Jump to content


Photo

Beautiful Homebuilt GEM

  • Please log in to reply
52 replies to this topic

#1 sternenhimmel

sternenhimmel

    Vostok 1

  • -----
  • Posts: 102
  • Joined: 05 Jun 2013

Posted 07 August 2013 - 07:35 PM

I have been researching the idea of building my own GEM from scratch, and have seen a number of excellent builds -- but haven't really found any good plans or CAD files just yet.

I stumbled upon this particular "light" build today:
http://www.theskyscr...quatorial-mount

I thought one, the polished aluminum is absolutely beautiful, but more importantly, the design seems relatively simple to construct with a lathe and a mill. The changes I would make would be to motorize the declination axis and simplify some of the geometry (e.g. no conical shapes, fewer rounded corners, etc).

Material costs probably won't be cheap for this project, but I reckon I could still save *money* on this project in exchange for my time. Thoughts?

I would be using the mount for lightweight primefocus astrophotography with autoguiding (with what will be custom electronics to drive the axes and interface -- hopefully).

This has also been a useful document:
http://bossanova9.or...uatorialMts.pdf

#2 Chucke

Chucke

    Explorer 1

  • -----
  • Posts: 96
  • Joined: 12 Mar 2010

Posted 07 August 2013 - 08:07 PM

I would suggest not having so much overhang on the Dec. Make everything as closely coupled as possible. You will be glad you did. If you do it right damping won't be an issue because there will not be any noticeable deflection to damp.

Look at Schaefer mounts as an example.

http://www.larryadki...llInrtoducti...


Chuck

#3 sternenhimmel

sternenhimmel

    Vostok 1

  • -----
  • Posts: 102
  • Joined: 05 Jun 2013

Posted 08 August 2013 - 02:19 PM

How does the overhang influence performance? Is it simply a matter of decreasing the moment arm?

So if I understand correctly, the Schaefer mounts move the large RA spur gear closer to the Dec axis? Like this:
http://www.siderealt...mount.German... ?

#4 Chucke

Chucke

    Explorer 1

  • -----
  • Posts: 96
  • Joined: 12 Mar 2010

Posted 08 August 2013 - 04:57 PM

Where the overhang comes in to play is that flexure varies with the cube of the overhang so you really want to minimize it.

Schaefer mounts put the gear up front in RA and at the top for Dec. The gear then acts as part of a large thrust bearing. In addition to reducing deflection from overhang it also reduces torsion of the shaft. With a rear mounted gear/clutch the shaft twists when a force is applied to the OTA. These flexures are not very large but they are greatly magnified by the lever arm of the telescope tube and the magnification of the eyepiece. The result is a springy telescope. Look at Paramounts and AP mounts. They all use front/top mounted gears for this reason. There is minimal overhang and a very short length of shaft to wind up from the point where the force is applied (the end of the tube)to the point where the shaft is restrained (the clutch). If you are interested, the formulae can be found in Machinery's Handbook. I first learned about these design issues in the 1980 RTMC Proceedings which had some simplified versions of the formulae.

Gerry Logan and Bill Schaefer were both influenced by Russell Porter's designs in the ATM books. Gerry's mounts are quite sturdy and were built without machine tools.

Chuck

#5 Sean Cunneen

Sean Cunneen

    Let Me Think

  • *****
  • Posts: 3217
  • Joined: 01 Aug 2007
  • Loc: Blue Island Illinois

Posted 08 August 2013 - 08:10 PM

I have been building mine since 2009. I can't believe it's been 4 years, but I recently found the receipt from my first aluminum order and sure enough...

If you are looking for tried-and-true, look at Opticraft mounts... The pillow-block mount is as basic, simple and reliable design. Cheap and customizeable.

I built mine because I wanted something big and stiff which I could use for potentially very large OTA's. Some folks build them because they have the capability to make very beautiful precision parts.

The basic design of a gem is simple enough, the hard part is the gearing. Most often your choice of gearing determines the aesthetic of the rest of the build. In my case, the ra gear and main bearings both fell into my lap on the same week. Since then it has been a process of refining and upgrading as I learned about materials and my own capabilities as a builder. There are very few "rights" and an awful lot of "wrongs" so be prepared for the long-haul. The reward is a one-of-a-kind build that always garners a lot of attention! You can follow the link to the build thread in my signature. Good luck!

