Jump to content


Photo

Starlock Trained PEC

  • Please log in to reply
10 replies to this topic

#1 galaxy_jason

galaxy_jason

    Vendor

  • -----
  • Vendors
  • Posts: 258
  • Joined: 22 May 2007
  • Loc: Texas

Posted 20 April 2013 - 09:29 AM

Beta testing the latest LX850 code on my production sample. The Starlock PEC training has been completely re-written and is looking very good. This code should be released soon. Attached is the raw and Starlock trained (guiding off of course). One train and three updates. I also downloaded the curve and applied one click of "smooth" with PEC tool. I think it helped a bit but probably not required.

Attached Files



#2 ahopp

ahopp

    Messenger

  • -----
  • Posts: 402
  • Joined: 24 May 2012

Posted 20 April 2013 - 09:52 AM

Jason, How do you do your balance? I have been doing it by feel, I wonder if I can improve balance by using a digital scale/pressure meter and use it when applying pressure on each side of the mount.

Tony

#3 galaxy_jason

galaxy_jason

    Vendor

  • -----
  • Vendors
  • Posts: 258
  • Joined: 22 May 2007
  • Loc: Texas

Posted 20 April 2013 - 10:00 AM

My only criticism of the mount, the unlocked clutch is s little tight but engineers tell me its a trade off in the design. RA can be balanced by feel because of the moment arm of the CW shaft. But to get it perfect I use a little postal scale to push up or down on the CW shaft or OTA.

#4 OzAndrewJ

OzAndrewJ

    Viking 1

  • *****
  • Posts: 951
  • Joined: 30 Nov 2010

Posted 20 April 2013 - 05:26 PM

Gday Jason

Is the new code still showing any residual drift after the post train processing????
I know the 1.1b code has reverted to the old normalising that results in drift, and your PEMPro plot doesnt indicate if it is displayed with drift removed.

Andrew Johansen Melbourne Australia

#5 galaxy_jason

galaxy_jason

    Vendor

  • -----
  • Vendors
  • Posts: 258
  • Joined: 22 May 2007
  • Loc: Texas

Posted 20 April 2013 - 06:25 PM

I think so the pec table was not being normalized. It now totals to sidereal

#6 OzAndrewJ

OzAndrewJ

    Viking 1

  • *****
  • Posts: 951
  • Joined: 30 Nov 2010

Posted 20 April 2013 - 07:06 PM

Gday Jason

I think so the pec table was not being normalized. It now totals to sidereal



Not sure i understand here.
The data has to be normalised after training.
Originally, this was done by simply making the positive raw PEC deltas match the negative ones. Due to the way PEC is applied, this results in drift, esp if you have a lot of largish PEC deltas.
In the later LX200 firmwares, this normalising got changed to do it based on elapsed time, not raw data. This method results in virtually no drift.

1.1b is still using the old method, so i was just wondering if that had got changed.
I would need to see the 60 bytes of raw data stored in the scope ( or the new code ) to see
if it has now changed.

Andrew Johansen Melbourne Australia

#7 galaxy_jason

galaxy_jason

    Vendor

  • -----
  • Vendors
  • Posts: 258
  • Joined: 22 May 2007
  • Loc: Texas

Posted 20 April 2013 - 08:00 PM

The developer told me the normalizing code was taken out and we were seeing drift. He put it back in. Maybe he meant the newer code had been taken out. And he mentioned something about a bug that had been there since the early days of the LX200 that had been fixed so maybe that is what you are talking about.

I don't get that much into the details, I just let them know what works and what doesn't. :)

#8 OzAndrewJ

OzAndrewJ

    Viking 1

  • *****
  • Posts: 951
  • Joined: 30 Nov 2010

Posted 20 April 2013 - 08:40 PM

Gday Jason

And he mentioned something about a bug that had been there since the early days of the LX200 that had been fixed so maybe that is what you are talking about.



Oohh there's lots of bugs and oddball processes re PEC
that have been there for a long while,
( esp if you are in the Sth Hemi :( )
but normalising wasn't one that is currently in the LX200 code.

I guess I'll just wait for the new firmware to come out in paperback :gotpopcorn:

Andrew Johansen Melbourne Australia

#9 Qwickdraw

Qwickdraw

    Surveyor 1

  • *****
  • Posts: 1717
  • Joined: 03 Mar 2012
  • Loc: Ann Arbor, MI

Posted 27 April 2013 - 05:47 AM

I am planning on purchasing the 14" in July. Does the LX850 allow for downloading of new code or will I be stuck with whatever rev it comes with?

#10 ahopp

ahopp

    Messenger

  • -----
  • Posts: 402
  • Joined: 24 May 2012

Posted 27 April 2013 - 05:58 AM

You can update the firmware on both the mount and the starlock system.

Tony

#11 Qwickdraw

Qwickdraw

    Surveyor 1

  • *****
  • Posts: 1717
  • Joined: 03 Mar 2012
  • Loc: Ann Arbor, MI

Posted 27 April 2013 - 06:44 AM

You can update the firmware on both the mount and the starlock system.

Tony


Nice, thx Tony






Cloudy Nights LLC
Cloudy Nights Sponsor: Astronomics