Jump to content

  •  

CNers have asked about a donation box for Cloudy Nights over the years, so here you go. Donation is not required by any means, so please enjoy your stay.

Photo

Electronics/mechanical geniuses out there?

  • Please log in to reply
9 replies to this topic

#1 rjhat3

rjhat3

    Mariner 2

  • *****
  • topic starter
  • Posts: 230
  • Joined: 23 Mar 2018

Posted 01 August 2020 - 12:01 PM

I am considering trying to build my own AP mount. I am familiar with electronics, not so much with mechanical. I have access to a vast array of electronics, including motors, encoders, control boards, connectors, and even individual electronic components.
A few days ago, I was replacing a clock motor at work. I realized that the motor was turning exactly 1 revolution per minute. The timing of the motor is achieved using the 60 hertz frequency of AC power.
I am wondering if I can use that motor along with a couple of encoders to measure accuracy. I would install a encoder control board, and devise a way to vary the 60 hertz frequency to the motor for corrections.
I realize I that software is gonna have to be written to achieve this.
Does this sound like a doable idea? Or am I thinking to far out in left field?

#2 bobzeq25

bobzeq25

    ISS

  • *****
  • Posts: 20,431
  • Joined: 27 Oct 2014

Posted 01 August 2020 - 12:51 PM

I am considering trying to build my own AP mount. I am familiar with electronics, not so much with mechanical. I have access to a vast array of electronics, including motors, encoders, control boards, connectors, and even individual electronic components.
A few days ago, I was replacing a clock motor at work. I realized that the motor was turning exactly 1 revolution per minute. The timing of the motor is achieved using the 60 hertz frequency of AC power.
I am wondering if I can use that motor along with a couple of encoders to measure accuracy. I would install a encoder control board, and devise a way to vary the 60 hertz frequency to the motor for corrections.
I realize I that software is gonna have to be written to achieve this.
Does this sound like a doable idea? Or am I thinking to far out in left field?

For Deep Space imaging, it's pretty far out.  The demands on tracking are extreme.  Camera pixels are .005mm.  Move the image over a few during a subexposure, and the image is toast.

 

It's just not intuitive how good a mount is required for DSO imaging.  With either a homebuilt or an economical (ie less that $1500) storebought, you need to autoguide, it's not economic to even make gears that are "good enough".

 

If you have a year to spend on this, maybe.  Maybe.


Edited by bobzeq25, 01 August 2020 - 12:52 PM.

  • Jon Isaacs likes this

#3 Arcamigo

Arcamigo

    Vostok 1

  • -----
  • Posts: 128
  • Joined: 31 May 2020
  • Loc: Gaston, Oregon, United States

Posted 01 August 2020 - 01:23 PM

The old Celestron C8 mounts used the 60 Hz frequency to control the rotation of the mount, but today's mounts are something completely different, especially the higher end ones. You're going to need access to a CNC milling machine that can handle large blocks of aluminum, if you want to build your own, because an AP mount has to remain rigid under the weight of the telescope, but not so rigid that it will fracture. Any flexure, even under temperature changes, will ruin your image. Then you have to think about how you will handle the gearing so there is little to no backlash. For AP you're trying to maintain arcsecond precision over hours.

 

The electronics will not use the 60Hz frequency. You will need precise, stable clock generator circuits from which you can derive the frequencies for your motors. This would be easier and more precise than trying to change the AC 60 Hz frequency.

 

I don't think firmware is that critical because if you design your interface to be INDI compatible, an external Raspberry Pi can do the celestial mathematics using a variety of available software. However, you will still need a microcontroller to do the INDI interfacing.

 

I do some of this stuff for my job, so I also know that everything I just wrote only addresses about 1% of the problem. I'm not trying to discourage you. Maybe you'll be the next Scott Losmandy!



#4 Eddgie

Eddgie

    ISS

  • *****
  • Posts: 26,582
  • Joined: 01 Feb 2006

Posted 01 August 2020 - 01:51 PM

I am considering trying to build my own AP mount. I am familiar with electronics, not so much with mechanical. I have access to a vast array of electronics, including motors, encoders, control boards, connectors, and even individual electronic components.
A few days ago, I was replacing a clock motor at work. I realized that the motor was turning exactly 1 revolution per minute. The timing of the motor is achieved using the 60 hertz frequency of AC power.
I am wondering if I can use that motor along with a couple of encoders to measure accuracy. I would install a encoder control board, and devise a way to vary the 60 hertz frequency to the motor for corrections.
I realize I that software is gonna have to be written to achieve this.
Does this sound like a doable idea? Or am I thinking to far out in left field?

For guided tracking you need to be able to vary the speed of the motor and this would be difficult using 60hz motor and direct drive.

 

Modern mounts use small motors that are geared up though a very large drive ration gearbox.  This way I can spin the motor fast but still make minute adjustments in speed.

 

In other words, the one revolution a minute is working against you.  Better to have a motor that spins very fast that can be fed though a gear box to decrease RPM and increase torque. This allows the very precise speed control necessary for tracking with high resolution sensors. A small motor spinning at lots of RPMs running through a gearbox to step down the speed is they way almost everyone is doing it and that is likely simply because it is the best way to do it.


  • bobzeq25 likes this

#5 Michael Covington

Michael Covington

    Author

  • *****
  • Vendors
  • Posts: 6,236
  • Joined: 13 May 2014
  • Loc: Athens, Georgia, USA

Posted 01 August 2020 - 03:13 PM

For guided tracking you need to be able to vary the speed of the motor and this would be difficult using 60hz motor and direct drive.

 

Modern mounts use small motors that are geared up though a very large drive ration gearbox.  This way I can spin the motor fast but still make minute adjustments in speed.