#6 sternenhimmel

sternenhimmel

    Vostok 1

  • -----
  • Posts: 102
  • Joined: 05 Jun 2013

Posted 08 August 2013 - 09:26 PM

Thanks for the insight Sean; your words of warning make me somewhat nervous though!

I'm only going to have free access to a machine shop for the next year. After that, I'm on my own. I am hoping that I can work on this project maybe 5 hours a week over the course of 4 months and have something capable of astrophotography with a 80mm refractor, 400-500mm focal length, and with autoguiding for both axes.

I would like to take this project from paper, to CAD, and to the shop. I want it to be well thought out and have as few hiccups as possible. That means learning from other people's mistakes, and modifying only slightly existing designs. Are there any freely/cheaply available GEM CAD files out there?

The pillow block mounts seem like a good compromise if I decide the project is too ambitious. However, these mounts have the RA gear hanging below the bearings -- from the discussion above, isn't this undesirable?

Also, from your experience, am I looking at saving money for the quality of mount I produce? Or will I be producing a mount that on the market could be had for cheap. My hope is to come out financially on top of the game, and not at a loss. I don't have a good sense of material costs, but if I use aluminum for the bulk of the parts, then the bulk of the costs would be in the gearing. Unless, that is, I make my own gears. Though the trend seems to be that people purchase the worm gears, then machine the spur gears.

#7 Starhawk

Starhawk

    Space Ranger

  • *****
  • Posts: 5509
  • Joined: 16 Sep 2008
  • Loc: Tucson, Arizona

Posted 08 August 2013 - 09:56 PM

I tried asking Dick Parker if it was possible to get a drawing set for the beautiful reproduction Clark and was told no, but they'd be for sale someday, and I should figure something else out for myself. I didn't assume a drawing package would be free, but this response was a surprise. Needless to say, I was disappointed, given the amount of photos and seemingly open discussion posted from that team.

Anyway, after that, a thought I had was to look at the old Unitron clock drive mounts and maybe use one of them as a guide. I'd like to have a nice drive able to run on something like 9V batteries, and the ability to use encoders, which I think could be figured out.

-Rich

#8 sternenhimmel

sternenhimmel

    Vostok 1

  • -----
  • Posts: 102
  • Joined: 05 Jun 2013

Posted 09 August 2013 - 12:51 AM

That's surprising to me as well. If I came up with a design that worked, I think I'd make it freely available for people ti improve upon.

I don't think my design will incorporate encoders, but I also don't see it being terribly difficult to add them. I just can't be spending the money on quality encoders.

On that subject, what motors are generally preferred for small GEMs? Steppers, servos, or DC?

#9 orlyandico

orlyandico

    Fly Me to the Moon

  • *****
  • Posts: 5362
  • Joined: 10 Aug 2009
  • Loc: Singapore

Posted 09 August 2013 - 02:27 AM

Steppers are easier for the hobbyist to comprehend, and can be driven with simple (non-CPU) electronics. Steppers with high gear ratios (not usable for GoTo, only for tracking) are cheap. But hybrid steppers with low-ratio gearboxes suitable for GoTo are not cheap ($200+ each, new). Also, steppers cannot detect if they are losing steps, unless you put a shaft encoder on them. The Takahashi NJP and EM400 do this, probably all the Temma mounts.

Servos meanwhile require a servo controller because they need an encoder for the closed-loop control. So more complex electronics (but no more complex than a stepper with an encoder).

Mounts like the Atlas/EQ6 etc. are hybrid stepper but have no encoders, so the electronics are simpler than the servo equivalents like the CGEM.

One thing: servos are DC motors, and usually spin at high RPM. This means you need a large gear reduction (about 50:1 at least, compared to 5:1 to 10:1 for steppers). A large reduction usually requires a more complex gearhead, with a lot of gear stages. And if that gearhead is cheap (read: CGEM) there will be intractable periodic error harmonics from that complex gearhead.

Steppers can make do with a 1- or 2-stage gearbox due to the low reduction ratio. This is why we hear about 8/3 periodic error in CGEMs but nothing similar for Atlas/EQ6.

