Page 1 of 4 123 ... LastLast
Results 1 to 10 of 35
Like Tree12Likes

Thread: Your LX90ACF can be super accurate!

  1. #1
    Piet Le Roux's Avatar
    Piet Le Roux is offline Main Sequence
    Points: 894, Level: 16
    Level completed: 94%, Points required for next Level: 6
    Overall activity: 4.0%
    Achievements:
    2 Posts Achievement20 Posts Achievement!50 Posts Achievement!5 Threads Achievement!365 Days+ Registered Achievement!
    Join Date
    Nov 2017
    Location
    Bloemfontein South Africa
    Posts
    76
    Points
    894
    Level
    16
    Thanks
    3
    Thanked 11x 9 Posts
    Downloads
    0
    Uploads
    0

    Default Your LX90ACF can be super accurate!



    This does not involve any physical modifications!

    For a long time the Az/RA and Alt/Dec ratio have been a bit of a mystery for me and something that should be left alone…someone even suggested that it should not be accessible in the menu. Now I think it’s a brilliant idea and not just does it make the system easily adaptable to different telescopes but also makes them easy to calibrate.

    First you need a solid working LX90 and to proof this we first have to preform a simple test (after a motor calibration and a “drive train” had been done) :

    Set the the scope up with its tube level (it would be wise to attach a small level with fine double-sided tape). Set all dials to “0” and do a mock “easy aliment” pressing “enter” every time it beeps. Now press “mode” until you see “select object” then select “utilities” and select “park scope” when it is finished doing that and all your dials are back at “0”, your scope is working fine. If this is not the case visit “Jan’s LX90 site” to see how to make adjustments to your Az and Alt drives to remove play in the worm gears and assembly.

    For the next step you must remove anything that would prevent your scope from rotating 360 degrees vertically, Like your spotting scope and 2” diagonal. Make sure that your OTA is perfectly level. Switch the scope on and keep on pressing “mode” until you get to “select object” press mode for 3-4 seconds until it displays the Ra and Dec values, press down arrow once and you will see Az and Alt values : all must be at zero. Now you do the “backflip” by pressing the up arrow until it has rotated 360 degrees, use a slow slew speed to get the OTA perfectly level again. Now note the Dec reading. In my case it was far out : about +4 degrees 43 minutes. This means that the scope would have stopped short 4.5 degrees in a 360 degree slew. And more than 0.5 degree in a 45 degree vertical slew…enough to mis an object completely! To eliminate error we do the same test but this time we do the “nosedive” by pressing the down arrow until the OTA have rotated 360 degrees again note the error . Convert the minute values to decimal by dividing it by 60 and lose the negative (-) . Ad the backflip and nosedive values and divide by two, mine came to 4.675 degrees.
    Under Setup you wil get “Telescope” press enter and use arrow until “Alt/Dec Ratio”, the default is +0.275075. This factor is dependent on the size of the spur gear : the LX90 has a gear with 120 teeth and its default ratio is 2.75075, a ETX has a gear with 60 teeth and its default ratio is 1.36889 because the worm gear has to make nearly dubbel the amount of turns for a 360 degree rotation using 120 teeth than for 60 teeth. So in my case the gear has to travel more than the 2.75075 ratio allows it, so the value has to be made bigger:
    (360+4.675)/360 * 2.75075 = 1.0129861111111 * 2.75075 = 2.786471 or 2.78647
    This is our new ratio: enter this in the place of the 2.75075 and press “enter” . Press enter again and check if it is the same value, if it is exit with “mode”, if the last digit have increased by one : enter a value of one less and press “enter”, check the value, if the value is correct exit with “mode”. Reboot.
    Run the test again to see the results : in my case the error over 360 degrees was now down to 10 minutes 34 seconds…. A great improvement from 4.675 degrees !

    The Az/RA Ratio has a “-“ value in the LX90 because the drive is mounted upside-down and turns in opposite directions. But its value can be modified in the same way : this time use the same terrestrial object you used to do drive training. Centre the object in your eyepiece and rotate 360 degrees clockwise and anticlockwise, record errors. If the scopes “odometer” went past 360 it would have stopped short and the value has to be increased, if it stops at less than 360 the value has to be made smaller: (360 - error)/360 * -2.75075

    I bought a brand new LX90ACF and was terribly disappointed with its goto performance. There is no technical support for Meade in South Africa and shipping it back to the USA was not an option for me. So I had to solve the problem.
    Last edited by Piet Le Roux; 04-04-2018 at 05:06 PM.
    E Sully, Rob357, Slabs1960 and 4 others like this.

