Jump to content


Photo

Wierd polar align calculations


  • Please log in to reply
21 replies to this topic

#1 Corpze

Corpze

    Member

  • Beta Tester
  • 207 posts
  • LocationSweden - Arvika

Posted 13 August 2016 - 10:34 PM

Hi, just wanna share my latest experience in this jungle of making the right settings in the right order.
I was going to make a new PA for the upcoming season here in Sweden.

Loaded the old pointing file, slewed to a 40deg Dec star near the meridian, cleared the old config, synched the star with CdC and set the new home pos.
Started a new pointingfile, three stars, same Dec, same pierside. Calculated the file and got a neat error of 340 degrees RA and 70 degree Dec.

Started over again but checked my location, time, date, old configs... Even in CdC... Everything looked ok.
Made the same procedure and with same result (just like Einsteins quote - only a fool expects different results from the same procedure. Well well.

Next time, started a new instance of autoslew, cleared config, new CdC instance, synched a star and new home position. But this time, autopoint did a 3 star,msame pier side for me. Calculated the file and got 7x12 degree off...

What dis i do wrong? It would make me understand Autoslew if i know what made it go so terrible off.

/Daniel

#2 ChristerS

ChristerS

    Member

  • Beta Tester
  • 87 posts
  • LocationFROSON, SWEDEN

Posted 14 August 2016 - 01:29 PM

Hi Daniel,

 

Last spring after updating to the AS 5.2.2.3 the OTA run into the pier when doing homefind, the problem was the pier side was wrong, hav never happend before so check if the pier side is correct. Try also reset AS settings there is an option for that.

 

/Christer


Cheers,
//Christer

www.astropix.se


#3 Corpze

Corpze

    Member

  • Beta Tester
  • 207 posts
  • LocationSweden - Arvika

Posted 14 August 2016 - 01:56 PM

Will check that, just updated to 5.2.2.3 as well.
How ever, the autopoint didn't get those huge error values... I did a PA after those values and when the weather allows mw to do a bigger autopoint model i will double check the PA error values again.
/Daniel

#4 RamaSpaceShip

RamaSpaceShip

    Member

  • Members
  • 165 posts
  • LocationIn the forest near Bordeaux, France

Posted 16 August 2016 - 06:32 PM

Hi Daniel,

 

When doing PA or a new pointing model, I always verify that the error associated with each star (displayed on the bottom line of Autoslew) is in reasonable limits (say less than 400 in each axis), if not I discard the star.

I do that because I noticed that such values fool the resulting pointing file such that its calculations are wrong. Why is there such error, I don't know. I suspect that the pier side value is misinterpreted causing the initial error.

 

Please do the same checks and tell me if you have similar behaviour.

 

Best regards.

 

Bernard



#5 Corpze

Corpze

    Member

  • Beta Tester
  • 207 posts
  • LocationSweden - Arvika

Posted 27 August 2016 - 10:07 PM

Ok, did some new test today with zero progress.
I reset autoslew and started off from scrath.
Choosed a 40 degree dec near the meridian with the right pierside. Synched and set a new home position.
Then i tried a synch in sequence but pinpoint couldnt solve that one.

I then tried to make a three star pointing model in autoslew, with huge errors, 140degree in dec and 70 min in RA... But i was only 1,3 degree off compared to CdC and the values in autoslew.
I have no idea whats going on :-/
I rebooted my pc several times, restarted AS with reseting the configuration each time...
Internet time synch and the right observer coordinates is right as well :/

#6 Waldemar

Waldemar

    Member

  • Beta Tester
  • 236 posts
  • LocationNetherlands

Posted 28 August 2016 - 07:19 AM

I had the same kind of pierside problem with the old 5.1.6 version. I just set pierside to west, which apparantly means east to the mount and that solved the problem, but it is klind of weird, Must be a bug that is still not fixed, although not everybody seems to have that issue.

 

Waldemar 


Waldemar
--------------
Mount: DDM85 Standard 12V ; Telescopes: TEC140 + 140FF + Quad TCC ; Celestron RASA; TMB 92SS + 27TVPH Cameras: ZWO ASI 174MM cool; SBIG STF8300M; ATIK4120EX; TRIUS SX-36
Filters: Astronomik and Baader NB + RGB; Lunt double stack 75 mm unobstructed signature series + BF 3400; Dutch weather and light polluted skies.  

#7 Corpze

Corpze

    Member

  • Beta Tester
  • 207 posts
  • LocationSweden - Arvika

Posted 28 August 2016 - 07:29 AM

I can try to set the "wrong" pier side to see if that solves it...
The first star was 0.0 degree dec and 0.3 degree RA off and the second star that much.
The strange thing is that last weekend i did a autopoint and that time, it manage to pinpoint the images, not this time :-/

 

As you see on the image, the reported coordinates from CdC and the coordinates from the mount isn't that faar away, so i presume it isn't any fault between my planetarium and the mount. Seems to be an translation error just like the pierside is wrong, hence the huge errors?

