Jump to content


Photo

Trouble with Autostar GOTO

  • Please log in to reply
33 replies to this topic

#1 ricknau

ricknau

    Sputnik

  • -----
  • Posts: 33
  • Joined: 26 Feb 2011

Posted 16 March 2013 - 03:04 PM

I have mounted Meade DS motors on my GEM. After doing the preliminary steps to calibrate and train, I tried to do the "easy" align process from the AutoStar menu. When the scope kicked into gear to drive itself to Sirius it moved very far in both axes but stopped nowhere near Sirius. (off by 20-30 degrees) I checked my settings and couldn't see anything wrong. When I begin the easy align process the heading displayed on the AutoStar says "German Polar" so I should be in the correct mode. I want to make sure I correctly understand the process which scrolls by like a tickertape on the AutoStar.

ASSUMING THE MOUNT IS VERY CAREFULLY POLAR ALIGNED (WHICH MINE IS), as I understand it all I should have to do is rotate the RA until the tube is directly above the mount. Then rotate the declination axis until the scope is pointed dead on the polar axis. I presume this is the "Polar Home Position" referred to in the AutoStar scrolling instructions. (The Instruction Manual says nothing about Polar Home Position despite what the rolling instrucutons say.) Then I just press ENTER and the AutoStar should drive to where it "knows" Sirius is, based on my location, date, and time of day.

One other curious quirk I've encountered that might be related is that when the AutoStar kicks into sidereal tracking mode it does not track very accurately. The stars creep across the .67 FOV in a minute or two. It makes me wonder if I had the motors plugged into the wrong jacks. (the ratios are very close) I feel certain that I didn't but the bad sidereal track makes me question myself now.

Let me give some other detail in anticipation of questons.

TRAINING
I spent a lot of time learning and executing the training of the AutoStar. Did "calibration" first then traning. I was able to nail the landmark quite consistantly with backlash numbers of ~400 dec, ~700 RA. I will work to improve the RA but this cannot be the cause of my gross GOTO problem.

POLAR ALIGNMENT
Was done very carefully. It's a permanent mount that I fine tuned. But even if it was slightly off it shouldn't produce the giant error I saw.

Things I've double checked.

Scope Model set to 4504 (Dick Seymours patch 43gg28 for GEM mounts)
Az/RA ratio set to 2.19022 for 96 tooth worm set
Alt/Dec ratio set to 2.28148 for 100 tooth worm set
Site (used zip code)
Date was good
Time was good to the second
daylight saving time YES

Something I didn't do was to see if the error was in one axis or the other or both. (arrrgh!)

I must be missing something really big to be off by 20-30 deg but I don't know what. Perhaps one of the drive ratio numbers should be negative?

Any ideas out there?

#2 Cliff Hipsher

Cliff Hipsher

    Apollo

  • ***--
  • Posts: 1147
  • Joined: 31 Dec 2008
  • Loc: North Chesterfield, VA

Posted 16 March 2013 - 03:41 PM

First this is we need to know which brand and model of GEM you have.

Second, do a drift alignment. This is THE MOST ACCURATE method to polar align.

Third, GoTo accuracy and Polar alignment have NOTHING IN COMMON. With the exception of a One Star Alignment, all Autostar does is generate an electronic "map" of the local sky. A 3 star alignment has the added benefit of reducing cone error.

Fourth, using you ZIP code is not accurate enough. Use your actual Latitude and Longitude. Most smart phones have that functionality. If you want to get real accurate you can get one of these: GPS and then use GPS2Astar (available in the files section here: Meade AutoStar on Yahoo ).

#3 OzAndrewJ

OzAndrewJ

    Apollo

  • *****
  • Posts: 1006
  • Joined: 30 Nov 2010

Posted 16 March 2013 - 03:51 PM

Gday Rick

It certainly sounds like a ratio problem.

Scope Model set to 4504 (Dick Seymours patch 43gg28 for GEM mounts)
Az/RA ratio set to 2.19022 for 96 tooth worm set
Alt/Dec ratio set to 2.28148 for 100 tooth worm set



The ratios also take into account the drive/transfer gearing as well.
Essentially, the ratios are merely
how many quadrature encoder transition represent 1 arcsec of motion at the output axle.

Edit.
Deleted improper test.
Test specified wont work in a GEM


Andrew Johansen Melbourne Australia

PS as this topic is very specific,
can i also suggest you visit the Roboscope site
http://groups.yahoo....roup/RoboScope/
as that is a site dedicated to this sort of retrofitting

#4 ricknau

ricknau

    Sputnik

  • -----
  • Posts: 33
  • Joined: 26 Feb 2011

Posted 16 March 2013 - 04:37 PM

First this is we need to know which brand and model of GEM you have.

