Go Back   IceInSpace > Equipment > ATM and DIY Projects

Reply
 
Thread Tools Rate Thread
  #1  
Old 30-09-2020, 10:25 AM
Greggles (Gregory)
Registered User

Greggles is offline
 
Join Date: Aug 2017
Location: Sydney, Australia
Posts: 75
Stepper motor upgrade - EQ Platform

Hi every one,

Just want to share my project to upgrade the motor on my 10"dobsonian DIY eq Platform from a EQ1 DC motor to a stepper motor run by arduino.

Firstly I would like to say a big thank you to Rod and Bojan who helped me out, I couldn't of done it without your support and encouragement.

I was always so scared to move to a stepper motor as I felt overwhelmed and confused trying to figure out all the bits and bobs to get one to work for an EQ platform. The way I learnt how to do it (besides all the help I got) was to take a step back and learn Arduino and then learn about steppers and then put it all together to work with an EQ platform. The code I wrote started as a switch turning on a led and I kept developing it to eventually work for my platform.

Functionality:

When turned on the motor will give a 2 second warning beep then the motor will proceed at a faster speed in reverse until the push button is triggered when the platform reaches the starting position. The platform will then move forward at a faster speed for 700 steps (eliminate the push button being activated again due to slow tracking speed) then begin to track at normal until it reaches the end where another button is pushed putting it in reverse and repeating the process.

I know some calculate actual amounts of rotation to signify when a reset needs to be done eliminating switches but I chose to go with switches as I like the ability to reset early or begin tracking at any point by reaching down and hitting the button.


Parts:

So the motor im using is a nema17 stepper motor with a 27:1 gear box Big thank you to Rod who shipped this out to me. What a legend. It is run by a DRV8825 stepper motor driver forced into fast decay mode in order to help with missed steps, an Arduino Nano, 12v power supply, buck converter, acouple of push buttons and a buzzer.

Putting it all together:

I first developed everything on a breadboard to ensure it was working and work on the code. Once I got it all I working I needed to move it to a more permanent board. Being my first electronics project I ended up going with a solder able breadboard as to reduce the possibility of me making any errors compared to strip/proto board.

I then soldered everything onto the board and mounted it to the platform. I connected the two push buttons onto the edges of the platform so they were triggered when the platform was at the start and end positions. And connected the motor as a direct drive with a coupling designed for CNC routers.

Motor powered by a 12v power supply. And board powered by the same power supply reduced to 5v through the buck converter.

If any one is interested in the Arduino sketch/code i'm happy to post this too.

Ready to go I connected it all together hooked it up and…. Oops. Something wasn't running right. The motor seemed to be slowing down and speeding up. Increasing the amps to the motor seemed to help but but I was running and A4498 at this stage rated for only 1A so I swapped to a DRV8825 at 1.5A. Unfortunately the problem persisted, seemingly worse. The motor was running at 1/16 microsteps and the motor was quite loud and when the speed issues where happening you could hear when it go louder and then subsequently quieter. Changing to 1/32 microsteps helped significantly, but it does seem I have some slight speeding up and slowing down at times. Not sure if this is PE or skipping steps or what. I have ordered a S109 driver just in case it helps with skipping steps as DRV8825 doesnt have the best reputation. If its PE there's a high precision 30:1 stepper motor on sale at the moment that’s looking inticing Not sure if this would help?

Alternately, as per Bojans advice, I could look into overhauling the whole drive system with belt drives etc.

Getting the perfect speed wasn't as easy as I though. The speed calculated as per my dimension was pretty close but not perfect. Using an eliminated reticent eyepiece with a 2x barlow I began making tweaks to the time between steps to get it as perfect as I could, this took for ever as often I would have to wait 10+ minutes to see any deviation. The other issue I was having was being unsure if some deviation was due to the PE, missing steps or platform slipping or not.

I’ve got the speed quite perfect when half way through tracking but from the start position the platform seemed to be tracking a little fast. Very frustrating. But not much I can do about it besides possibly get new more accurate sectors or redeveloping the platform.

Over all the change to a stepper motor is great. The precision is leaps and bounds over the DC motor I have.

- Once calibrated I don’t have to spend 30 minutes every single session fiddling with the potentiometer of the DC motor trying to get it right.
- will confirm tonight but with a 2X barlow and 12.5mm eyepiece star will stay in field of view for 45 mins+.
- kept approximately centre with deviations back and forth for approx 15+ minutes.
- dslr with 1.5x barlow exposure time moved from 4 seconds to approx 8 seconds for relatively consistent photos. 10 seconds which was my goal seems to be just too long and picks up the errors in the drive more often.

