DIY motorized horseshoe mount - help please

[Osprey]

I am in the design phase of building a horseshoe mount.  End result I want to achieve is a motorized control / goto type system.  I would like to utilize a raspberry pi to be the brains but will not rule out a 2560 Mega.  So far, I have collected the following information:

Nema 17 stepper will make 200 steps per full revolution

Shaft diameter is 5 mm

Earth rotates 360 degrees in 23h 56m 04s or 15.041085 degrees per hour.

 

Armed with this minimal amount of information, first: is there software that already exists for horseshoes so I do not have to go through all the math?

Second: If not, how would I find out how many steps / revolutions would it take to equal 15.041085 degrees?

Third: How different is solar tracking from "regular" tracking?

 

The refractor I have built is to be used for solar observing as the optics are only school laboratory grade, 2 scratches on the main objective.  I'm not very far into this journey and while I am more engineer than computer whiz...I still try.  Help appreciated

[fwm891]

@Osprey You need to first break it down to give you the number of seconds in a sidereal day:

23 x 60 x 60 = sec in 23 hours = 82,800

Plus 56 x 60 = 56 mins in seconds = 3,360

Plus the odd 4.0905 sec

Total = 86,164.0905 seconds in a sidereal day which is the length of time the Earth takes to rotate 360° or 234.34469 sec per 1° = 3.989078 mins per 1° rotation

Your stepper motor will give 200 steps per minute = 47,868.9391666 steps

If you drive your horseshoe directly from the 5mm diameter shaft at 200 steps per minute then:

5 x pi = 15.70796 mm surface travel per shaft rotation in 1 minute.

There are 3.989078 mins per degree rotation = 62.6602776 mm travel per 1° x 360 for full rotation = 22,557.6999 mm or 22.5577 metres for your horseshoe circumference

Divide by pi = 7.180.34 meters diameter.

You will need to gear down the motor by some factor to bring the mount to reasonable proportions or use micro stepping...

Hope that helps...

Edited January 26, 2020 by fwm891
text added

[Osprey]

16 hours ago, fwm891 said:

Total = 86,164.0905 seconds in a sidereal day which is the length of time the Earth takes to rotate 360° or 234.34469 sec per 1° = 3.989078 mins per 1° rotation

Your stepper motor will give 200 steps per minute = 47,868.9391666 steps

@fwm891 I need to ask where you came up with 47,868.9391666 steps?  If I divide 47,868.9391666 by 200 steps/min, I get 239.344696 which I assume should be 234.344696 sec/deg.  Please clarify for me

Bill

[fwm891]

@Osprey Oops see below...

Typo in above... Total = 86,164.0905 seconds in a sidereal day which is the length of time the Earth takes to rotate 360° or 239.434469 (original calc showed as 234.34469 sec) sec per 1° = 3.989078 mins per 1° rotation

3.989078 mins per 1° rotation  then 360° = 1,436.06808 mins for 360°

If your motor does 200 steps per min. Then for 360° rotation = 1,436.06808 x 200 = 207,213.616 steps per sidereal day

Sorry about the typo

 

[Osprey]

@fwm891 lol...no worries, all good  I am currently documenting all my work as to upload to my website...

[Osprey]

5 hours ago, fwm891 said:

@Osprey Oops see below...

Typo in above... Total = 86,164.0905 seconds in a sidereal day which is the length of time the Earth takes to rotate 360° or 239.434469 (original calc showed as 234.34469 sec) sec per 1° = 3.989078 mins per 1° rotation

3.989078 mins per 1° rotation  then 360° = 1,436.06808 mins for 360°

If your motor does 200 steps per min. Then for 360° rotation = 1,436.06808 x 200 = 207,213.616 steps per sidereal day

Sorry about the typo

 

@fwm891 I think I found another typo.... 1,436.06808 x 200 = 287,213.616 not 207,213.616 steps / sidereal day

[fwm891]

😀

[Andy_ZH]

Hello Bill,

as I wrote in the other thread, I am also tinkering with my mount, therefore I'd like to join the discussion.

I think something is missing in the data that you gave. I think for sure, you cannot drive your mount directly from the shaft of your Nema17 stepper motor, and this is probably the reason why the step count in the calculations above is so little. Horseshoe mounts are sometimes driven on some kind of tension wheel, and if this is the case for your mount, you have to include the size ratio between the horseshoe circumference and the driving wheel.

> Armed with this minimal amount of information, first: is there software that already exists for horseshoes so I do not have to go through all the math?

Except of considering the mechanics, there should be no difference between the horseshoe mount and any other mount , es e.g. a fork mount. Therefore, you could use any software that you can find online, e.g. onestep.

