DIY - Automated Polar Alignment

[stanlite (Grady)]

Well given my frustration with getting good polar alignment (more so the setup and pack up phases of the operation) and the limited spells of astro weather up here in Brisbane of late I have thought up a contraption i hope will semi automate polar alignment for my G11.

What I am planning to do is essentially use a Stepper motor/s attached to worm gears (replacing the alt and az knobs on the G11) or gearboxes to move my Alt and AZ axis during polar alignment.

I will run a check on my polar error with PemPro for example to find i am off by 100 arc seconds then slew the stepper motor for that axis say 100 steps and measure the change in error, with the change measured i should be able to work out the number of arc seconds to each step of the stepper to then be able to correct in one further iteration the remain polar misalignment.

I am going to be using an arduino board as the control interface between computer and motors.

I am however, wondering which gear method would be better worm or planetary?

PS. after a short search of the internet i have discovered a similar attempt however this person is attempting to direct drive the mount with the stepper. He is also using the mounts slew function to help correct polar misalignment visually where as I will just be correcting for measured error.

http://www.scopefocus.info/polar-alignment

[Nico (John)]

Grady

The G11 has a polar alignment error correction routine that is found through, I think, the setup menu. Before it is used however you have got to do a 3 star alignment. Do a few more 'syncs' on different stars then run the routine. It will require that you use the ALt and Az controls to bring the star back on then once thats done restart.

This is a similar rountine found in some standalone programs such as align master. Hope that helps.


Nico

[stanlite (Grady)]

I know the routines your are talking about ... i am attempting to remove the error of the human hand from the equation and make it a mechanical adjustment run via a computer to a much higher resolution than woulld be possible by hand.

The worm gear i am looking at coupled with the step size of the motor should give a resolution of about .5 arc second per step in the alt axis and 0.3 arc seconds per step in the az axis.

I could simply work out the error correction (say 7.5 arc minutes in Alt) needed and tell the stepper motors to move the 900 steps to get the alt axis aligned.

[Nico (John)]

I suppose you could obtain the error through the Gemini controller and apply the alt/az offsets, can't get any more accurate than that given that the algnment is based upn the eyeball anyhow.

[stanlite (Grady)]

i would be using drift alignment ... this is meant to be an improved and faster way of drift aligning

[lazjen (Chris)]

I have always wondered why someone hasn't done this before (cost, I assume) - the idea seems obvious/great to me and I hope you succeed. And if you do I'd be tempted to apply it to my G11.

[AndrewJ]

Gday Grady
I started doing something like this several years back but gave up due to lack of enthusiasm
I had the Alt axis of my wedge motorised and it works very well.
The AZ is also easy to motorise but i never got round to it.
I was also just going to drift align via PC and tweak the motors to reduce it as reqd.
Note, my method only works due to the special design of my wedge, which was done more for ease of loading than motorising.

Andrew

[stanlite (Grady)]

Well i got the stepper motors yesterday and the ardunio board and easyboard (stepper controller) and worm gears today. So I have soldered everything up and i am currently using the SGL focus control to test the motor everything is working well.

On the worm gear side of things i went with plastic Worms and Gears - 24 Diametrical Pitch (Series 1). a 20 tooth gear for the AZ axis and a 50 tooth Gear for the Alt axis. as is these gears will fit right in place of the existing control knobs so just need to thread them right and they should screw right into place.

no need for adapters for the gears.

The motors are another story, one of them came nonfunctional out of the box so i am going to email the supplier to see if i can get a refund/new one. The other works perfectly so i might try to design an adapter that will let it site 90degrees t the worm gears using the existing losmandy bolt holes. this might also allow me to use one engine and shift it between axis's if i can't get a new one.

so far total bill stands as so

Ardunio board from JayCar - $30
Easycontrol board - $14
Stepper Motor (only counting the functioning one atm) - $23
Plastic housing = $8
Worm and Worm Gears = $32

Total: $107

I estimate the cost of fabrication for the adapter for the motor will probably be in the $40 range. So might come in under estimated budget of $200

[RobF (Rob)]

Grady, I've been polar aligning for a couple of years with a pretty rough VB.NET program that monitors Y drift error reporting from Maxim. I've "calibrated" the rate of drift very roughly into quarter turns of my az knobs - no need to change alt as I set up in the same spot outside and rarely fiddle with it.

