Topics

Help needed choosing gearing between steppers and worm drives


Robert Gast
 

I am in the process of converting an Explore Scientific Nano EXOS GQ3 mount in to a Go To and Atrophotography/Eaa setup. Im planing to build my own system with a different micro and all kinds of other stuff to see how long I can do unguided long focal length shots with software tricks and a few inexpensive sensors. I have decided I want to start with OnStep first and maybe just expand my ideas in to the project later on.

My biggest problem is that I am not understanding how to choose the gear reduction/belt drive. I see that most people using onstep and #rduinoscope are using a bout a 1:4 reduction, but I have also seen 1:18 planetary gearboxes being used. Originally I was going to build custom 1:58 reductions on the RA and the DEC. right now I am brazing aluminum mounts to temporarily use the motors in direct drive, the .9 degree nema17s I have are 24v 2amp so they have enough torque to direct drive the worm gear on a much bigger mount. I am guessing since I want unguided accuracy using the biggest reduction possible on the RA would usually be the best choice (ive got a 60K reduction box around here lol) but if I try to go to big on the reduction then the goto will be ridiculously slow even driving the stepper at max rpms right? So is 1:4 a recommended reduction or is it just what a lot of people use because the parts are easily available on the 3D printer market? Also when calculating arcsecond per step resolution. I just divide .9 by the 138 reduction of the mounts worm wheel then I divide that by whatever gear ratio I plan to use, am I suppose to also divide by the microsteps per step? Im not really sure if that counts as resolution since micro steps arent accurate or reliable?

Lastly I plan on using 16bit absolute magnetic encoders on each stepper (i have worked with plenty of DIY brushed servos, this is my first stepper project), since I will be able to to correct position and fix misses microsteps by monitoring how much the shaft moves can I get away with just going direct drive and then use micro stepping only for higher resolution? This set up would seem ideal because you could switch between full stepping when slewing and 256 micro stepping when decelerating or tracking.


Howard Dutton
 

1. Read the Wiki.
2. Figure out the mount's gear reduction (count turns of the slow-motion knobs to reach a given angular motion of 180 or 360°) if necessary.
3. Figure out how much OVERALL reduction you want, perhaps 360:1 to 1440:1 (or even higher,) depending on the motor.
4. Think about how much non-periodic error there is (which even PEC isn't going to help with) in those worm gears and are precision servos really worth the cost/effort?
Low(ish) cost stepper/servo motors have recently been discussed and these can work with step/dir signals from OnStep (as can other DC servos.)  A search of our group for "servo" probably will find some of that info.
5. Remember auto-guiding is a low cost "servo" drive system.  In-fact it's about as closed loop as you can get with the primary limitation being that it's not very responsive (exposure time etc.)
6. Micro-stepping is about incremental improvement.  2x isn't quite twice as accurate as 1x but it's close.  4x again not quite 2x but close.  8x... etc by the time you're to 16x or 32x the positioning accuracy returns have diminished to the point of being virtually non-existent but you still are suppressing vibration which is important too.  And some seem to think that the motor's torque becomes less with fine micro-stepping this is for the most part* NOT TRUE.  The "delta" (change in) torque between each micro-step is less but the overall torque remains the same.

* There are a few special circumstances WRT full-step mode where this isn't true but these seldom matter in our application.  In-fact full-step mode at low/mid RPM's (we frequently use) lowers available torque and increases vibration vs. fine micro-stepping.


Robert Gast
 

Thanks Howard I appreciate the reply and have had some confusion about microstepping cleared up. But im still not sure about my main question which is how much gearing if any I should use, I realized reading your post that I left some important info out here. This is the setup I will be running on

Exos Nano has 138 teeth
Steppers are .9 degree 400 step, 2amp 24v sanyo denkis with anti-vibration mounts and ferro fluid dampeners 
Direct Drive equates .00652 degrees per step or ~23.5 Arc Seconds per step
Telescope/Camera Res is .94 arcseconds per pixel

First off let me say yes I want to do astrophotography but mostly im doing to view things I otherwise wouldnt see without long exposure and long focal length, like small galaxies. I want the pictures to look good enough with decent detail that I study what im looking at but im not looking to make award wining photos. I want to experiment with going unguided I.E not using a guide cam, although I plan to use encoders on the motors and the mounts axises, along with an inertial measurement unit, to help accuracy and calculate any periodic and non periodic errors. These encoders chips  magnet are 5 bucks a piece from digikey so much less than a decent guide cam. I know some of this is out of the realm of onstep, as I said I will be using onstep at first or expanding on to its code base right now I just want to finish the mechanical build so I can get to the soldering iron and codeing!