Third, GoTo accuracy and Polar alignment have NOTHING IN COMMON. With the exception of a One Star Alignment, all Autostar does is generate an electronic "map" of the local sky.

Fourth, using you ZIP code is not accurate enough.


1) Homemade

3) Understood, but if the alignment is way off the easy align will have no chance of locating the first alignment star. (which apparently is Sirius) I should have not clouded my problem discription with comments of polar alignment.

4) Not looking for killer accuracy right now. Just trying to get it work. I'll try your suggestions once I get there.

It certainly sounds like a ratio problem.

The ratios also take into account the drive/transfer gearing as well.

A simple test if you have it right ( magnitude wise )
is to point your scope at a landmark ( with DEC = 0 )
( like you would for drive training )
PressnHold mode to bring up the RA/DEC display.
Note the RA.
Now slew 360deg in RA using the Hbx till you get back to the landmark.
Try not to reverse, just slow down before you get there.
Once realigned, does the Hbx Read the same RA??
Now repeat for DEC.

If the numbers arent real close before and after,
the problem is in the encoders or ratios,
either misreading or wrong.

Andrew Johansen Melbourne Australia

PS as this topic is very specific,
can i also suggest you visit the Roboscope site
http://groups.yahoo....roup/RoboScope/
as that is a site dedicated to this sort of retrofitting


The drives are direct coupled, DS shaft to worm gear shaft so I think the tooth number itself should be what I use for my ratios. Am I correct in this? I got the numbers from here:

http://bedair.org/Autostar.html

I will try the test you recommend tonight.

Thanks to you both.

#5 Cliff Hipsher

Cliff Hipsher

    Apollo

  • ***--
  • Posts: 1147
  • Joined: 31 Dec 2008
  • Loc: North Chesterfield, VA

Posted 16 March 2013 - 05:04 PM

After you get your rough polar alignment done, go into Autostar and do a One Star Alignment. This is what Meade calls a Polar Alignment because you adjust the mount in lat and AZ to get Polaris centered in the eyepiece. When you hit Enter, Autostar will go to a second star and you use the hbx to center it. When that is done, do a 3 star...

#6 OzAndrewJ

OzAndrewJ

    Apollo

  • *****
  • Posts: 1006
  • Joined: 30 Nov 2010

Posted 16 March 2013 - 05:06 PM

Gday Rick

The drives are direct coupled, DS shaft to worm grear shaft so I think the tooth number itself should be what I use for my retios.



Yep, if direct coupled, the ratio is simply a function of the the encoder vane count, gearbox ratio and wormwheel toothcount.
I havent read Steves site recntly, as i always calc ratios from first principles. Will see what i get my way.
Edit
Just did the numbers
and based on 96 and 100 tooth wormwheels and a std DS drive unit, the nos you got are correct.
The slow tracking in RA indicates the sign of that ratio is correct
but the ratio may be wrong.


PS, i wrote my earlier instructions pre coffee. :confused:
Even tho you have a GEM, the encoder test needs to be done in AltAz mode ( for the RA axis ), as the GEM will be tracking, and we need a stationary scope.
Not sure if you can tell the model you have selected its AltAz.
The DEC test can still be done if reqd on a star, but not on a landmark.


#7 ccs_hello

ccs_hello

    Fly Me to the Moon

  • *****
  • Posts: 6551
  • Joined: 03 Jul 2004

Posted 16 March 2013 - 05:20 PM

Ricknau,

Welcome to Cloudy Nights!

Just wish to make sure:

You are using a DS motor (not the square-ish all self-contained DH4 motor) with some sort of adapter kit (e.g. GOTO4all kit.)

If your motor is DS (with 22-T output gear) and direct-coupling to mount's axial, then your calculation is correct.

P.S. DH4 motor looks like the pictures shown in this very old CN thread.

Clear Skies!

ccs_hello

#8 ricknau

ricknau

    Sputnik

  • -----
  • Posts: 33
  • Joined: 26 Feb 2011

Posted 16 March 2013 - 06:20 PM

It is this kit with a autostar 497

http://bedair.org/492Kit.html

#9 ccs_hello

ccs_hello

    Fly Me to the Moon

  • *****
  • Posts: 6551
  • Joined: 03 Jul 2004

Posted 16 March 2013 - 06:36 PM

Yes. That's DS motor.
If that 22-tooth gear is direct-coupling with the mount axis, then your calculation is correct.
(Of course, if you disconnect the power and see the stars drift away slower, then you have the sign wrong.)