Pretty ordinary weather here in Sydney but once I get some deep space shots with the new motor I will post them here as well as on my Instagram @astrogreggles

Hope this was helpful/interesting. Please let me know if you have any questions and comments.

Cheers Greg
Attached Thumbnails
Click for full-size image (eq1.jpg)
182.8 KB81 views
Click for full-size image (eq2.jpg)
193.7 KB67 views
Click for full-size image (eq3.jpg)
203.2 KB59 views
Reply With Quote
  #2  
Old 30-09-2020, 11:53 AM
bojan's Avatar
bojan
amateur

bojan is online now
 
Join Date: Jul 2006
Location: Mt Waverley, VIC
Posts: 6,927
Good job

BTW, Arduino Nano is using low precision crystal (>100ppm) for clock (and every other timing), 10pm would be more appropriate (for AP... but it wouldn't hurt for visual as well).
Of course, this is way better than free running DC motor you had before.
Deviation period of 15 minutes you are mentioning suggests PE (in gearbox probably.. what is the rotational rate of motor/gearbox shaft?).
Reply With Quote
  #3  
Old 30-09-2020, 12:56 PM
Greggles (Gregory)
Registered User

Greggles is offline
 
Join Date: Aug 2017
Location: Sydney, Australia
Posts: 75
Hey Bojan,

Ah I didn't know that about the Nano.

200 step motor x 1:26.85 gearbox * 32 microsteps = 171840 steps per rev.
one step = 5550 microseconds. so one revolution = 15.895
one step = 0.08 Arc seconds.

That should be correct.


Quote:
Originally Posted by bojan View Post
Good job

BTW, Arduino Nano is using low precision crystal (>100ppm) for clock (and every other timing), 10pm would be more appropriate (for AP... but it wouldn't hurt for visual as well).
Of course, this is way better than free running DC motor you had before.
Deviation period of 15 minutes you are mentioning suggests PE (in gearbox probably.. what is the rotational rate of motor/gearbox shaft?).
Reply With Quote
  #4  
Old 30-09-2020, 01:35 PM
bojan's Avatar
bojan
amateur

bojan is online now
 
Join Date: Jul 2006
Location: Mt Waverley, VIC
Posts: 6,927
0.08 arcsec resolution is a bit of overkill... ~0.25 is still ok, so you will be fine with 1:4 or 1:6 timing belt reduction instead of gearbox.

Last edited by bojan; 30-09-2020 at 01:49 PM.
Reply With Quote
  #5  
Old 30-09-2020, 02:22 PM
bojan's Avatar
bojan
amateur

bojan is online now
 
Join Date: Jul 2006
Location: Mt Waverley, VIC
Posts: 6,927
Quote:
Originally Posted by bojan View Post
... Deviation period of 15 minutes you are mentioning suggests PE (in gearbox probably.. what is the rotational rate of motor/gearbox shaft?).

Correction - 15 mins suggest PE on last reduction step: main shaft - sector.
Reply With Quote
  #6  
Old 30-09-2020, 04:43 PM
Rod
Registered User

Rod is offline
 
Join Date: Sep 2006
Location: Mount Martha
Posts: 369
Hi Greg

Great to see your new system going so well. I think it's a great achievement to get a platform tracking this well on the first try. My first one was nowhere near as accurate. I got better results with later designs and a lot of help from friends at the ASV.

Most of the platforms I've built were made with a jig. I know you used Reiner Vogel's excellent design which doesn't need one but I think you could use a jig to sand or grind out any error in the sectors. I can send you a short video of a jig being used if that helps you understand how they work.

I see Bojan is making some suggestions to help with the electronics. Perhaps try those first.

I hope to see some of your photos one day.

Rod.
Reply With Quote
  #7  
Old 30-09-2020, 06:46 PM
Greggles (Gregory)
Registered User

Greggles is offline
 
Join Date: Aug 2017
Location: Sydney, Australia
Posts: 75
Thanks Rod!

Pretty happy, although I never know when to be content.. always wanting more accuracy haha.

Is the jig you use like the one that ed uses in his demo video?

Thabks for the info Bojan!
Reply With Quote
  #8  
Old 30-09-2020, 07:35 PM
JohnL
Registered User

JohnL is offline
 
Join Date: Feb 2020
Location: Victori, Australia
Posts: 3
Hi Greg,

I am planning to make an EQ platform too, with a very similar arduino+stepper motor to drive it, but probably a barn door style with a curved threaded rod and a piano hinge.