 

In other words, the one revolution a minute is working against you.  Better to have a motor that spins very fast that can be fed though a gear box to decrease RPM and increase torque. This allows the very precise speed control necessary for tracking with high resolution sensors. A small motor spinning at lots of RPMs running through a gearbox to step down the speed is they way almost everyone is doing it and that is likely simply because it is the best way to do it.

Actually, in the 1980s we all had those 1-rpm motors or something like them -- they are 3600-rpm motors that are geared down internally -- and precise speed control is easy, by varying the frequency of the AC power.  I designed several circuits to do that; the best is called "Alcor" and was published in the JBAA.  http://www.covington...alcor/alcor.pdf

 

What you cannot do is make it run more than a few percent faster or slower than normal speed.  So, you can make guiding corrections, but you cannot slew to objects.


  • Jon Isaacs likes this

#6 AhBok

AhBok

    Mercury-Atlas

  • *****
  • Posts: 2,677
  • Joined: 02 Dec 2010
  • Loc: Lakeland, TN

Posted 01 August 2020 - 03:26 PM

I did this very thing in the early 70’s. It was actually not uncommon to see homemade driven eq mounts back then. Mine also used a motor speed control. Unfortunately, mine was fine for visual use with my DYI 8” newt. AP was a completely different thing. The best I ever got was with my Pentax piggybacked on the newt using lenses under 300mm. Of course, back then I was thrilled with my poor-mediocre results!

#7 AhBok

AhBok

    Mercury-Atlas

  • *****
  • Posts: 2,677
  • Joined: 02 Dec 2010
  • Loc: Lakeland, TN

Posted 01 August 2020 - 03:29 PM

I should mention I drove my mount in RA only. Dual axis control was high tech back then!!
  • Jon Isaacs likes this

#8 t-ara-fan

t-ara-fan

    Apollo

  • -----
  • Posts: 1,325
  • Joined: 20 Sep 2017
  • Loc: 50° 13' N

Posted 01 August 2020 - 06:17 PM

 
Does this sound like a doable idea? Or am I thinking to far out in left field?

The mechanical part is the hardest.  The electronics are fairly straightforward as you figured out, using steppers or synchronous motors.

 

My advice: build a motorized tracker first. See how it goes.  That is probably 1% of the complexity of a full EQ MOUNT.  I have seen some 3D printed trackers on the Intertubes.  They look decent.

 

Trying to build an EQ mount is kind of like building a car. Yes you could do it but the cost/quality quotient would be ... astronomical lol.giflol.giflol.giflol.giflol.gifcool.gif

 

Some people build EQ mounts ... old school with no goto. Look up "Byers drive"  or click here http://www.edbyersco.com/in_stock.html


Edited by t-ara-fan, 01 August 2020 - 06:19 PM.

  • Jon Isaacs and bobzeq25 like this

#9 rjhat3

rjhat3

    Mariner 2

  • *****
  • topic starter
  • Posts: 230
  • Joined: 23 Mar 2018

Posted 01 August 2020 - 08:56 PM

Thanks for the info, guys. I do have a buddy that’s a machinist. He could help me with the fabrication, however he couldn’t tell you what Polaris was, much less a telescope mount.
I see the issue with the one hertz motor. Something I should’ve thought about was the torque needed, and that motor wouldn’t do it.
I have never changed the input frequency on one of those, but knowing that it will only change a few percent is very discouraging.
And yes, you guys blew a hole in my plan!🤣 Let’s face it, it really wasn’t a plan, as much as an idea. So, how about another idea?
What if I purchased a couple of precision encoders, an encoder interface board, and modified my Atlas to make it more precise? It guides very well with PHD2 and a guide cam. What it doesn’t do, is keep track of where the night sky is over a long period of time. In other words, after I do an alignment, I can point to an object from the database no prob. Six hours later, it’s not even in the FOV. This has got to be an error in timing, or mechanics.
I wonder if I could attach precision encoders to each axis, and interface them with the firmware to “correct” this error.
I’m not sure how precise the encoders are on the Atlas EQ-G. I’ve had it apart, and it appears that the encoders are integrated into the motor assemblies, so changing them really wouldn’t be an option. I could bypass them, however. Does anyone have detailed info on the encoders for this mount? The issue would be that the motor control board expects to se a certain count from the encoders, and equates this to a certain degree of movement. With higher resolution encoders, this would likely need to be adjusted.

#10 hcf

hcf

    Messenger

  • -----
  • Posts: 450
  • Joined: 01 Dec 2017

Posted 02 August 2020 - 12:51 AM

What it doesn’t do, is keep track of where the night sky is over a long period of time. In other words, after I do an alignment, I can point to an object from the database no prob. Six hours later, it’s not even in the FOV. This has got to be an error in timing, or mechanics.
 

A different approach to this problem, is to use platesolving alongside GoTo. For AP, you already have a camera attached, and most common AP tools, integrate platesolving with GoTo for accurate framing.

 

For visual, you could add a camera, in the mix, maybe the guidescope/cam or even simpler, and use platesolving.

Take a look at my DIY platesolving GoTo mount for visual which is always accurate. As an added bonus in my  setup, you can unclutch the EQ mount and move it by hand, and other than aligning the camera with the scope, no star alignment is needed.

 

https://www.cloudyni...-project-ps-g2/


Edited by hcf, 02 August 2020 - 12:58 AM.



CNers have asked about a donation box for Cloudy Nights over the years, so here you go. Donation is not required by any means, so please enjoy your stay.


Recent Topics






Cloudy Nights LLC
Cloudy Nights Sponsor: Astronomics