images. Both have integration times of 120 seconds. The scope is an RC8 with a QSI583 binned x2. This gives me a plate scale of about 1.39 arc seconds/pixel. The first is unguided, the second guided. You can see the difference.

I'm sure that collimation is also at play here so its not just the mount/PA.

The eggy stars are across the field so I don't think its field curvature.

I've since refined my PA - the PA report suggests that I can lower my mount 1.8 arc minutes while az is excellent. This gives me Sky RMS of 92.2. This is with a 20 point model which I used to smarten up the PA. he attached pics were taken before I did this.

The next step is to obtain a bigger model, turn on Protrack and then check the PE.

Out of interest, is anyone surprised by this performance? The scope has a focal length of 1604mm which is not exactly long and I'd have hoped to have longer subs than 2 minutes without eggy stars - but that may be explained by the PA.

Pete
Attached Thumbnails
Click for full-size image (Screen Shot 2013-02-24 at 12.22.38 PM.png)
111.2 KB25 views
Click for full-size image (Screen Shot 2013-02-24 at 12.22.49 PM.png)
100.5 KB23 views
Reply With Quote
  #11  
Old 24-02-2013, 08:00 PM
PRejto's Avatar
PRejto (Peter)
Registered User

PRejto is offline
 
Join Date: Jan 2011
Location: Rylstone, NSW, Australia
Posts: 1,397
Welcome to IIS Patrick! (He is the author of T-Point!)

Pete, I still think you need better PA particularly in Altitude. And I'm sure I will be corrected if wrong, but after you get PA much better, you need to do 200+ run of t-point to derive any benefit from protrack.

Peter
Reply With Quote
  #12  
Old 24-02-2013, 08:30 PM
PTW (Patrick Wallace)
Registered User

PTW is offline
 
Join Date: Feb 2013
Location: Abingdon, Oxon UK
Posts: 5
Pete -

Quote:
I've since refined my PA - the PA report suggests that I can lower my mount 1.8 arc minutes while az is excellent. This gives me Sky RMS of 92.2. This is with a 20 point model
92.2 arcsec RMS is a very poor result for a Paramount, so there's obviously a lot to get right before assessing ProTrack results.

Do an all-sky calibration run of say 200 observations and see what TPoint Super Model says. And after adjusting polar alignment do a "recalibration" (as opposed to a full calibration), selecting the portable-mount option. This will retain your hard-won existing model and just tweak the ME and MA terms (plus IH, ID and CH).
Reply With Quote
  #13  
Old 06-03-2013, 09:42 PM
pvelez's Avatar
pvelez (Pete)
Registered User

pvelez is offline
 
Join Date: Aug 2006
Location: Sydney
Posts: 1,250
I suspect that I have found the problem - or a problem at least.

My pc clock is gaining time!

I synchronised my pc time via the internet and then checked it with Verify the SkyX Time function - a difference of 6 seconds.

I started a calibration run - 5 minutes into the run, the discrepancy is 82.8 seconds!

How can a longT Point run accurately determine where the mount is pointing compared to where it should be if the time reported by my pc is wrong?

So it should come as no surprise that the PA report is all over the shop.

Has anyone seen this before? I've read about clocks losing time because of the CMOS battery on the motherboard running down but this clock is gaining time.

All very, very weird - I am running a fitPC2 with XP if that helps.

At least it is consistent - after 10 minutes it has gained 152 seconds ie 2.5 minutes

Is there a way to turn off the link between TSX and the pc clock? I can't find it if its possible. My thinking is a workaround would be to sync the clock then run TSX without it being infected by the dodgy clock. Just a thought.

Pete
Reply With Quote
  #14  
Old 06-03-2013, 09:49 PM
Logieberra's Avatar
Logieberra (Logan)
Registered User

Logieberra is offline
 
Join Date: Mar 2010
Location: Canberra
Posts: 1,635
SB forum posts often refer to the 'D4Time' (Dimension4) app to regulate the PC clock with major internet time servers. Free download. Worth a shot?
Reply With Quote
  #15  
Old 06-03-2013, 09:52 PM
Logieberra's Avatar
Logieberra (Logan)
Registered User

Logieberra is offline
 
Join Date: Mar 2010
Location: Canberra
Posts: 1,635
Welcome Patrick.

P.S. I enjoyed reading the TPoint background stuff on your website. Gave me a new appreciation for what we can easily take for granted in TheSkyX! Cheers.
Reply With Quote
  #16  
Old 06-03-2013, 10:10 PM
pvelez's Avatar
pvelez (Pete)
Registered User

pvelez is offline
 
Join Date: Aug 2006
Location: Sydney
Posts: 1,250
Quote:
Originally Posted by Logieberra View Post
SB forum posts often refer to the 'D4Time' (Dimension4) app to regulate the PC clock with major internet time servers. Free download. Worth a shot?
Thanks for this

I've downloaded ClockMon which is keeping it stable at about 5 seconds difference which i suspect is within the TSX tolerance.

I'm running a short TPoint calibration to see what it looks like with the time fixed - fingers crossed.

Pete
Reply With Quote
Reply

Bookmarks

Thread Tools
Rate This Thread
Rate This Thread:

Posting Rules
You may not post new threads
You may not post replies
You may not post attachments
You may not edit your posts

BB code is On
Smilies are On
[IMG] code is On
HTML code is Off

Forum Jump


All times are GMT +10. The time is now 04:26 AM.

Powered by vBulletin Version 3.8.7 | Copyright ©2000 - 2024, Jelsoft Enterprises Ltd.
Advertisement
Testar
Advertisement
Bintel
Advertisement