A couple of questions:
How do you polar align your one?
Do you use the 1.5x Barlow to allow the camera to focus? Or another reason?
Reply With Quote
  #9  
Old 30-09-2020, 09:12 PM
Greggles (Gregory)
Registered User

Greggles is offline
 
Join Date: Aug 2017
Location: Sydney, Australia
Posts: 75
Hey john. When not at home I roughly polar align with a compas and then use the star drift method to fine tune.

These days I mostly set up at home in the same spot so I have marked true south using the solar moon method. Quick and easy!

And your spot on about the 1.5x barlow. I cant achieve focus otherwise.

Greg

Quote:
Originally Posted by JohnL View Post
Hi Greg,

I am planning to make an EQ platform too, with a very similar arduino+stepper motor to drive it, but probably a barn door style with a curved threaded rod and a piano hinge.

A couple of questions:
How do you polar align your one?
Do you use the 1.5x Barlow to allow the camera to focus? Or another reason?
Reply With Quote
  #10  
Old 01-10-2020, 04:24 AM
Rod
Registered User

Rod is offline
 
Join Date: Sep 2006
Location: Mount Martha
Posts: 369
Quote:
Originally Posted by Greggles View Post
Thanks Rod!

Pretty happy, although I never know when to be content.. always wanting more accuracy haha.

Is the jig you use like the one that ed uses in his demo video?

Thabks for the info Bojan!
Hi Greg

Ed doesn't really use a jig. I made one because at the time I did not have a table saw. Her is an example of what I mean:

http://sawdustfactory.nfshost.com/eq%20platform/

Scroll down the page to see it.

The advantage is that you are sanding the sectors smooth as you rotate them around a virtual polar axis.

Rod.
Reply With Quote
  #11  
Old 01-10-2020, 07:07 AM
Greggles (Gregory)
Registered User

Greggles is offline
 
Join Date: Aug 2017
Location: Sydney, Australia
Posts: 75
Thabks Rod,

Think I'm going to give this a go! Are you able to shoot me
that video. I understand the theory (I think) just want to see it in action and how to pivot around polar axis.

I've always been concerned my segments mathematically may not have been perfect. From my understanding this method would eliminate that doubt. One question is how do you get the initial shape of the sector before sanding it smooth on the jig?

Greg

Quote:
Originally Posted by Rod View Post
Hi Greg

Ed doesn't really use a jig. I made one because at the time I did not have a table saw. Her is an example of what I mean:

http://sawdustfactory.nfshost.com/eq%20platform/

Scroll down the page to see it.

The advantage is that you are sanding the sectors smooth as you rotate them around a virtual polar axis.

Rod.
Reply With Quote
  #12  
Old 01-10-2020, 11:29 PM
Greggles (Gregory)
Registered User

Greggles is offline
 
Join Date: Aug 2017
Location: Sydney, Australia
Posts: 75
Hey every one,

Got some good testing in tonight with interesting results.

So from the start position the mount is tracking to fast, middle is tracking quite well, and towards the end it is tracking too slow. Having said that, at the middle of the platform i was able to have mars in the center of view or my asi462 with a 3x barlow, 600x600 for 10 minutes before clouds came over and swallowed Mars

Now from my understanding the stepper motors speed do not change dependent load. so i believe it is either due to the direct drive system minutely slipping or my vertical sectors are not correct.

i can rationalize the slipping as its faster when the weight of the platform is on the left of the drive shaft pushing the platform right. Where as it at is slowest when the majority of the weight of the platform is on the right hand side of the drive shaft and the motor is trying to pull up the platform with the platform wanting to fall in the opposite direction. i have no idea if i am explaining this in a way any one is going to understand!!


however i need some help determining if the vertical sector shapes could be an issue? for the life of me I cant figure out if the slow and fast speeds mentioned above could come from the sector shapes being incorrect?

Anyway, if any one knows if either of the above two things could be the issue please let me know.

Kind Regards,



Greg
Reply With Quote
  #13  
Old 02-10-2020, 06:36 AM
bojan's Avatar
bojan
amateur

bojan is online now
 
Join Date: Jul 2006
Location: Mt Waverley, VIC
Posts: 6,927
I would check the shape and position of the sector first. And polar alignment, if not accurate will produce the same effect.

Slippage is unlikely (however, I can see some marks on the shaft(s) where the sector is riding that could have been created by grinding on sector, possibly earlier?).
Try to see how much force/imbalance you need for the sector to start to slip...