To eliminate gearhead periodic error with servos, you need a premium gearhead, such as those made by Pittman, Maxon, Dunkermotoren, etc. I believe Roland has stated that AP mounts use an in-house fabricated gearhead with Swiss-made Maxon motors.

#10 EFT

EFT

    Vendor - Deep Space Products

  • *****
  • Vendors
  • Posts: 2544
  • Joined: 07 May 2007
  • Loc: Phoenix, AZ

Posted 09 August 2013 - 11:15 AM

Thanks for the insight Sean; your words of warning make me somewhat nervous though!

I'm only going to have free access to a machine shop for the next year. After that, I'm on my own. I am hoping that I can work on this project maybe 5 hours a week over the course of 4 months and have something capable of astrophotography with a 80mm refractor, 400-500mm focal length, and with autoguiding for both axes.

I would like to take this project from paper, to CAD, and to the shop. I want it to be well thought out and have as few hiccups as possible. That means learning from other people's mistakes, and modifying only slightly existing designs. Are there any freely/cheaply available GEM CAD files out there?

The pillow block mounts seem like a good compromise if I decide the project is too ambitious. However, these mounts have the RA gear hanging below the bearings -- from the discussion above, isn't this undesirable?

Also, from your experience, am I looking at saving money for the quality of mount I produce? Or will I be producing a mount that on the market could be had for cheap. My hope is to come out financially on top of the game, and not at a loss. I don't have a good sense of material costs, but if I use aluminum for the bulk of the parts, then the bulk of the costs would be in the gearing. Unless, that is, I make my own gears. Though the trend seems to be that people purchase the worm gears, then machine the spur gears.


Not to discourage you, but if you are an excellent machinist with good access to materials, you might be able to come out ahead on costs for a project like this. However, if you are not an excellent machinist with materials available, then it is rarely going to result in a cost savings to build something of high quality. In that case, you undertake a project like this because of the challenge or the satisfaction of doing it yourself, not to try to save money and don't forget that your time is worth something.

#11 sternenhimmel

sternenhimmel

    Vostok 1

  • -----
  • Posts: 102
  • Joined: 05 Jun 2013

Posted 09 August 2013 - 11:54 AM

I am by no means an excellent machinist. I'm a beginning machinist with very little experience. I might have access to some scrap material, but a lot of it will have to be purchased. I will have the guidance of excellent machinists, if that's worth anything.

I'm a maker at heart, and if I see a way to do something myself, I'll (try to) do it myself. If nothing else, it would be an excellent project to become more familiar with the mill and lathe.

I believe I saw a video of the Atlas slewing and it sounded like a stepper motor that was being microstepped. Would that be a correct assumption? If I plan on using a guide scope, would encoders really be necessary (ignoring GoTo capability for now)?

#12 orlyandico

orlyandico

    Fly Me to the Moon

  • *****
  • Posts: 5362
  • Joined: 10 Aug 2009
  • Loc: Singapore

Posted 09 August 2013 - 12:28 PM

Yes the Atlas is using microstepped hybrid steppers.

The Atlas doesn't have encoders on its steppers. The Takahashi NJP does. I would think there is value to those encoders, but you can live without them.

BTW there is a project out there that uses an Arduino to emulate Atlas mount electronics. So you can build your own mount with your own gear ratios (not matching the Atlas), but control it using a Synscan handset or EQMOD.

https://github.com/TCWORLD/AstroEQ

#13 EFT

EFT

    Vendor - Deep Space Products

  • *****
  • Vendors
  • Posts: 2544
  • Joined: 07 May 2007
  • Loc: Phoenix, AZ

Posted 09 August 2013 - 12:38 PM

I am by no means an excellent machinist. I'm a beginning machinist with very little experience. I might have access to some scrap material, but a lot of it will have to be purchased. I will have the guidance of excellent machinists, if that's worth anything.

I'm a maker at heart, and if I see a way to do something myself, I'll (try to) do it myself. If nothing else, it would be an excellent project to become more familiar with the mill and lathe.

I believe I saw a video of the Atlas slewing and it sounded like a stepper motor that was being microstepped. Would that be a correct assumption? If I plan on using a guide scope, would encoders really be necessary (ignoring GoTo capability for now)?