 

The pierside is how ever the right side as reported from autoslew (pierside east) and m counterweight was indeed pointing east.

 

Attached File  error.jpg   138.29KB   1 downloads


Edited by Corpze, 28 August 2016 - 08:11 AM.


#8 si6pack

si6pack

    Member

  • Beta Tester
  • 127 posts
  • LocationGauting, nearby Munich

Posted 28 August 2016 - 09:26 AM

I do not whether my comment will help to solve your problem, however I would like to comment 2 things:

1) When I looked at you posted AS window, I see that both axis were calibrated, but what I miss is the "homefind" command. If I forget to issue my homefind, then I get wired pointing results. Therefore I would speculate, that issuing a "homefind" might solve you issue...or do you have absolute encoders?
2) In my experience, doing an autopoint over an existing autopoint file (config file) never added to the pointing quality. Since 3 to 4 nights with a clear sky are rare at my side, I have to pull my setup back in my garage very often. When I start over again, the first step I do is to make 100% sure that my old config file is deleted in AS, then I do the polar alignment, manully delete the polar config (sometimes it is deleted automatically by AS, sometimes not; no pattern here) and then do the 22-star pointing file...then, Everything works just great.

Hope this helps,
Christian
DDM85S
ASA12N-OK3 or TEC140
FLI 8300 & FLI FW2-7 or XPress H694 & XPress USB FW

#9 ChristerS

ChristerS

    Member

  • Beta Tester
  • 87 posts
  • LocationFROSON, SWEDEN

Posted 28 August 2016 - 10:40 AM

Hi!

In the image I see GPS 0Sat, i.e. no satellites found that could make som problem.

//Christer


Cheers,
//Christer

www.astropix.se


#10 Corpze

Corpze

    Member

  • Beta Tester
  • 207 posts
  • LocationSweden - Arvika

Posted 28 August 2016 - 11:02 AM

Si6pack:
I have done a homefind, but maybe i have to do another after i have reset the configuration?

Motor on
Homefind
Synch a star near the meridian
New home position
New home find?
Autopoint or manual three star?

Christer;
I di not use GPS but will check the settings so AS hasn't added a GPS for me :)

/ Daniel

#11 si6pack

si6pack

    Member

  • Beta Tester
  • 127 posts
  • LocationGauting, nearby Munich

Posted 28 August 2016 - 12:15 PM

This is what I do:

First night:
Motor on
Homefind
Clear old config
3-star polar alignment using SEQ
Solve frames
Using AS - load autopoint
Calculate 3-star config
Polar align
Correct slew via Az and Alt
If not cleared automatically by AS, manually clear (polar) config
Slew somewhere on same side of meridian
Take picture and solve
Sync solved picture (Maxim)
In AS: cross ckeck if config is cleared, then set new homeposition
In SEQ: 22-star pointing file
Solve frames
In AS: load auotopoint file
Calaculate config normally Ra < 0,2 arc sec, Dec dito
Save config for next start
Start SEQ run - MLPT enabled, dither 4 pix, grouped by slot
After SEQ finished, Telescope parked (near homeposition, counterweight shaft on E), then motors off, shut down all soft, shut down computer, axes fixed by inbus screws...

Second night (and following, so weather will):
Unfix screws in Ra and Dec
Switch on motors, Ra and Dec calibration
Homefind
Start SEQ run for desired object
99% perfect frames

Whenever I did an additional polar alignment with my config file used, it became crap...if I follow what I described above, everything fine...forgot to mention AS 5.2.1.9 and SEQ 2.0.9; latest AS version (5.2.2.3?) screwed whole system...! Therefore I am back on 5.2.1.9

Regards,
Christian
DDM85S
ASA12N-OK3 or TEC140
FLI 8300 & FLI FW2-7 or XPress H694 & XPress USB FW

#12 Corpze

Corpze

    Member

  • Beta Tester
  • 207 posts
  • LocationSweden - Arvika

Posted 28 August 2016 - 02:13 PM

This is what I do:

First night:
Motor on
Homefind
Clear old config
3-star polar alignment using SEQ
Solve frames
Using AS - load autopoint
Calculate 3-star config
Polar align
Correct slew via Az and Alt
If not cleared automatically by AS, manually clear (polar) config
Slew somewhere on same side of meridian
Take picture and solve
Sync solved picture (Maxim)
In AS: cross ckeck if config is cleared, then set new homeposition
In SEQ: 22-star pointing file
Solve frames
In AS: load auotopoint file
Calaculate config normally Ra < 0,2 arc sec, Dec dito
Save config for next start
Start SEQ run - MLPT enabled, dither 4 pix, grouped by slot
After SEQ finished, Telescope parked (near homeposition, counterweight shaft on E), then motors off, shut down all soft, shut down computer, axes fixed by inbus screws...

