Loading ...
Sorry, an error occurred while loading the content.

Re: SMT training (was Seeking ultra precise LX200GPS go-to)

Expand Messages
  • johansea
    Gday Pete ... Oooohhh goody. Can you explain it to me :-) ... A lot of help from Dick on how the code goes together, and then it was just a simple task to
    Message 1 of 34 , Aug 1, 2010
    • 0 Attachment
      Gday Pete

      --- In LX200GPS@yahoogroups.com, "barringtonri" <pecorp@...> wrote:

      > I'm starting to understand SMT now Andrew.

      Oooohhh goody. Can you explain it to me :-)

      > How you figured it out from scratch is beyond me.

      A lot of help from Dick on how the code goes together,
      and then it was just a simple task to decompile a few thousand lines of machine code and compare it to real world data.
      Quite simple really ;-) ( and much easier than sudoku )

      > I don't think that we've ever discussed SMT in such detail in
      > the Yahoo group. The concept is good but the implementation
      > seems to be lacking.

      That applies to a lot of Meade code. Lack of attention to detail has crippled many useful processes
      ( inc PEC, tours and a few other things )

      > If tonight's clear one more attempt will be made,

      I luv a glutton for punishment :-)
      After yr third attempt, it would be interesting if you posted the results of your three attempts, even if only to show how random the results are.

      Andrew Johansen Melbourne Australia
    • johansea
      Gday Pete ... If i do that, i just use back , and many times it restores the cached edit screen. ... been drinking in the dome huh???? ... It still shows
      Message 34 of 34 , Aug 1, 2010
      • 0 Attachment
        Gday Pete

        --- In LX200GPS@yahoogroups.com, "barringtonri" <pecorp@...> wrote:
        >
        > Don't you love it when after 30 minutes of careful writing
        > you hit a wrong key and Yahoo erases the whole message.

        If i do that, i just use "back",
        and many times it restores the cached edit screen.

        > Figuring that Meade must have had some confidence in SMT

        been drinking in the dome huh????

        > I doggedly attempted yet another SMT update.
        > The MyScope plot of the July 31st model isn't perfect,
        > but it may be workable:

        It still shows rather ragged results, and overlaps, but is certainly better than the first runs.
        However, the fact the three plots are so inconsistent tells me immediately that the data must be treated as suspect.

        > I agree with you Andrew. SMT modeling not what one would
        > call robust. MyScope contains the only SMT analysis routine
        > that I'm aware of. We Meade users have just been assuming
        > that SMT works first time every time, and trying to find
        > excuses as to why go-to operation is still a bit off.

        Part of the problem is the randomness of the data ( and this applies to PEC as well )
        Many users "fluke" good data, and then post there are no problems, and that everything works. Others never get anything to work and others get part good and part bad. Properly understanding how variable the data is, is paramount to understanding why some users get a good experience and others dont.

        Anyway, thanks for the nice datasets to highlight it.

        > There is obviously a reason that routines such as MaxPoint
        > and APC exist and I'll undoubtedly wind up using one of them.

        One BIG reason is they are on a PC. Ie they have huge data storage abilities and also very accurate floating point calcs.
        The Austostar uses limited storage sizes and improvised floating point emulation ( "single" precision using an 8bit processor ), so it cannot compete with a PC at the accuracy required for many of the trig functions involved.
        This comment doesnt remove the fact that the actual firmware isnt correct to begin with, just that even if it was correct, it is still limited to the accuracy of its data and processor.


        Andrew Johansen Melbourne Australia
      Your message has been successfully submitted and would be delivered to recipients shortly.