 

> Third: How different is solar tracking from "regular" tracking?

As mentioned above, there is nearly no difference except a little higher speed of the sun. Please note: the moon also has its own speed.

But I would not try to calculate the accurate number of steps in too much detail beforehand. Usually, you can fine-tune this later when your hardware is ready. A rough estimation would be sufficient.

In my case with the Vixen new Atlux, the calculation is: 180 teeth, driven by a worm, with a 2:1 pulley / belt results in 1 degree per stepper motor rotation. With 200 steps and some micro steps, I am just a the (my) limit of accuracy. Another degree of freedom is that we could use a 0.8 degree / 400 step motor. But also as mentioned above, in case you don't have a worm that drives your mount, you probably need to have an additional gear to increase your stepper accuracy.

For direct driven 3d printer extruders, you can get for very little money stepper motors with attached gear that would probably fit.

I made good experience with Trinamic 2209 drivers, they are also excellent when driving from a Arduino. They have less on resistance which reduces the necessary cooling of the driver.

Could you please add some kind of picture or drawing of your mount? From books and pictures, I remember that real horseshoe mounts are used by dobsonians, and I cannot imagine a refractor sitting in the horseshoe. Usually, the scope is not accessible at the horseshoe end, and that's why it was meant for dobs.

 

Best

Andy

 

 

[JamesF]

I have a feeling that @skybadger has built something like this.  Might be misremembering though.  If not perhaps he may be able to offer advice.

James

[Osprey]

@Andy_ZH I really wasn't sure how to drive this mount yet...the biggest variable is to be product availability vs cost.  Are geared cogs and tracks readily available at a reasonable cost vs. a toothed belt drive system?  As stated, I'm still in the designing phase, catching up on documentation.  Was looking at some grooved pulleys, like 3d printer ones, earlier today to get an idea on how to design / construct my mount.  I also bit the bullet and went an purchased the sheets of plywood necessary once the drawings are to my liking.

All said...here's what's in my mind:

I would like to utilize high torque stepper motors to either drive a belt system (preferred) or a worm / track system. How is still unknown at this time...

I would like to utilize as much 3d printing into this setup as I can.  I can hold my own with SketchUp Make and 3D Builder.

I would like a raspberry pi to handle the computerized portions but will give way to an arduino / combo if necessary. (Will learn code if necessary - am aware of Indi and KStars)

As for anything above and beyond...I am putty armed with computers and a 3d printer lol.

Attached are the images you requested...the mount is still in the design phase, I need to add in the bars to support the side bearings of the scope...coming soon

H.E.Li.O.S is a 6.5" refractor.  The dew shield, whole focusing assembly, and side bearings have been 3d printed by my Anet A6.  I had to modify several pieces of the focusing assembly from their original versions I found on thingiverse (https://www.thingiverse.com/thing:3586112).  

So there's where I sit...I know I still have a bit of road to haul but "living is all in the journey"

 

@JamesF I will give him a lookup, thanks for the tip!!

[skybadger]

Hi, my name was mentioned ?

I used a DC friction drive on my horseshoe, with incremental encoder for speed control. Big drive wheel, eventually use pec, should be nice and smooth. Sadly not quite there yet, have the drive unit and motors but destroyed the mount in a classic wheelbarrow accident. 

Will be reassembled eventually.

I agree withe observation that a refractor is going to be an interesting choice !

 

Mike

[Osprey]

Will keep everyone up to date, including photos as I go.  @skybadger friction drive...would that be like a rubber-ish type wheel directly in contact with the edge of the shoe?  I've seen images like that while figuring out my own design but the question remains in my head: Friction more / less reliable than a toothed belt?  I understand that if built properly, both should be close to equal, each having their own pros / cons.

[lenscap]

Hi Osprey,

As mentioned above, Onstep works with Alt/Az as well as EQ mounts. Have a look at the Wiki;  https://onstep.groups.io/g/main/wiki/home

I suggest you apply to join the Onstep forum as well. There is lots of useful info in both.

The Wiki includes a spreadsheet where you can decide the level of tracking precision you want, and try out lots of gear ratios & motor steps & micro steps to see what works.

Onstep does not run on the Pi.

It does run on the Mega2560 , the STM32 & other commercial micro-controller boards as well as a choice of PCB's designed by the developers.

You say you want tracking & goto.

Learning to write code to give a choice of say, Siderial, Lunar & Solar tracking speeds is straight-forward.

Writing software to enable a mount to star-align & goto is orders of magnitude more complex than just tracking. I think you would need to achieve at least a semi-professional level of coding ability to get a result, in a reasonable time, that works well & can interface with a handset or laptop by usb, or perhaps a phone , maybe by wifi. No coding is required with Onstep. The developers provide it free. Just download, configure, run.