Second night (and following, so weather will):
Unfix screws in Ra and Dec
Switch on motors, Ra and Dec calibration
Homefind
Start SEQ run for desired object
99% perfect frames

Whenever I did an additional polar alignment with my config file used, it became crap...if I follow what I described above, everything fine...forgot to mention AS 5.2.1.9 and SEQ 2.0.9; latest AS version (5.2.2.3?) screwed whole system...! Therefore I am back on 5.2.1.9

Regards,
Christian

Thanks for your input - so you don't do a synch on a star after you have cleared the old config?
maybe thats what's causing my troubles?
other than that, i am using the AS version 5.2.2.3.

 

So here is your routine:

"First night:
Motor on
Homefind
Clear old config
3-star polar alignment using SEQ              - I do a synch on a near meridian star and i new home position before heading over to sequence.
Solve frames"                                             - All my solves failed, maybe because of my synch?

 

/Daniel



#13 Corpze

Corpze

    Member

  • Beta Tester
  • 207 posts
  • LocationSweden - Arvika

Posted 03 September 2016 - 09:20 PM

Ok... this is getting mroe and more worse - now i can't even connect to my mount - i am running win 10 and is getting constantly error 43 - USB device is not recognized by windows... I have tried to reinstall the older ftdi driver located under C:\programx86\Autoslew with no success...
No wonder autoslew/the mount have been behaving strange (if this is the problem)
anyone seen this behaviour before?

 

/Daniel



#14 Jay

Jay

    Member

  • Beta Tester
  • 125 posts
  • LocationFinland

Posted 05 September 2016 - 06:45 AM

Hi Daniel,

 

Don't know if this helps any but I've had my share of problems with the Windows 10, especially with COM / USB ports when installing drivers. Sometimes installing different drivers doesn't seem to help because the "old" drivers should be have been deleted first ?

 

I also used to have "USB error" issues but with good cables and better USB hub, they have mostly disappeared.  

 

After quite a lot of work I finally have everything running again (with the new laptop and Win 10), except can't still get the AAG Cloud sensor connected for some reason.  

 

So the problem might not be with the Autoslew drivers...

 

Jay


Edited by Jay, 05 September 2016 - 06:46 AM.


#15 Corpze

Corpze

    Member

  • Beta Tester
  • 207 posts
  • LocationSweden - Arvika

Posted 05 September 2016 - 04:09 PM

I backed down to win7, i was up and running within an hour, everything worked at once...
/Daniel

#16 ChristerS

ChristerS

    Member

  • Beta Tester
  • 87 posts
  • LocationFROSON, SWEDEN

Posted 06 September 2016 - 09:59 AM

Hi Daniel,

P.Keller fixed some FTDI windows 10 issues in the latest 5.2.2.3 release, make sure you run that version if you go back to Win10.

Works ok for me on Win10.

 

/Christer


Cheers,
//Christer

www.astropix.se


#17 Corpze

Corpze

    Member

  • Beta Tester
  • 207 posts
  • LocationSweden - Arvika

Posted 06 September 2016 - 10:46 AM

I tried to unistall and reinstall AS several times, i tried to reset all the com-ports and even unistall/reinstalled all the drivers again with no luck.
The pc i use, dell Optiplex 960 isn't supported for win10 either.
I will go for win7 now :-)

#18 maxmirot

maxmirot

    Member

  • Beta Tester
  • 221 posts

Posted 22 September 2016 - 02:43 PM

Seems we have a lot of complex issues with Windows 10.

 

I was working with WIN 10 using 5.1.1.9 and everything worked fine.  

 

I updated to the current version of Autoslew and clicked on replacing the USB driver for install. Then I had many USB errors.

 

I then installed the latest FTDI USB driver from the FTDI website. This fix the communication errors.

 

Max



#19 Corpze

Corpze

    Member

  • Beta Tester
  • 207 posts
  • LocationSweden - Arvika

Posted 22 September 2016 - 08:03 PM

Yea i had too many wierd errors, it seemed to be best to install Windows 7 again.
I have now 100% perfect 5min subs, even through zenit.

/Daniel

#20 maxmirot

maxmirot

    Member

  • Beta Tester
  • 221 posts

Posted 30 September 2016 - 02:54 PM

Did you fix this problem? Could it be an autoslew bug 5.2.2.3?

 

Max






0 user(s) are reading this topic

0 members, 0 guests, 0 anonymous users