  2. The Following 2 Users Say Thank You to Piet Le Roux For This Useful Post:

    jinzai (04-04-2018),Trombatissimo (04-09-2018)

  3. #2
    E Sully's Avatar
    E Sully is offline Bright Giants
    Points: 5,749, Level: 51
    Level completed: 99%, Points required for next Level: 1
    Overall activity: 1.0%
    Achievements:
    2 Posts Achievement20 Posts Achievement!5 Threads Achievement!365 Days+ Registered Achievement!750 Days+ Registered Achievement!
    Join Date
    Jun 2012
    Location
    New York
    Posts
    206
    Points
    5,749
    Level
    51
    Thanks
    3
    Thanked 47x 42 Posts
    Downloads
    0
    Uploads
    0

    Default Re: Your LX90ACF can be super accurate!

    Very interesting procedure.
    8" LX200 EMC Classic
    ETX-90EC

  4. #3
    OzAndrewJ's Avatar
    OzAndrewJ is offline White Dwarf
    Points: 4,609, Level: 46
    Level completed: 30%, Points required for next Level: 141
    Overall activity: 0%
    Achievements:
    2 Posts Achievement365 Days+ Registered Achievement!750 Days+ Registered Achievement!20 Posts Achievement!First 1000 Experience Points
    Join Date
    May 2012
    Location
    Australia
    Posts
    48
    Points
    4,609
    Level
    46
    Thanks
    0
    Thanked 13x 11 Posts
    Downloads
    0
    Uploads
    0

    Default Re: Your LX90ACF can be super accurate!

    Gday Piet
    Az/RA and Alt/Dec ratio have been a bit of a mystery for me
    There are lots of explanations of how it works on several sites.
    The ratio is a flexible value that allows the std Autostar firmware to drive any scope.
    It is simply the number of quadrature encoder ticks required to move the output axle through 1 arcsec
    The 360deg test you figured out for yourself is the std way to test if the encoder feedback is working correctly, but i have to note that as there are literally thousands of LX90s out there working with the default ratio, it doesnt need "calibrating".
    What i suspect is that like many before you, your encoder disk has a bit of grot on one line.
    ( The encoder disk in the LX90s are actually exposed to the atmosphere )
    Lets do the math

    // How to calc ratios
    // a) We need to calc how many ticks are required to get one turn of the worm = Full GbxReduction * Encoder vanes * 4
    // b) We need to get how many ticks are associated with one tooth of the worm wheel = 360deg/toothcount * HbxRatio

    // ( Vanes * 4 ) * GBxRatio * XferGearRatio = ( 1,296,000 / ToothCount ) * HbxRatio
    // ticks/revofmotor * revsofmotor/revsout * Outgear/InGear = arcsec/Tooth * ticks/arcsec

    // HbxRatio := Vanes * 4 * GbxRatio * ( Out / In ) * WormTeeth / 1,296,000
    // = Vanes * GbxRatio * ( Out / In ) * WormTeeth / 324,000 In = toothcount on GBxOut
    // Out = toothcount on Worm

    LX90s have a 154 tooth wormwheel, 108 vane encoder and gearbox ratio of 53.5859375:1 with 1:1 xfer gears
    Sooooo Ratio = 108 * 53.5859375 * 1 * 154 / 324,000 = 2.7507447916666666666666666666667 :-)
    Sooo lets look another way. What type of error would we expect for a damaged encoder line, and does it get close to what you got.
    In 360deg we have 1,296,000 * 2.75074479 = 3,564,965.25 encoder transitions
    with 108 vanes that is 33,009 encoder transitions per vane per 360 deg
    or 09d 10m 09s. ( ie 2d 17m 32s per transition, but the transitions arent always perfectly equally spaced )
    You got close as dammit 1/2 a full vane, so maybe you have grot on a vane that means one channel doesnt read on that vane
    If you have a logic analyser, you can grab the encoder output for one rev at say speed 3, and then look to see if there are any glitches where one or both channel misread.

    Andrew Johansen Melbourne Australia
    DEnc and jinzai like this.

  5. The Following User Says Thank You to OzAndrewJ For This Useful Post:

    jinzai (04-04-2018)

  6. #4
    jinzai's Avatar
    jinzai is offline Bright Giants
    Points: 3,302, Level: 37
    Level completed: 68%, Points required for next Level: 48
    Overall activity: 0%
    Achievements:
    2 Posts Achievement20 Posts Achievement!5 Threads Achievement!Gallery Achievement!50 Posts Achievement!
    Join Date
    Aug 2017
    Location
    Davenport, IA
    Posts
    390
    Points
    3,302
    Level
    37
    Thanks
    39
    Thanked 98x 87 Posts
    Downloads
    0
    Uploads
    0

    Default Re: Your LX90ACF can be super accurate!