BTW, do you have (or plan to have in the near future) some sort of hand-paddle with faster/slower buttons to compensate for inaccuracies in alignment? you may add additional motor to elevation shaft as well, your processor is capable of handling that without problem.
Reply With Quote
  #14  
Old 02-10-2020, 07:26 AM
bojan's Avatar
bojan
amateur

bojan is online now
 
Join Date: Jul 2006
Location: Mt Waverley, VIC
Posts: 6,927
One more question: when you stop tracking, in your eyepiece, does the star starts moving in the same line as drifting when tracking?
This could help you to determine the cause of the issue with greater certainty.

Last edited by bojan; 02-10-2020 at 12:10 PM.
Reply With Quote
  #15  
Old 02-10-2020, 12:19 PM
bojan's Avatar
bojan
amateur

bojan is online now
 
Join Date: Jul 2006
Location: Mt Waverley, VIC
Posts: 6,927
When I had a more careful look at your sectors... I realized that arrangement doesn't look right at all.
You have two segments, and they are not in the same plane. And it is very hard to adjust them properly.
Further, if sectors surfaces are not ground properly in some sort of jig, (which will be pretty heavy and large), to form the cone-like surface, you will have issues like you describe (including slippage) because the contact points (lines actually) with shafts are not concentric (they should intersect at the third bearing or at wherever the pivot point is).
Shafts should also be grounded for maximum tracking accuracy IMO.
Reply With Quote
  #16  
Old 07-10-2020, 10:01 PM
Greggles (Gregory)
Registered User

Greggles is offline
 
Join Date: Aug 2017
Location: Sydney, Australia
Posts: 75
Hey all,

Apologies for not replying in a while i've been busy working on my sectors.

I ended up swapping the steel out for aluminium and ground down the sectors using a jig. Photos attached

Crappy weather in Sydney so untested but hopefully get some testing in and some photos of mars!

Bojan, in regards to my sectors not being on the same plane, i believe this is part of the VNS design and the angle is calculated so at a right angle it intersects the the polar axis. although i don't know if I'm explaining this properly.

Regards,

Greg
Attached Thumbnails
Click for full-size image (eq jig 1.jpg)
202.8 KB46 views
Click for full-size image (eq jig 2.jpg)
203.5 KB52 views
Click for full-size image (vertical segment.jpg)
181.5 KB36 views
Reply With Quote
  #17  
Old 08-10-2020, 05:48 AM
bojan's Avatar
bojan
amateur

bojan is online now
 
Join Date: Jul 2006
Location: Mt Waverley, VIC
Posts: 6,927
Greg,

It is OK for sectors not to be in the same plane after grinding and polishing in the jig (like you did). I mentioned this because I thought you didn't know about it, my apologies :-)
You should have better results now, but pay attention to shafts as well, as they will be contributing factor as well.

PE is created mostly in the last transmission stage, because the contribution of the previous stages will always be divided by the reduction ratio of the last one. In your case it could be ~50 (judging by the size of the jig on photos)

Last edited by bojan; 08-10-2020 at 08:01 AM.
Reply With Quote
  #18  
Old 09-10-2020, 01:38 PM
hamishbarker
Registered User

hamishbarker is offline
 
Join Date: Mar 2017
Location: Nelson, new zealand
Posts: 63
from what you are describing, assuming the slow/fast description is referring to error in the RA motion direction and not N/S (dec) it sounds to me that the centre of gravity of your telescope AND the moving part of the platform combined may be below the polar axis.


because they are below, at the beginning of motion the actual centre of gravity is going downhill to the bottom of the arc of motion around the polar axis. then they have to climb back uphill after passing the meridian.

Ideally, the CoG is located on the polar axis, then the stepper has to do just constant work against friction but is not lowering or raising the CoG.

if your telescope simply sits on the platform and the platform isn't part of its azimuth board, you may be able to try some adjustments without modifying the platform.

you could place your telescope a few cm towards the north end of the platform to bring the centre of gravity closer to the polar axis to see if that improves things.

you could also try placing the telescope towards the south of the platform, to see if this makes your tracking speed problem worse.

finally, try placing the telescope a bit to the east of the centre. then it should start with less speed error and gradually get slow as it starts to have to work uphill.

Quote:
Originally Posted by Greggles View Post
Hey every one,

Got some good testing in tonight with interesting results.

So from the start position the mount is tracking to fast, middle is tracking quite well, and towards the end it is tracking too slow. Having said that, at the middle of the platform i was able to have mars in the center of view or my asi462 with a 3x barlow, 600x600 for 10 minutes before clouds came over and swallowed Mars

Now from my understanding the stepper motors speed do not change dependent load. so i believe it is either due to the direct drive system minutely slipping or my vertical sectors are not correct.

i can rationalize the slipping as its faster when the weight of the platform is on the left of the drive shaft pushing the platform right. Where as it at is slowest when the majority of the weight of the platform is on the right hand side of the drive shaft and the motor is trying to pull up the platform with the platform wanting to fall in the opposite direction. i have no idea if i am explaining this in a way any one is going to understand!!


however i need some help determining if the vertical sector shapes could be an issue? for the life of me I cant figure out if the slow and fast speeds mentioned above could come from the sector shapes being incorrect?