Since you identified GOTO and tracking error, I would suspect the encoder wheel inside the DS motor may have problems.
E.g., grease stuck in encoder slit, opto interrupter is not aligned with the encoder wheel.

I would also recommend you ask the question in Y! Roboscope group as well.

Clear Skies!

ccs_hello

#10 OzAndrewJ

OzAndrewJ

    Apollo

  • *****
  • Posts: 1006
  • Joined: 30 Nov 2010

Posted 16 March 2013 - 06:50 PM

Gday Rick

Just to do the 360deg encoder test with a GEM,
i checked a cheat method, which appears to work.
Simply tell the handbox the scope is an ETX-80
Set it to AltAz mode.
Reset the ratios to what you have calculated
Set the mode to terrestrial

Now set the OTA to DEC = 0 and use pressnhold
to bring up the menus and select the Alt/Az display.
Slew in Az/RA until you can see a target.
Note the AZ
Keep slewing in the same direction for 360deg in Az
until the target is recentred
Note the Az
Repeat for the Alt/DEC axis.
If the encoders are working correctly
you will get very close results.
The DS motor encoders have 36 vanes on them, so if one is blocked
you will get something like a 10deg error.
If you get something else, it involves a rethink.

Andrew Johansen Melbourne Australia

#11 ricknau

ricknau

    Sputnik

  • -----
  • Posts: 33
  • Joined: 26 Feb 2011

Posted 17 March 2013 - 01:39 AM

My first try at your test, Oz (is that what you go by?), came out a tiny bit off in the Az readout match. Like maybe a minute or two. I didn't write it down because I became puzzled by something else I observed.

Even though I am in "terrestrial" mode the Az number slowly increments even though the motor is not moving. It changes by 0.1 minutes about once every 3 seconds then once about every 7 seconds, repeatedly. In other words... change, wait 3, change, wait 7, change, wait 3, change, wait 7, etc.

I did a reset but it didn't change that behavior. This doesn't seem right and will never give me the same Az number doing the 360 deg test.

Also I just noticed that the Alt/dec number changes when I move only the Az/RA axis. There is some kind of crosstalk going on. I read somewhere about fixing noise issues with the DS motors. I'll have to re-find that.

???

P.S. ... Go Webber!

#12 OzAndrewJ

OzAndrewJ

    Apollo

  • *****
  • Posts: 1006
  • Joined: 30 Nov 2010

Posted 17 March 2013 - 02:24 AM

Gday Rick

Oz (is that what you go by?)


Andrew will do

Even though I am in "terrestrial" mode the Az number slowly increments even though the motor is not moving


How do you know the motor isnt moving????
Its not something you can normally "see"
unless the motor encoder is exposed.

That said, in terrestrial mode, the Az and Alt displays
"should" stay constant no matter what.
If you are configured for AltAz AND terrestrial
and the Az readout is changing, something weird is going on.

P.S. ... Go Webber


Good for advertising, but he never lives up to expectations.
So far, the race has just been a high speed game of chasey.
At least in the celebrity race, they got serious :jump:

Andrew Johansen Melbourne Australia

#13 ricknau

ricknau

    Sputnik

  • -----
  • Posts: 33
  • Joined: 26 Feb 2011

Posted 17 March 2013 - 09:44 AM

How do you know the motor isnt moving????
Its not something you can normally "see"
unless the motor encoder is exposed.


I can hear the motor when it moves at that rate (roughly sidereal). While the numbers are changing but before I give it the command to move the motor makes no sound whatsoever. Plus I carefully observed it for 10 or 15 minutes.

That, combined with the dec readout changing when it shouldn't, would seem to indicate I'm screwed. Heck, the dec numbers change while that motor is unplugged from the junction box!

I might need a new 497 controller. I just picked this one up last month from a classifid ad.

BTW here's a photo of my rig. I'm doing the test indoors. The RA is a modified tripod head with built in worm gear. I built the dec axis (and worm drive) from parts bought from McMaster-Carr.

Attached Files



#14 Jeff Lee

Jeff Lee

    Viking 1

  • -----
  • Posts: 727
  • Joined: 17 Sep 2006

Posted 17 March 2013 - 12:55 PM

Polar home is achieved by using your adjustment bolts not moving the axis by releasing the clutches.