Doing it yourself can be very rewarding, a good learning experience on the machines, and a good way to gain a better understanding of how mounts really work. That is what I do with a lot of things.

The Atlas uses steppers. Steppers are easier to deal with than servos with encoders, but they are not necessarily as accurate because they work on the assumption that the steps that the motor was told to make were actually made and made correctly. If you use any CNC machines, you may quickly gain and appreciation for this problem.

#14 Sean Cunneen

Sean Cunneen

    Let Me Think

  • *****
  • Posts: 3217
  • Joined: 01 Aug 2007
  • Loc: Blue Island Illinois

Posted 09 August 2013 - 03:17 PM

Remember that time when you typed the words "GOTO"? Right there you made it more expensive to build than to buy. Commercial mounts are all--in-one packages and goto is very expensive to DIY. Structurally, Modern mounts can be purchased for approximately the same material costs as what it takes to build so the GOT functionality is basically gratis. However the difference comes in quality. You do not need to cut costs therefore your components can be the best available. Dependability and the ability to repair/replace a broken component means your mount will last for decades rather than a few years. The other benefit of DIY is the ability to upgrade and modify should better technology make itself available. You cannot rely on backwards-compatibility from commercial manufacturers. The ones that do make their equipment with upgrades in mind(AP and Losmandy come to mind) Are so expensive for that very reason. So to see the real economic benefit of DIY, you have to think long-term.

#15 EJN

EJN

    Vanguard

  • *****
  • Posts: 2171
  • Joined: 01 Nov 2005
  • Loc: 53 miles west of Venus

Posted 09 August 2013 - 06:40 PM

I would suggest not having so much overhang on the Dec.


That mount looks to have similar Dec. overhang to the Losmandy GM-8,
and that is a highly regarded mount.

GM-8

#16 sternenhimmel

sternenhimmel

    Vostok 1

  • -----
  • Posts: 102
  • Joined: 05 Jun 2013

Posted 09 August 2013 - 08:26 PM

GoTo isn't a a highly desired feature for me but I also wouldn't see it being an expensive feature? It can be implemented with software, some driver boards, and an arduino or raspberry pi as mentioned above with the open source code. However, this is a method that requires assumptions about the position of the axes, unless encoders are used.

I'll start drawing up plans and iterating through design improvements. There are many things the details of will need hammering out...

#17 Sean Cunneen

Sean Cunneen

    Let Me Think

  • *****
  • Posts: 3217
  • Joined: 01 Aug 2007
  • Loc: Blue Island Illinois

Posted 09 August 2013 - 09:25 PM

Unless you get lucky with surplus or you pull them out of something, encoders will run you $100 ea, easy. Add in the boards, drivers and an oscilloscope and the price goes up. Don't forget all the tuning. But don't pay too much attention to me, I have always been a star-hopper, so my bias comes through.

#18 orlyandico

orlyandico

    Fly Me to the Moon

  • *****
  • Posts: 5362
  • Joined: 10 Aug 2009
  • Loc: Singapore

Posted 09 August 2013 - 10:56 PM

Vexta hybrid steppers with gearheads are $50 each on ebay. New they are $250+.

The link I've shown above provides GoTo electronics for UKP 80 or about $150 USD.

#19 sternenhimmel

sternenhimmel

    Vostok 1

  • -----
  • Posts: 102
  • Joined: 05 Jun 2013

Posted 09 August 2013 - 11:24 PM

Anyway, I wouldn't be worried about implementing a simple GoTo system. As orlyandico pointed out, there are affordable options out there -- encoders would be an upgrade if ever they was needed.

But I digress, the electronics I can get a grasp of, and there are lots of resources available out there for custom drive solutions for telescope mounts. The machining aspects, and the details of GEM construction on the other hand, is sparsely documented. It will require a significant effort to piece together everything and come up with something that will function well.

#20 Starhawk

Starhawk

    Space Ranger

  • *****
  • Posts: 5509
  • Joined: 16 Sep 2008
  • Loc: Tucson, Arizona

Posted 10 August 2013 - 09:18 AM

How much observing have you done? If this is a starter mount, then it would be putting off quite a bit of skill building in observing for the sake of building something which may not turn out to be the sort of mount you really want.