    I am pretty sure that this applies to some of the ETX series, as well. Basically, anything that has the same Auto/Audiostar firmware. Perhaps the numbers would change owing to the possible gear tooth count differences, but -- the basic formula is the same.

    Thanks for that effort, at any rate. It is much appreciated here.
    Michael
    Meade ETX-125 Observer (127mm f/15) Maksutov-Cassegrain
    Explore Scientific ED80CF f/6 Apochromatic Refractor
    Celestron AVX
    QHY 30mm Mini Guidescope
    QHY 5II Guide camera
    QHY163C

  7. #5
    OzAndrewJ's Avatar
    OzAndrewJ is offline White Dwarf
    Points: 4,609, Level: 46
    Level completed: 30%, Points required for next Level: 141
    Overall activity: 0%
    Achievements:
    2 Posts Achievement365 Days+ Registered Achievement!750 Days+ Registered Achievement!20 Posts Achievement!First 1000 Experience Points
    Join Date
    May 2012
    Location
    Australia
    Posts
    48
    Points
    4,609
    Level
    46
    Thanks
    0
    Thanked 13x 11 Posts
    Downloads
    0
    Uploads
    0

    Default Re: Your LX90ACF can be super accurate!

    Gday Michael
    I am pretty sure that this applies to some of the ETX series, as well.
    Yep, ETX, DS, DSX, DS2000, LX75, LX80, LX90, LT, StarNavigator all have different gearing and ratios, but the handbox firmware is the same.
    The only current gotcha is the new ETX90s ( with connectors in fork arm ) use different ratios to the earlier ETX-90s but some firmwares dont know about this. ( To fix that with the new units, just select ETX-80 as the model and all will be sweet )
    That said, when testing for encoder problems, there are 2 possibilities
    a) Grot on the encoder
    b) Bad circuitry
    If you do the 360deg test in AZ ( or 90deg using a digital level in Alt ) and compare the known "mechanical move" to what the Hbx reports it "thinks" it has moved, you can determine if the encoder has a problem.
    If you start in a known spot with a known handbox reading, then slew out say 90-100deg then slew back past target then back to target, it removes any backlash errors from the numbers. If the start and end positions report the same, then it is more likely a blocked encoder ( as it loses the same vane count on each slew ).
    If the results are very different, then its more likely the encoder circuit is misbehaving.
    Again, the best test for this is a logic analyser trace.

    Andrew Johansen Melbourne Australia
    jinzai likes this.

  8. #6
    Piet Le Roux's Avatar
    Piet Le Roux is offline Main Sequence
    Points: 894, Level: 16
    Level completed: 94%, Points required for next Level: 6
    Overall activity: 4.0%
    Achievements:
    2 Posts Achievement20 Posts Achievement!50 Posts Achievement!5 Threads Achievement!365 Days+ Registered Achievement!
    Join Date
    Nov 2017
    Location
    Bloemfontein South Africa
    Posts
    76
    Points
    894
    Level
    16
    Thanks
    3
    Thanked 11x 9 Posts
    Downloads
    0
    Uploads
    0

    Default Re: Your LX90ACF can be super accurate!

    My LX90 was brand new out of the box but made in China. So I suspect that the machining is not as accurate as on earlier productions. The first test is to check if the encoder is working well and if it can return accurately to its start position after a 2 star calibration I am satisfied with that. If the AZ and Alt 360 degree tests does show a error less than a degree its not worth changing the default. I think the default takes in account machining errors and wear, so it may get closer to the default as the gears start to seat better. But Andrew I tried to keep to the KISS principle and now you have made it much more complicated !

  9. #7
    OzAndrewJ's Avatar
    OzAndrewJ is offline White Dwarf
    Points: 4,609, Level: 46
    Level completed: 30%, Points required for next Level: 141
    Overall activity: 0%
    Achievements:
    2 Posts Achievement365 Days+ Registered Achievement!750 Days+ Registered Achievement!20 Posts Achievement!First 1000 Experience Points
    Join Date
    May 2012
    Location
    Australia
    Posts
    48
    Points
    4,609
    Level
    46
    Thanks
    0
    Thanked 13x 11 Posts
    Downloads
    0
    Uploads
    0

    Default Re: Your LX90ACF can be super accurate!

    Gday Piet
    The first test is to check if the encoder is working well and if it can return accurately to its start position
    Yep, but all that proves is the encoder is working or its disc is evenly damaged :-)
    The Normal test in Az is to start slewing until you hit a target ( so gears are preloaded ). Note the reported Az
    Now keep slewing in the same direction until you hit the target again.
    If it doesnt report the same Az, there is a problem, even if it can slew out and back accurately.
    Alt is harder as you cant do a full 360 spin, so thats why people normally use a digital level.