#15 ricknau

ricknau

    Sputnik

  • -----
  • Posts: 33
  • Joined: 26 Feb 2011

Posted 17 March 2013 - 01:19 PM

Hi Jeff. Thanks for replying. What do you mean by "adjustment bolts"?

And what/where is polar home? The scrolling Autostar instructions say to set to "polar" home as described in the manual. The manual says it is with the OTA level and pointing North. (This would have the counterwieght shaft horizontal.) I think this may be where I have gone astray. I started with the OTA directly above the mount pointing at Polaris.

#16 jrcrilly

jrcrilly

    Refractor wienie no more

  • *****
  • Administrators
  • Posts: 33887
  • Joined: 30 Apr 2003
  • Loc: NE Ohio

Posted 17 March 2013 - 01:25 PM

The manual says it is with the OTA level and pointing North. (This would have the counterwieght shaft horizontal.)


That's correct for the LXD650/750, but those used a different controller. I've never seen an Autostar GEM rig that didn't use the position you are using. What model manual are you using?

#17 ricknau

ricknau

    Sputnik

  • -----
  • Posts: 33
  • Joined: 26 Feb 2011

Posted 17 March 2013 - 02:03 PM

Andrew,

I have errored. My Autostar has not gone bonkers. I am new to the Autostar. When I did the "pressnhold" per your instructions it displayed RA and dec even thought I has set it to Alt/Az mode. Not knowing better I just assumed these were the numbers you wanted me to monitor. But by pressing the down arrow I now see the Alt and Az display! Yea! The Alt AZ display does not have any creeping values. The RA dec display I had been using is clearly some kind of running clock that also changes with motor rotation.

Anyway... the results of the 360 test is that the start and stop values after one rotaion are same +/- about 5'. This is after repeating it 4 times. This is most likely the slop in my ability to match the point exactly. It can't explain the gross error the "Easy Align" process had in locating Sirius. I really think (hope) I was starting from the wrong polar home position. See above.

#18 ricknau

ricknau

    Sputnik

  • -----
  • Posts: 33
  • Joined: 26 Feb 2011

Posted 17 March 2013 - 02:26 PM

That's correct for the LXD650/750, but those used a different controller. I've never seen an Autostar GEM rig that didn't use the position you are using. What model manual are you using?


I've downloaded a few varieties of manual. One is a condensed version for use with "various models of Meade telescopes". As far as I know none of them will be pertinent to a GEM rig. That's why the Autostar had to be patched. So there's plenty of room for confusion.

EDIT
Sorry if I appear to be ignorant on Autostar features and capabilities as well as Meade telescope types. Your reply inspired me to dig (and think) deeper. So now I've read up more on a varietry of Meade scopes and can see that the LXD-55 is a GEM. And, just like you said, polar home for GEMS is above the mount. And the Autostar has it as a selection.

The bad news is that elimiates my process as being the problem. Andrew's suggested 360deg around encoder test has eliminated the ratios as being the problem. So what's left but a broken controller? Whaaaa! :bawling: :confused:

#19 OzAndrewJ

OzAndrewJ

    Apollo

  • *****
  • Posts: 1006
  • Joined: 30 Nov 2010

Posted 17 March 2013 - 04:56 PM

Gday Rick

When I did the "pressnhold" per your instructions it displayed RA and dec even thought I has set it to Alt/Az mode.....
But by pressing the down arrow I now see the Alt and Az display! Yea!



That will definitely give the symptoms you described. :)
As you have found, you need to use the RA/DEC display when tracking and the AltAz display when terrestrial.

The RA dec display I had been using is clearly some kind of running clock that also changes with motor rotation



Yes and no. RA is a datum fixed to the stars.
When the OTA is stationary, the stars "drift" across it, and hence the Autostar will keep track of where "in the sky" the scope is pointing ( using time ), as well as where it is relative to the ground.
In this scenario, the RA display will change ( but DEC should stay steady )
Both Alt and Az should stay constant.

Based on the fact your spin tests gave very close to 360deg results, we need to assume the ratios entered are correct in magnitude and the encoders are working.
Next thing is ratio sign.
Sirius is moderately close to the equator.
If your DEC ratio sign was wrong, it may explain the error you saw.
( ie was the error mainly in DEC ??? )
Try it with the DEC sign reversed.
To confirm the Az sign is easy.
Just start the scope and it will begin tracking
If it goes the right way, the sign is correct.

Andrew Johansen Melbourne Australia

PS If you grab a copy of the LXD75 manual from Meades site
it will show you what the expected polar home position is
and how to check tube alignment etc.
This manual is probably the closest to what you are doing,
so following it will be 99% applicable to your setup.