-Rich

#21 sternenhimmel

sternenhimmel

    Vostok 1

  • -----
  • Posts: 102
  • Joined: 05 Jun 2013

Posted 10 August 2013 - 10:40 AM

Observing? A fair amount over the years, mostly with an 8" Newtonian. But my real goal is an astrophotography platform for relatively light scopes. I see the merit in buying an introductory setup like the CG-5, and going from there as I become more advanced and my appetite expands. However, I also get the feeling it would be better to invest in something more substantial right off the bat, so that it will last me years to come. If that's my philosophy, then to me it makes sense to build something myself.

Again, I'll only have access to this machine shop for another year, and I would like to become somewhat proficient in my machining skills. This seems like a good project to achieve both my goals while the resources are still at my disposal.

#22 Chucke

Chucke

    Explorer 1

  • -----
  • Posts: 96
  • Joined: 12 Mar 2010

Posted 13 August 2013 - 04:20 PM

Here is an example. 0.4" overhang. I could have made it less but assembly in the field would have been more difficult.

Chuck

Attached Files



#23 orlyandico

orlyandico

    Fly Me to the Moon

  • *****
  • Posts: 5362
  • Joined: 10 Aug 2009
  • Loc: Singapore

Posted 14 August 2013 - 12:24 AM

That's a handsome mount... home-built?

#24 Starhawk

Starhawk

    Space Ranger

  • *****
  • Posts: 5509
  • Joined: 16 Sep 2008
  • Loc: Tucson, Arizona

Posted 14 August 2013 - 09:15 AM

Sternenhimmel, the light weight astrophotography mount is nothing less than the holy grail in this hobby. The real problem comes down to tracking accuracy, which can be broken down a little unconventionally this way:

Mount:
Drive mechanical accuracy
Drive movement accuracy

Mount and telescope:
Telescope focal axis flexure (SCTs are famous for this issue)
Telescope to guidescope flexure
Telescope/ mount rigidity and natural frequency
Telescope and mount environmental disturbance response (wind, jarring).

Telescope:
Focal length
F/#

So, mount mass ends up being positively correlated with dealing with about a third of that list. What isn't obvious is picking a telescope with a longer focal length means any tracking error is easier to see in the image- and this is linear. However, f/# drives what you are required to do. So, an f/10 system can get to minutes of exposure time very easily, so even if the error is fairly low, the opportunity to damage the image is large. But image brightness is proportional to the inverse square of the f/#. And this is why very fast systems like the f/2 hyperstar exist. I invented this dimensionless brightness number to predict performance:

BR= 1000/(f/#)^2

BR (f/10)= 1000/100= 10

BR (f/2)= 1000/4= 250

So, the f/2 image is 25 times brighter than an f/10 one, meaning the magnitude it reaches in 30 seconds is equivalent to what the f/10 system needs 12 minutes and thirty seconds to do. So, the mount tracking burden is far less than it is for the slower telescope.

So, for small and light, you're talking about a light weight telescope, short focal length, and low f/#. This is a specific solution set more in the direction of camera lenses or things like the C6 hyperstar. Things like the Vixen Polarie, Takahashi SkyPatrol, or Celestron NexStar SE mounts fit in this zone depending on size.

Hopefully this will help for seeing what you would be trying to get with larger mounts.

-Rich

#25 windjammer

windjammer

    Lift Off

  • -----
  • Posts: 18
  • Joined: 02 Jun 2013

Posted 14 August 2013 - 05:02 PM

Hi -

>>astrophotography platform for relatively light scopes

yes please!

You might be interested in my upgrade projects on an EQ2 to try and achieve this (I know, quixotic). One of the projects is electronic setting circles using home made encoders which might be useful for the goto aspects of your mount - you can find all the mechanicals, electronics and software on the site:

www.astrobling.weebly.com

I get totally the pleasure of machining beautiful precision parts, and some of the designs mentioned look fantastic and lovely. At the risk of sounding heretical, a lot of this is overkill - the trick surely is to combine cheap simple mechanics with smart software and electronics to get over the mechanical shortcomings. For example, any piece of junk can be made to track accurately for as long as you like if you have closed loop guiding - autoguiding.

Simon






Cloudy Nights LLC
Cloudy Nights Sponsor: Astronomics