I'm sure you should be able to home in pretty quick with what you're planning. Even just a minute or so of average drift gives you a pretty good idea which way to go. Of course you need longer monitoring windows as you get closer - assuming you really do want it spot on. For my NEQ6 I generally fine a tad of error so its only correcting one direction in Dec is helpful.

[allan gould]

Grady
Perhaps you are over compensating as the losmandy manual states that a full turn of the Altitude knob gives a shift of 3.2 degrees while the azimuth gives 0.8 degrees for a full turn. It should be easy to plate solve and find out how much you are from the pole at CW down and parked position and then turn the knobs by a known amount and then take another plate solve to show how it's affected your pole position and if you have turned in the right direction.
Or am I missing something?
Allan

[DavidTrap (David)]

What about taking the output from PoleAlignMax and feeding that into your stepper motor controller?

DT

[stanlite (Grady)]

Hey Allan

I know about the plate solving option. What putting a stepper motor on the mount will allow me to do is make precise adjustments from my computer without having to touch the mount.

Originally Posted by allan gould

Grady
Perhaps you are over compensating as the losmandy manual states that a full turn of the Altitude knob gives a shift of 3.2 degrees while the azimuth gives 0.8 degrees for a full turn. It should be easy to plate solve and find out how much you are from the pole at CW down and parked position and then turn the knobs by a known amount and then take another plate solve to show how it's affected your pole position and if you have turned in the right direction.
Or am I missing something?
Allan

[stanlite (Grady)]

For the moment I find it easier to make the calculations manually and just use the focuser program. I have not yet got my head around C++ lol

Originally Posted by DavidTrap

What about taking the output from PoleAlignMax and feeding that into your stepper motor controller?

DT

[rogerg (Roger)]

I look forward to monitoring your success with this. It has astounded me for many years that in the very least the big players Meade and Celestron don't have automated polar alignment built in. In principal it seemed easy enough 5 years ago but seems no one (or very few) have had the time or skills to do it.

Good luck

[stanlite (Grady)]

Update:

Droped the alt knob off to be duplicated today and attached the worm gear to the azimuth axis

as you can see i have tried to keep it simple and relatively compact ... there is enough room for the stepper to reach the gear without impinging upon the mount itself.

I bored into the alt knob instead of my original idea of attaching the worm gear directly to the azimuth screw because the worm is just ever to big to fit. This i think will be better in the long run for the extra room it grants plus it means i don't have to thread the plastic and can still use the knob if the worm is not engaged.

[stanlite (Grady)]

well today marks another step forth in my project and thankfully is one filled with successes.

I have managed to get the Azimuth axis setup and functioning below are pictures. I ran it through the full 10 degrees in both directions and apart from a few tune ups to mesh the gears correctly i had no issues, at least without the scope setup.

From my eye test today the 20:1 gear ratio and motor result in a 200000 step run from hard lock to hard lock on the axis which results in a resolution of 0.153 arc seconds a step.

While i am waiting for a custom part for the alt axis (due to the fact the knob has seized on requiring a replacement) I am now unsure if the motors i got will be powerful enough to move lift the alt axis they will be fine to lower it however. If this is the case i am not sure i will bother continuing with that axis as someone has said before you can set it once and it is good.

[stanlite (Grady)]

more photo's

[h0ughy (David)]

love your work Grady - but what about a suggestion - gear the worm so that it can get some torque

[stanlite (Grady)]

The motor has .5 NM of torque already and the worm is a 20:1 ratio worm as it is ... I have found it has no problem on the azimuth axis Huff. I was originally thinking of gearing it yes but decided to test it direct drive. If I have issues with it fully loaded with my gear i will revisit this option. As it stands though the less complicated i can make it the better.

My concern with the Alt axis is that it is much stiffer then the az axis which even when i have it loaded up seems to have a very light touch.

problem with gearing also is it takes longer to make adjustments. I timed it at 4 mins to move a full 8.5 degrees of of travel in the az axis ablit this was at a reduced speed to avoid missing steps.

[rustigsmed (Russell)]

i guess if the alt can only lower, that should be alright if it starts off higher?