Anyway, if any one knows if either of the above two things could be the issue please let me know.

Kind Regards,



Greg
Reply With Quote
  #19  
Old 09-10-2020, 01:52 PM
hamishbarker
Registered User

hamishbarker is offline
 
Join Date: Mar 2017
Location: Nelson, new zealand
Posts: 63
the jig looks good. I see you have both north and south sectors rather than just a north end pivot. this keeps your platform shorter and stiffer, fine.

slippage should not be an issue as long as everything is arranged to put the CoG on the polar axis, or just a smidge to the west of it in order to just balance out friction so that the motor hardly has to do any actual work and just regulates the movement at precisely the drive speed.

I found it well worth spending some time getting my head around drift alignment. I remember it by using an acronym: LEN. LOW EAST NORTH as follows:

if the polar axis of the platform is to LOW (when looking at a star near the eastern horizon),

or is too far EAST of due south (when looking at a star on the meridian near the celestial equator)

then it will appear to drift NORTH in declination when viewed in the eyepiece. Nudge the scope north before watching for drift so that you can tell which side of the field of view is to the north.

using a crosshair eyepiece with the lines rotated so that any RA error is along one line makes it possible to immediately see any DEC error as the star will deviate one side of the line or the other.

I keep a few shims with the platform as I was too lazy to add adjustable feet (and also worried that they would make for flexure).

Quote:
Originally Posted by Greggles View Post
Hey all,

Apologies for not replying in a while i've been busy working on my sectors.

I ended up swapping the steel out for aluminium and ground down the sectors using a jig. Photos attached

Crappy weather in Sydney so untested but hopefully get some testing in and some photos of mars!

Bojan, in regards to my sectors not being on the same plane, i believe this is part of the VNS design and the angle is calculated so at a right angle it intersects the the polar axis. although i don't know if I'm explaining this properly.

Regards,

Greg
Reply With Quote
  #20  
Old 10-10-2020, 09:54 AM
brian2014
Registered User

brian2014 is offline
 
Join Date: Sep 2014
Posts: 1
Testing an Eq Platform

I built a platform a few years ago and developed a fairly simple test procedure.

Set your platform up about 3 to 4 metres from and parallel to a wall.
Make a bracket up for a laser collimator to point at the wall and square to the centreline of your platform.

Start your platform running and every 10 minutes mark a cross on the wall from the laser collimator.

After 1 hour you will have a line of dots on the wall.
End to end distance represents represents one 24th of 360 degrees of travel, which is 15 degrees. Measure the distance to the wall and some high school trig gives you your measured angle. So you can adjust your motor speed to match. Star checking is still the best way, but it found this method gave very good results.
If you measure the 10 min gaps you tell if your platform is slipping.

If you look at the vertical angle the line of spots makes on the wall, it represents latitude the platform is working at. You can then adjust the level to suit your observing site.

Daylight Polar Alignment.
I found using a compass was to inconsistent, so I use this solar method.
Mark a centreline on your base board. Before dark, do an initial setup of the base board. Get a plastic protractor and align it to your centreline. Find something to act as a vertical post with a sharp edge(a builders square works well) and position it at the centre of the protractor. You should now have a nice shadow falling on the edge of your protractor. I then open up Sky Safari under search select the Sun. You can get up a page called Object Info. On that page under Celestial Coordinates is the current Azimuth of the Sun. Have a look at your protractor, if you are polar aligned your shadow line on the protractor should match the current azimuth of the sun. Adjust as required. I estimate I can align to better than a quarter of degree, which is all you need for visual observing. A bigger protractor will give a better result if needed.

Hope this helps. Have fun.

Brian
Reply With Quote
Reply

Bookmarks

Thread Tools
Rate This Thread
Rate This Thread:

Posting Rules
You may not post new threads
You may not post replies
You may not post attachments
You may not edit your posts

BB code is On
Smilies are On
[IMG] code is On
HTML code is Off

Forum Jump


All times are GMT +10. The time is now 12:11 PM.

Powered by vBulletin Version 3.8.7 | Copyright ©2000 - 2024, Jelsoft Enterprises Ltd.
Advertisement
Bintel
Advertisement
Testar
Advertisement