Whatever approach you choose , good luck & enjoy the journey.

Edited January 28, 2020 by lenscap

[skybadger]

Osprey, I use a toothed belt, smooth side facing the horseshoe and toothed side facing the motor sprocket. That way I get a long friction surface rather than a single point. The wheel is supported by two support bearings, over which the belt runs ,so that is where most friction will be experienced. Seems to work nicely by hand...

I too recommend onstep, but last time I looked it couldn't handle DC servo motors.

[lenscap]

2 hours ago, skybadger said:

I too recommend onstep, but last time I looked it couldn't handle DC servo motors.

Correct. Onstep is a stepper motor controller.

2 hours ago, skybadger said:

I use a toothed belt, smooth side facing the horseshoe and toothed side facing the motor sprocket.

Out of interest, how do you do that?

[Andy_ZH]

Hi Osprey,

 

thanks for the pics. I have some more questions. Sorry if I seem to be pedantic, but I just cannot imagine some things:

You are aiming for solar observing. The sun runs on the ecliptic. But the (e.g. at equinox), when you scope points to the ecliptic, the scope will be within the plane of your horseshoe. Do you have enough space there? Without mirror, you will not be able to access the eyepiece point. Even worse, in summer when the sun is above the sky equator, your observation point will be between the horseshoe and the floor. Or did I get something wrong? I would have guessed that for a refractor, a fork mount would be more appropriate.

> Writing software to enable a mount to star-align & goto is orders of magnitude more complex than just tracking.

Well, yes, there is some more coding involved, but this is also no rocket science, but some more advanced geometry calculations. The difficulty comes when you try to do too much within a microcontroller (Arduino) where C is the only option of programming language.

I described my planned Raspberry Pi project here:

if you like, I can share my code with you. It is not yet ready for open-sourcing, but I can drive two stepper motors quite precisely. There is no need for an Arduino, my software runs just directly from the pi and it simply needs two 2209 drivers.

Your refractor seems to be quite fast. Will you use some H-alpha filter with it? If not, I wonder if it will have an enormous chromatic aberration.

Friction drive (without teeth): This would be ok in case you just do visual observations. But since there are no teeth on a friction drive, your tracking / goto could drift (=slide) away over time. Therefore, you would also need some encoders in order to get feedback of your actual position. I believe with a pure friction drive, stepper motors don't make sense anymore since you are loosing the advantage of the stepper with its discrete steps.

How about the following suggestion: Use a threaded bar to drive the horseshoe. With your wooden horseshoe, after applying some pressure and driving the horseshoe with the threaded bar for a while, it will grind its own thread into the wood. For instance, a M12 threaded bar has 1.75 mm per turn, which would give you your necessary gearing ratio. E.g. with a horseshoe circumference of 1.75 meters, you would have 1000 motor turns per day, which should be perfect.

Best

Andy

[Osprey]

54 minutes ago, Andy_ZH said:

Hi Osprey,

 

thanks for the pics. I have some more questions. Sorry if I seem to be pedantic, but I just cannot imagine some things:

You are aiming for solar observing. The sun runs on the ecliptic. But the (e.g. at equinox), when you scope points to the ecliptic, the scope will be within the plane of your horseshoe. Do you have enough space there? Without mirror, you will not be able to access the eyepiece point. Even worse, in summer when the sun is above the sky equator, your observation point will be between the horseshoe and the floor. Or did I get something wrong? I would have guessed that for a refractor, a fork mount would be more appropriate.

> Writing software to enable a mount to star-align & goto is orders of magnitude more complex than just tracking.

Well, yes, there is some more coding involved, but this is also no rocket science, but some more advanced geometry calculations. The difficulty comes when you try to do too much within a microcontroller (Arduino) where C is the only option of programming language.

I described my planned Raspberry Pi project here:

if you like, I can share my code with you. It is not yet ready for open-sourcing, but I can drive two stepper motors quite precisely. There is no need for an Arduino, my software runs just directly from the pi and it simply needs two 2209 drivers.

Your refractor seems to be quite fast. Will you use some H-alpha filter with it? If not, I wonder if it will have an enormous chromatic aberration.

Friction drive (without teeth): This would be ok in case you just do visual observations. But since there are no teeth on a friction drive, your tracking / goto could drift (=slide) away over time. Therefore, you would also need some encoders in order to get feedback of your actual position. I believe with a pure friction drive, stepper motors don't make sense anymore since you are loosing the advantage of the stepper with its discrete steps.

