Go Back   IceInSpace > Equipment > ATM and DIY Projects

Reply
 
Thread Tools Rating: Thread Rating: 2 votes, 5.00 average.
  #1  
Old 09-07-2014, 02:22 PM
vignesh1230's Avatar
vignesh1230 (Vignesh)
Registered User

vignesh1230 is offline
 
Join Date: Jun 2011
Location: Melbourne, Australia
Posts: 35
Talking Project Phyllix -DIY EQ mount

Project Phyllix - An EQ mount

Hey guys, I am a student at Nossal High School, currently in Year 11, and seeing as this is the first year we have had the subject Systems Engineering, Ive decided to create an equatorial mount for my Year 12 Systems Engineering project(2015) that we do for the whole year. I realise its a big undertaking but I think the finished product is going to be worth it and if it works well, should rack up some nice marks for the subject, and be a sweet mount to put in an observatory that I plan to build soon.


The primary design goals of this mount are

  • Stability
  • Accuracy
  • Ease of Setup
  • Fit for Observatory Use for years to come
  • ASCOM friendly
  • Friendly Go-To Control Interface - If I have time, I may develop a hand controller.

Now, seeing as a friend of mine owns a machining company and I can get stuff machined cheaper than normal rates, I've decided to go for a 'fancy' telescope mount, and not simply 2 shafts held in with a plate and pillow blocks. Ill keep the design of the mount open source so if anyone wants to hop in and refine the design with their wisdom they can and/or make one themselves


So far I have gone through about 6 revisions of the design where i've just started from scratch again, and I think I'm getting close to a functional design. Having not actually seen or operated a Go-To equatorial mount physically (I have a small CG-3..), I have based most of my design off the White Swan 180, and WS 240, along with the Astro-Physics mount range.
At this stage, It is designed to be 100% made out of Aluminium T6-6061 except for the worm wheels and worm gear. Right now, the latest design is missing its base, DEC shaft and saddle, But as soon as I finish designing it, I shall post an image of it and the cross section.


In the meantime, here's a render I just did of Rev 4, Its missing a lot of things, But generally what it will look like.. Dont mind the Base, It was a trial and error thing that didnt work out. Currently its at 25kg, 12kg on RA and 13kg on DEC so its pretty beefy for a high load capacity. The mount is able to be disassembled into 2 parts by removing 4 bolts ( or 3 if you take off the base)



**Why dont the image tags work on the thread but show in the preview??

Click image for larger version

Name:	ca8b14be1f.jpg
Views:	105
Size:	64.8 KB
ID:	165591



4 Bearings Each Axis
  • 1x AXK130170 on RA Gear, 2x HJ 324120 Needle Roller Bearings on Axis shafts, 1x NTA-3244 on Clutch knob side
  • 1x AXK120155 on DEC Gear, 2x HJ 324120 Needle Roller Bearings on Axis shafts, 1x NTA-3244 on Clutch knob side



Current Drive Idea

  • Arduino Mega2560 - To be replaced later on with a custom board with Atmega2560 chip
  • Nema17 steppers with either 14:1 planetary gearbox or 5:1 timing pulley drive
  • DRV8825 1/32 micro-stepping capable stepper motor driver
  • Worm Drives - DEC 173mm 270 teeth, RA 190mm 297 teeth


I have ordered an Arduino, and as soon as I get it, will start coding the control system.


I don't know whether I should include encoders in the design or not. Opinions would be nice. If I was to employ encoders in this mount I would either use the
CUR AMT203 encoder - Binary (Absolute), Quadrature with Index (Incremental) - Although im not sure if they truly are absolute,
or the
CUR AMT103 encoder - Quadrature with Index (Incremental).

Something tells me that the Absolute encoders are probably a better idea, but they do cost twice the price of each AMT103. The absolute encoders also need me to play around with SPI, which I havent done before, while the AMT103 would be easier to implement, just simply connecting up to the interrupts of the Arduino and counting from there.
I'm guessing the optimal placement of the encoders would be on the worm itself(seeing as on axis encoders are expensive$$$$) to also overcome inaccuracies in the motor gearbox or timing pulley drive.

Most of this year is just planning and designing - Im trying to get a head start on this project so that next year my workload would be reduced.


Some Questions:
1. What should the telescope control system need to do separate from the hand-controller or PC.
eg. Should it have a way to correct for refraction, have its own alignment system, etc..

2. Are encoders on the worm worth it, or just count the number of steps in software.

3. Will encoders on the worm help with reducing PE by a lot?

4. Should the CoG of the telescope mount be over the centre of the base or does the weight of the telescope and CW change the position of the CoG?

5. Losmandy Style Clutch or like the ones on the Astrophysics mounts?


This is a project that is going to keep going as long as I have the mount, so I'll still be working on it after High School.

Any suggestions, help, and tips are more than welcome.

Last edited by vignesh1230; 10-07-2014 at 08:59 PM.
Reply With Quote
  #2  
Old 10-07-2014, 03:39 PM
alistairsam's Avatar
alistairsam
Registered User

alistairsam is offline
 
Join Date: Nov 2009
Location: Box Hill North, Vic
Posts: 1,837
hi,

That's pretty impressive and I'll look forward to your updates.
I'd be interested in how you tackle the ASCOM interface without writing a driver.