Now back to the main question at hand if I go using worm sets I have on the RA to gear at 1:58 or 1:205 I will be getting decent to good resolution at .5 arcseconds or .111 arcseconds per step respectively. The thing is im not seeing anyone use more than 1:18, Im afraid if I start trying to get nuts with the gearing to raise the step resolution not only will I have to deal with any error periodic or non periodic and backlash the gears introduce but more importantly I will just kill GoTo speed to an unacceptable level.

So taking in to account I want to do long focal length unguided imaging, and that my mounts worm wheel is 138 teeth, the question is what ratio between the motor and mount gears is a nice sweet spot for accuracy and go to speed, and how high of a ratio can you use before go to becomes un usable?


Howard Dutton
 

On Fri, May 3, 2019 at 08:51 AM, Robert Gast wrote:
Direct Drive equates .00652 degrees per step or ~23.5 Arc Seconds per step
This isn't something I'd try.

Again:
3. Figure out how much OVERALL reduction you want, perhaps 360:1 to 1440:1 (or even higher,) depending on the motor.


Howard Dutton
 

On Fri, May 3, 2019 at 08:51 AM, Robert Gast wrote:
Now back to the main question at hand if I go using worm sets I have on the RA to gear at 1:58 or 1:205
The 58:1 might do very slow goto, depends on the stepper motor.


Howard Dutton
 

On Fri, May 3, 2019 at 08:51 AM, Robert Gast wrote:
Steppers are .9 degree 400 step, 2amp 24v sanyo denkis
2A rated NEMA17? stepper motors are designed for high speeds.  Our favorite stepper driver (SSS TMC2130) can't fully power them (there are other options though and you wouldn't necessarily even need to fully power them.)


Howard Dutton
 

On Fri, May 3, 2019 at 08:51 AM, Robert Gast wrote:
So taking in to account I want to do long focal length unguided imaging, and that my mounts worm wheel is 138 teeth, the question is what ratio between the motor and mount gears is a nice sweet spot for accuracy and go to speed, and how high of a ratio can you use before go to becomes un usable?
If this were my mount I'd use 400 step motors with a 3:1 timing belt reduction (provided the form/fit is workable.)  Keeps backlash low and is affordable.


Howard Dutton
 

The other high quality motor we like could get the job done fairly cheaply:
https://www.mpja.com/Stepper-Motor-1Deg-266V-Stepper-Motor-Used/productinfo/31129+MS/

These were just about gone but they must have found more surplus... they require a bit of finesse getting the gear/shroud removed but are worth the effort.


Khalid Baheyeldin
 

Robert,

I am like you, in that I have no interest autoguiding, stacking and post-processing, and have been experimenting with how far I can go with long exposures, before re-evaluating where I am and re-considering autoguiding. What I do is more akin to EAA than astrophotography. There is a very promising feature in OnStep (Full Dual Axis Compensation) that shows a considerable positive impact, and I have been experimenting with it. Still not fully 'there' yet. I also added a PEC sensor, but was unable to record it, since the easiest way to record is autoguiding, and I was not able to get that working via KStars/Ekos using the main scope and camera (nothing to do with OnStep).

Back to your questions ...

For resolution suitable for astrophotography, we generally aim for 12,800 steps per degree (or more), which is calculated for you in the spreadsheet based on the inputs. This provides ~ 0.28 arc second per step resolution. This is about half the best atmospheric limit on earth (0.4", but usually around 1" in places where humans live).

See this recent message. for some examples.