How about the following suggestion: Use a threaded bar to drive the horseshoe. With your wooden horseshoe, after applying some pressure and driving the horseshoe with the threaded bar for a while, it will grind its own thread into the wood. For instance, a M12 threaded bar has 1.75 mm per turn, which would give you your necessary gearing ratio. E.g. with a horseshoe circumference of 1.75 meters, you would have 1000 motor turns per day, which should be perfect.

Best

Andy

Andy, 

I plan on having a digital camera involved.  Currently I have a G3 camera from Orion that does quite well with the little 80mm I retrofitted.  I have a mylar type filter on the 80 mm. 

With my web degree, I know the basics of C and C++.  How they pertain to real life programming may be a little sketchy.  Nothing Google or Youtube hasn't been asked yet....lol.  I would be greatly interested in seeing your code.  Maybe I can learn more and contribute something towards the cause some day?

Yes, the refractor is very fast...objective FL is only like 200 mm if I remember right and the secondary lens is around 180 mm.  All told, the length of the notch in my horseshoe will be around 975 mm (38 3/8 approx).  As for the aberration...most likely because these were only school lab quality lenses.  Nothing you would find from a custom lens retailer.  There will be limitations not only by quality but also by this being my first full build.  This is why I chose solar vs deep sky.

Your idea of letting the threaded rod cut its own grooves is quite interesting...the mount is to be made from plywood but who is to say that I cannot find some maple or oak laying around and steam it to fit the curve? My party and I can cry if I want to lol!!! (ok, that remark shows my age lol)

Bill

[skybadger]

I worry about this advice.

Friction drives are chosen because it's easy to make 'perfect' circles. The overall accuracy is likely better not worse than a geared wheel. 

The choice of whether to drive a stepper or a servo is purely about your ability to measure and manage the speed, in your case either will do fine, for geared or friction drives. There is no difference in 'drift' susceptability. 

Use of a 'pressed'' gear like you describe has been successful but I wouldn't do it into the end grain of a sheet of ply... is that where the strip of oak is envisaged ? There will be much more variation than routed disk on bearing.

 

Edited January 29, 2020 by skybadger

[Osprey]

4 hours ago, skybadger said:

Use of a 'pressed'' gear like you describe has been successful but I wouldn't do it into the end grain of a sheet of ply... is that where the strip of oak is envisaged ? There will be much more variation than routed disk on bearing.

@skybadger Yes, I would mold and trim the leading edge of the plywood in either maple or oak...if this was the method I chose to use.  Today, my goal is to fabricate the circle jig for the mount and bonus would be to cut out the first few pieces of the mount. Being a full time caregiver...my days are very fluid...goals turn to dreams at times.

[skybadger]

You can see mine at www.skybadger.net if it helps. Should have said earlier. I need to dig out a photo of the update drive train described above.

[Osprey]

@skybadger Not to nitpik but you have a lot of broken links within your site...possible removed pictures or incomplete hyperlinks.  Tried to look at your mount but the hyperlink was broken...all I could view was the thumbnail. Just so you know

Bill

[skybadger]

Fair point, I don't get a lot time to maintain it and know my way around. 

This should work, I viewed it last night. 

https://www.skybadger.net/equipment/worthing12inchscope.shtml

[Osprey]

Here are some update photos of my build.  Main wheel is 48" in diameter and the smaller is 15".  Please pay no attention to the mess...it IS a garage after all 😛 I'm thinking that I will finally get to use that 1" flat aluminum stock I bought last year for my observatory...use it to line the bigger wheel edge.  This will help out with minor imperfections in my router work.  I had to overcome a router phobia as almost a year ago, my Craftsman router flipped out of my hands and landed on my right thigh - still spinning.  24 stitches later and many people scratching their heads about how a shark would end up in Lake Huron...I was home, shoveling snow (the router ripped my jeans, slowing the bit, but by the time it was over, I had what appeared to be a perfectly shaped shark bite - upper and lower jaws - on my thigh.).

[Mick J]

Coming along, loving the block supports and string/strap arrangement to the ceiling, router story sounds horrendous, big power tools eh.  

[Osprey]

11 hours ago, Mick J said:

loving the block supports

@Mick J LOL...don't forget the barrels...they need love too!!! #Barrels lives matter #LMAO

 

Not as horrendous as it sounds.  The bit only cut fatty tissue...but close enough that I could see the muscles...didn't bleed much.  Think I was in shock because the only thing that really hurt was when the P.A. was giving me lidocaine shots...those hurt the most.  Over all...great story to sit around the table and drink coffee to lol

 

*Lake Huron is all freshwater.  This even happened Feb 2019 - plenty of ice still on the big lakes... no sharks were harmed in the telling of this story*