I've used the DRV8824 in a fork mount with a stepper and the logic I used for autoguiding was that the ST4 pulse would be monitored by the MCU and would either add or subtract from the current period that the pulses are based on.
This worked well and I didn't account for refraction.

There's a thread here about sidereal rates and I believe you need to track at the King rate to account for most refraction.

Encoders at the axis would be good but would need to be hi-res to provide the resolution required. Renishaw encoders are used in High end mounts with direct drive.
however, as long as you keep backlash and PE low with your gear mesh, you would have repeatability by keeping count of the steps similar to what's done with the EQ series mounts where no encoders are used. Just steps.

The render does look very good. will it be milled or cast?

Cheers
Alistair
Reply With Quote
  #3  
Old 13-07-2014, 08:57 AM
Stefan Buda
Registered User

Stefan Buda is offline
 
Join Date: Jul 2013
Location: Melbourne, VIC
Posts: 823
Hi Vignesh,

I have made a couple of small EQ mounts in the past and the most difficult parts to make are the worms and the worm wheels. Are you sure your friend's workshop has the specialized tooling that is required?

Cheers,
Stefan
Reply With Quote
  #4  
Old 14-07-2014, 11:35 AM
vignesh1230's Avatar
vignesh1230 (Vignesh)
Registered User

vignesh1230 is offline
 
Join Date: Jun 2011
Location: Melbourne, Australia
Posts: 35
Quote:
Originally Posted by alistairsam View Post
hi,

That's pretty impressive and I'll look forward to your updates.
I'd be interested in how you tackle the ASCOM interface without writing a driver.

I've used the DRV8824 in a fork mount with a stepper and the logic I used for autoguiding was that the ST4 pulse would be monitored by the MCU and would either add or subtract from the current period that the pulses are based on.
This worked well and I didn't account for refraction.

There's a thread here about sidereal rates and I believe you need to track at the King rate to account for most refraction.

Encoders at the axis would be good but would need to be hi-res to provide the resolution required. Renishaw encoders are used in High end mounts with direct drive.
however, as long as you keep backlash and PE low with your gear mesh, you would have repeatability by keeping count of the steps similar to what's done with the EQ series mounts where no encoders are used. Just steps.

The render does look very good. will it be milled or cast?

Cheers
Alistair

Thanks Alistair

EDIT: About the Ascom(i forgot about it when i typed this up at school), I was either planning on emulating a mount made by Sky watcher and using it with EQAscom, or actually write a driver myself. Ive had a thorough look through all the documentation available on the telescope mount drivers, and I think, given time, Ill be able to make a working driver, and then make refinements and so on.

Its good to know that someone else has used the DRV8825 chip to control a telecope mount haha. I was originally planning to just use pulse guiding to guide the telescope but I guess i can add an ST4 port just in case. In regards to not accounting for refraction, how good was the pointing accuracy, and how much work did the guider have to do?

Definately cant afford Axis encoders like the Renishaw. Hence the idea to put an encoder on the worm itself, to control the PE better.

The whole thing is designed to be milled and lathed, for simplicity's sake.


Quote:
I have made a couple of small EQ mounts in the past and the most difficult parts to make are the worms and the worm wheels. Are you sure your friend's workshop has the specialized tooling that is required?
The worms and worm wheels are the parts that I'm most worried about. I'm not sure how I'm going to tackle them. I have been to his workshop and he does own a lot of different CNC lathes, but I'm pretty sure he dosen't have like Class AA hobs or anything. Ill check with him..
He may be able to do everything but the gears actually, but there seems to be 2 gear machining companies near his workshop, so ill be trying to get a quote soon

Last edited by vignesh1230; 14-07-2014 at 09:16 PM.
Reply With Quote
  #5  
Old 15-07-2014, 08:35 PM
Wavytone
Registered User

Wavytone is offline
 
Join Date: Jul 2008
Location: Killara, Sydney
Posts: 4,147
Vignesh,

Loss of mass just means a cumbersome mount that is annoying to lug around - it doesn't necessarily mean it is particularly stiff in response to the loads (and torques applied on the shafts). One of the reason dobsonians are surprisingly stiff, despite the use of lightweight low density materials like plywood or composites.

Use FEA to work out where this flexes (and where it doesn't) in order to work out when it needs stiffening, and where you can trim the lard off it ? My guess is you could halve the mass without much difficulty, and easily done if this is to be produced by CAD/CAM.

With a telescope on top and thus sitting on a pier or tripod, stiffness is best measured by the frequency is oscillates at when a step or impulse load is applied - the higher the frequency the better.

The second characteristic that matters is how fast the oscillations are damped - the faster the better. In this respect lighter mass is a huge advantage - the amplitude is reduced, and the damping is faster in comparison to a similar but heavier design.