The overall reduction needed is 720:1 (or more) assuming a 400 step motor and 1/16 step, or 360:1 for 1/32.

Assuming 138 teeth, you need 5:1 reduction, and the figures in the sheet will be:

400, 1/32, 5:1, 138 = 0.147 arc second per step.

Best way for reduction is pulleys and belts.

The Sanyo Denki's that you have are overpowered at 2A. You cannot use TMC2310 with them, but there are other steppers that can do that (e.g. the S109), but you need good heat dissipation, and most likely a fan in the controller box.

There are similar 400 step motors that run at 0.9A at Steppers Online.


Robert Gast
 

Hey you guys thanks for all the information it has basically answered all my questions in one way or another, but has raised one last question lol. I see that in Khalids calculations you are including the micro stepping as part of the overall resolution. As stated earlier if the microsteps are more than ~16 the accuracy goes out the window, well how do we calculate the resolution if we are using 512 stepping in order to run the motor smoother, or is that just not a recommended practice?

I see you have both brought up my motor specs, I bought these with very little documentation and after digging through data sheets at sanyo denki I have realized that these were custom built and not part of the standard commercial line sanyo denki sells. I bought them barely used and know that there job before I got them were super hi end cnc machines. I think the 2 amps i quoted were quoted to me assuming the motors are run unipolar wich then makes there resistance equal to the .8 ohm printed on the spec sticker, using an LCR meter the coils measure a little over 1.5ohm. I have been driving them at one amp using an a4988 step stick with no troubles at all. 

@khalid, thanks for the tip full dual axis compensation I will definately research that. As far as running unguided I have a solution for pec, and other types of compensation, If you put hi resolution 16 bit abs encoders on the mount itself at each axis you can then track the movement of the mounts axis including all mechanical error and then use this to compare to your wanted position. See I plan to encode the backshaft of each stepper and then the mounts axis, along with mounting an IMU on to the telescope itself. I then plan to use a micro called a Propeller to read all these sensors and make the right motor control calculations, and implement a sort of huge pid loop between the motor shafts and the telescopes postion, I have also been reading about Vector control or FOC which sounds interesting. The propeller has 8 totally separate cores with shared ram so each motor/encoder on the system can be assigned it own fully deterministic core that can be synced to gps time. The propellers way of doing things totally removes interrupts out of the equation too, so there is no reason to mess to be off any counts at all. I also want to use the TMC5130 stepper controller too because those can actually accept input from the encoders on the backshafts. Basically I want to use onstep to make sure everything i build works and is mechanically sound before I dive in to getting more accuracy, after that I would like to somehow use the propeller for all motion control and then modify onstep to communicate with the propeller letting onstep handle all the astronomical calcs, system utils, wireless coms, etc etc,                                                


Khalid Baheyeldin
 

On Fri, May 3, 2019 at 05:52 PM, Robert Gast wrote:
I see that in Khalids calculations you are including the micro stepping as part of the overall resolution.
It is the spreadsheet calculates the resolution that way:

StepsPerDegree = ( Motor steps per rotation X microstep X GR1 X GR2 ) / 360

ArcSecondsPerDegree = 60 X 60

StepsPerSecond = StepsPerDegree / ArcSecondsPerDegree X 15

Resolution = 15 / StepsPerSecond

15 being the sidereal rate (apparent motion of the sky, in arc seconds per second).

As stated earlier if the microsteps are more than ~16 the accuracy goes out the window, well how do we calculate the resolution if we are using 512 stepping in order to run the motor smoother, or is that just not a recommended practice?
There are no drivers that we commonly use with OnStep that can do 1/512.
The most that can be done is 1/256 (TMC2130, indirectly via interpolation, and ST820 natively, but not a well used option yet).

I have been driving them at one amp using an a4988 step stick with no troubles at all. 
The question is how much torque will they provide when driven at 1A? If you experiment and find that it is adequate and they run cool enough, then our concerns are not an issue. If they don't provide the needed torque (unlikely if you use enough reduction), then be prepared to add more reduction, or use other motors.