#20 ricknau

ricknau

    Sputnik

  • -----
  • Posts: 33
  • Joined: 26 Feb 2011

Posted 17 March 2013 - 05:56 PM

Andrew, see my EDIT just above. I just found the LXD55 manual. The RA does track in correct direction. I will try to reverse the sign of dec as soon as I can get a clear night. Fully overcast tonight.

Thanks for the advice. Same to all who are pitching in.

#21 OzAndrewJ

OzAndrewJ

    Apollo

  • *****
  • Posts: 1006
  • Joined: 30 Nov 2010

Posted 17 March 2013 - 06:10 PM

Gday Rick

I will try to reverse the sign of dec as soon as I can get a clear night. Fully overcast tonight.



Overcast is good :jump:
as you can waste time indoors with a beer and cheese whilst watching TV if reqd.
Use a planetarium to display the night sky for your location.
Adjust the time in the planetarium to find when Sirius is at the meridien.
( The actual date/time you get is immaterial, ie it doesnt need to be current. )
Now go to yr mount ( it can be set up anywhere and pointing anywhere )
Put the DEC head to polar home and start the Hbx.
When it asks for time, use the date/time you found in the planetarium.
Do a one star align.
Just hit enter when it asks to centre polaris.
The next star will be Sirius
When it slews to it and asks you to centre it,
your OTA should be pointing vertical and the DEC readout of the OTA
( from the setting circles ) should be the same as Sirius.

Lots of tests can be done this way without needing clear sky.

Andrew Johansen Melbourne Australia

#22 ricknau

ricknau

    Sputnik

  • -----
  • Posts: 33
  • Joined: 26 Feb 2011

Posted 18 March 2013 - 08:44 PM

Now go to yr mount ( it can be set up anywhere and pointing anywhere )


Except the bearings that hold my motorized head at the angle of my latitude (note the large shaft protruding to the right in my photo) are permanently mounted outside. :(

Not say that I couldn't design a similar experiment with the head sitting flat on some kind of stand.

But the beer and cheese is sounding more appealing right now! :D

#23 OzAndrewJ

OzAndrewJ

    Apollo

  • *****
  • Posts: 1006
  • Joined: 30 Nov 2010

Posted 18 March 2013 - 09:03 PM

Gday Rick

Except the bearings that hold my motorized head at the angle of my latitude (note the large shaft protruding to the right in my photo) are permanently mounted outside.



So take the beer and cheese outside on a tray :cool:
All i'm pointuing out is that for this level of diagnostic work,
you dont need a clear sky, or even the ability to see stars.
We are only trying to test if changing the DEC ratio sign
gets you into the ballpark.
Faking the time, to put a known target in a known position
makes it easy to see how close it is getting.

Andrew Johansen Melbourne Australia

#24 labmand

labmand

    Messenger

  • *****
  • Posts: 471
  • Joined: 08 Feb 2011
  • Loc: Michigan USA

Posted 19 March 2013 - 09:47 AM

Thanks Andrew, like the beer and cheese idea, I'm learning so much reading this thread. Good stuff.

#25 ricknau

ricknau

    Sputnik

  • -----
  • Posts: 33
  • Joined: 26 Feb 2011

Posted 19 March 2013 - 09:20 PM

All i'm pointing out is that for this level of diagnostic work,
you dont need a clear sky, or even the ability to see stars.
We are only trying to test if changing the DEC ratio sign
gets you into the ballpark.
Faking the time, to put a known target in a known position
makes it easy to see how close it is getting.


Ok, I'm slow but I get there! :lol: I must say that while I missed a chance 2 nights ago, last night was big rain. Tonight is good though and I've given it a go with the sign of Dec reversed. I did it without the OTA. That made it much easier to set up.

Here's what happened. After doing a motor calibration I set the mount at polar north and did Easy Align. It looked like the the RA fairly nailed Sirius' RA but the declination motor had the imaginary tube pointed roughly 160-170 deg out. I was encouraged by RA and decided to return the dec sign to +. Did easy align again. This time it pointed to roughly 2/3 of the way between Sirius and Procyon. Returned once again to polar north. Redid easy align. It pointed to the same place between Sirius And Procyon.

Things of note: At one point the Autostar crashed. Running off the battery pack with AA batts. Battery readout says batts at 100%

Work night so I couldn't keep beating at it. Off to bed now






Cloudy Nights LLC
Cloudy Nights Sponsor: Astronomics