Systems engineering is one thing; classical control system theory is a rather overlooked subject - as a subject it is intensely mathematical and difficult - and much more important in many domains if you are aiming to be an EE.
Reply With Quote
  #6  
Old 16-07-2014, 10:30 AM
LightningNZ's Avatar
LightningNZ (Cam)
Registered User

LightningNZ is offline
 
Join Date: Oct 2011
Location: Canberra
Posts: 951
With regards to making the worms and wheels, you can use a precision threaded rod that has been cut down part of its length to act as a thread cutter. Use a small portion of the rod to do the cutting and another part of it will be your worm. Go through the ATM forum at CloudyNights.com - there are quite a few discussions on this sort of thing there, though you'll likely have to push through quite a few pages of threads before you find exactly what you're after.

Good luck, you're very ambitious but you also sound very talented and determined.
Reply With Quote
  #7  
Old 17-07-2014, 07:22 PM
vignesh1230's Avatar
vignesh1230 (Vignesh)
Registered User

vignesh1230 is offline
 
Join Date: Jun 2011
Location: Melbourne, Australia
Posts: 35
Quote:
Originally Posted by Wavytone View Post
Use FEA to work out where this flexes (and where it doesn't) in order to work out when it needs stiffening, and where you can trim the lard off it ? My guess is you could halve the mass without much difficulty, and easily done if this is to be produced by CAD/CAM.

With a telescope on top and thus sitting on a pier or tripod, stiffness is best measured by the frequency is oscillates at when a step or impulse load is applied - the higher the frequency the better.

The second characteristic that matters is how fast the oscillations are damped - the faster the better. In this respect lighter mass is a huge advantage - the amplitude is reduced, and the damping is faster in comparison to a similar but heavier design.
Wavytone,

Would it be beneficial to run stress tests on individual parts in approximately where the loads would be and then run a full analysis on the fully modeled assembly?

I haven't particularly been focusing alot on weight reduction, yet.. Just making sure that the mass of each part was reasonable. I actually forgot all about the resonant frequencies, and how that affects the stiffness! Thanks for that!



Quote:
With regards to making the worms and wheels, you can use a precision threaded rod that has been cut down part of its length to act as a thread cutter. Use a small portion of the rod to do the cutting and another part of it will be your worm.
I don't think i've seen that yet, I have seen someone on this forum making their own worms and wheels with the same technique/similar I believe. It seems pretty promising and I think ill have a go at making some like that. I don't know how great the PE will be by using presicion threaded rod, rather than making one from scratch. But it does make things easier!

Thanks Cam, Im trying my best on this project haha
Reply With Quote
  #8  
Old 03-02-2015, 07:39 PM
vignesh1230's Avatar
vignesh1230 (Vignesh)
Registered User

vignesh1230 is offline
 
Join Date: Jun 2011
Location: Melbourne, Australia
Posts: 35
Its been a long time since I updated this thread.

Year 12 has finally begun and I can start building this thing!

So what has happened in the meantime?
  • Design is 90% complete. Figuring out holes for the latitude adjustment and holes for a Dovetail saddle plate (I would make one if I could find some details and specifications for designing one, but if not ill have to buy something like the Williams Optics one)
  • Decided on a tapered bearing at the rear end (32010x) and a needle roller bearing at the front (NK50/25), mostly to make it easier for me design wise and I felt that it would be redundant to have another tapered bearing at the top when there is a thrust bearing to take the axial load.
  • Nema17 motors go through a 4:3 gearing system to drive the worm.
  • Worm is going to be mounted on easily available skateboard bearings(they may not be designed for this purpose but having skated with them and seeing how smooth they roll at low speeds makes me confident that they'll be alright to use)
  • It is going to be running EQAscom with an Arduino mimicking a Skywater mount through the AstroEQ software, for the time being.
  • All bearings and moving parts have been bought - just awaiting delivery*

    * The size of the tapered bearing is much bigger than what i had imagined but the measurements all check out.
Click image for larger version

Name:	b8eb588c56.jpg
Views:	67
Size:	20.2 KB
ID:	177178



(I had the drive system mounted on the design before but decided to take it away because it was just getting in the way when i was modelling it)
But yeah, just thought I'd update this thread.
Attached Thumbnails
Click for full-size image (1.jpg)
36.4 KB56 views
Reply With Quote
  #9  
Old 03-02-2015, 10:01 PM
mswhin63's Avatar
mswhin63 (Malcolm)
Registered User

mswhin63 is offline
 
Join Date: May 2009
Location: Para Hills, South Australia
Posts: 3,620
Good luck with the project.

My son has just started Plant and Mechanics he has offered his services to help me build a Trike Not a EQ mount project but these course look really interesting, nothing at all what I did in Year 11.
Maybe he can help me build a Horseshoe mount for my DOB
Reply With Quote
  #10  
Old 04-02-2015, 07:52 AM
astro_nutt
Registered User

astro_nutt is offline
 
Join Date: Mar 2005
Location: Melbourne
Posts: 1,013
Hi Vignesh.
That is one awesome project you have undertaken. I'm no engineer but By the sound and looks of it, it will have some serious uses after you present it for your Systems Engineering.
Good luck!
Reply With Quote
Reply

Bookmarks

Tags
diy, eq mount

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 10:00 AM.

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