    I think the default takes in account machining errors and wear,
    Nope.
    The ratio is a digital calculation based on gear teeth and encoder vanes.
    No amount of running in will change it :-)
    now you have made it much more complicated !
    Sorry bout that

    Andrew Johansen Melbourne Australia

  10. #8
    Piet Le Roux's Avatar
    Piet Le Roux is offline Main Sequence
    Points: 894, Level: 16
    Level completed: 94%, Points required for next Level: 6
    Overall activity: 4.0%
    Achievements:
    2 Posts Achievement20 Posts Achievement!50 Posts Achievement!5 Threads Achievement!365 Days+ Registered Achievement!
    Join Date
    Nov 2017
    Location
    Bloemfontein South Africa
    Posts
    76
    Points
    894
    Level
    16
    Thanks
    3
    Thanked 11x 9 Posts
    Downloads
    0
    Uploads
    0

    Default Re: Your LX90ACF can be super accurate!

    Having done electronics, digital electronics and metal machining I cannot totally agree but thanks for your input

  11. #9
    OzAndrewJ's Avatar
    OzAndrewJ is offline White Dwarf
    Points: 4,609, Level: 46
    Level completed: 30%, Points required for next Level: 141
    Overall activity: 0%
    Achievements:
    2 Posts Achievement365 Days+ Registered Achievement!750 Days+ Registered Achievement!20 Posts Achievement!First 1000 Experience Points
    Join Date
    May 2012
    Location
    Australia
    Posts
    48
    Points
    4,609
    Level
    46
    Thanks
    0
    Thanked 13x 11 Posts
    Downloads
    0
    Uploads
    0

    Default Re: Your LX90ACF can be super accurate!

    Gday Piet
    Having done electronics, digital electronics and metal machining
    Me too, plus over 15 years working with Meades firmware and hardware on a range of scopes.
    There is no way machining quality can affect the specific ratio required for a given scope.
    Sure slop in the assy can affect pointing due to flexure etc
    but it has zero affect on the encoders and ratios, ( unless some swarf gets stuck to the encoder disc )

    Andrew Johansen Melbourne Australia

  12. #10
    Piet Le Roux's Avatar
    Piet Le Roux is offline Main Sequence
    Points: 894, Level: 16
    Level completed: 94%, Points required for next Level: 6
    Overall activity: 4.0%
    Achievements:
    2 Posts Achievement20 Posts Achievement!50 Posts Achievement!5 Threads Achievement!365 Days+ Registered Achievement!
    Join Date
    Nov 2017
    Location
    Bloemfontein South Africa
    Posts
    76
    Points
    894
    Level
    16
    Thanks
    3
    Thanked 11x 9 Posts
    Downloads
    0
    Uploads
    0

    Default Re: Your LX90ACF can be super accurate!

    Ok lets look at my 4.675 degree error : 4.675/360 *100 =+-1,3% error, there is 36 gaps in a decoder wheel for a etx, thus 36 pulses every motor revolution if just one gap was blocked it would result in a 1/36 * 100 = 2.8% error unless the LX90 has a 72 gap wheel that means the digital error could be 1.4%... then I guess its possible!? it also means I would then have to open the drive that I don't want to do! But then again if it was missing a pulse it would over shoot every time when slewing and that's not the case!

    PS I am 56 .... more than 35 years on the job!
    Last edited by Piet Le Roux; 04-05-2018 at 04:25 AM.

 

 
Page 1 of 4 123 ... LastLast

Similar Threads

  1. Meade LX90ACF 8" Motor Unit Failure. RA drive controller board
    By AWOL in forum Meade Telescope Forums
    Replies: 7
    Last Post: 03-12-2016, 04:01 PM
  2. Replies: 0
    Last Post: 11-03-2012, 04:13 AM
  3. Meade 8" LX90ACF
    By oko77 in forum Astronomy Beginners Forum
    Replies: 8
    Last Post: 09-06-2012, 12:54 AM
  4. Follow the Sun with LX90ACF
    By Harkhoost in forum Meade Telescope Forums
    Replies: 6
    Last Post: 10-28-2010, 09:25 PM
  5. Super-accurate Mars distance
    By Mac Roberts in forum Amateur Astronomy Forum
    Replies: 0
    Last Post: 08-14-2003, 06:15 PM

Posting Permissions

  • You may not post new threads
  • You may not post replies
  • You may not post attachments
  • You may not edit your posts
  •  
Powered by vBulletin® Version 4.2.0
Powered by vBulletin®
All times are GMT. The time now is 03:14 AM.