@khalid, thanks for the tip full dual axis compensation I will definately research that.
It is pretty simple: start with a 6 star align, then go to Tracking in the Android App, and select Full, then Dual. That is all. OnStep will compensate for atmospheric refraction and alignment inaccuracies.

I am pretty sure that with short focal lengths and sensors that do not have a small pixel pitch this will be adequate, maybe with PEC maybe without.

However, with 1310 mm focal length and a Canon T4i (5184 x 3456 @ 4.3 micron pixels), it is not enough, and I am still experimenting (at a rate of one night per month, given our lousy weather). I still have to report the latest finding to Howard.

As far as running unguided I have a solution for pec, and other types of compensation, If you put hi resolution 16 bit abs encoders on the mount itself at each axis you can then track the movement of the mounts axis including all mechanical error and then use this to compare to your wanted position.
OnStep has an encoder feature via the WiFi MCU. You can read about the design from Howard's reply earlier today on another thread.

But for me, encoders are not an option. Anything that requires mechanical work is complicated and only suitable for mounts that are 'open' enough to add components on their exposed shafts.

See I plan to encode the backshaft of each stepper and then the mounts axis, along with mounting an IMU on to the telescope itself. I then plan to use a micro called a Propeller to read all these sensors and make the right motor control calculations, and implement a sort of huge pid loop between the motor shafts and the telescopes postion, I have also been reading about Vector control or FOC which sounds interesting. The propeller has 8 totally separate cores with shared ram so each motor/encoder on the system can be assigned it own fully deterministic core that can be synced to gps time. The propellers way of doing things totally removes interrupts out of the equation too, so there is no reason to mess to be off any counts at all.
Sounds like that product from Europe that was sold some years ago (maybe still in production?)
TDM = Telescope Drive Master.

I also want to use the TMC5130 stepper controller too because those can actually accept input from the encoders on the backshafts. Basically I want to use onstep to make sure everything i build works and is mechanically sound before I dive in to getting more accuracy, after that I would like to somehow use the propeller for all motion control and then modify onstep to communicate with the propeller letting onstep handle all the astronomical calcs, system utils, wireless coms, etc etc,
Interesting stuff. Please report on your progress. Always interesting to see how people overcome issues.


Butchf
 

wow nice wish i'd have seen these before.


Robert Gast
 

It should be easy to add encoder to the shafts of any mount, no? I was thinking the counter weight shaft for the DEC and for RA my mount has a plastic cap on the back of the RA housing when i pop it off there is access to the the steel rod through the RA.

Like i said the encoders are basically $3.50 for 12 bit absolute or $7 for 16 bit absolute at digikey.com, there just chips with hall effect sensor arrays in them combined in a way that you only have to center a diametrically polarized two pole magnet (90 cents also at digikey) over them, you can probably just stick them straight on the bottom of the shafts. Guiding with a camera seems expensive and overly complicated for such slow feedback. I dont think of the problem as trying to track a star, the problem is moving an object at a constant rate regardless of outside conditions, there are lots of highly accurate robotics projects dealing with the same sorts of problems and they solve it without a guide star and machine vision. I just feel like Orion selling a motorized mount for $500 dollars that cant track well enough for a 1 minute exposure without adding guiding is a bit ridiculous with all the cheap technology's available today.

Thanks for all the help, time to start gear shopping i guess.. the way my mount is made i think a belt drive is my only option on the dec axis, using a 1:5 gearhead and coupler on the RA is probably going to be the easiest thing to fabricate a mount for. I just want to get the dang gears and belts on the thing so i can be finished designing the motor mounts and welding them and move on to the acuall fun stuff


Khalid Baheyeldin
 

On Fri, May 3, 2019 at 09:57 PM, Robert Gast wrote:
It should be easy to add encoder to the shafts of any mount, no?
That is a relative thing. For me, this stuff is hard. Not so much for others.

I was thinking the counter weight shaft for the DEC and for RA my mount has a plastic cap on the back of the RA housing when i pop it off there is access to the the steel rod through the RA.
That is my point: my mount (Vixen SXD) is fully enclosed and there is no way I can attach something as big as a high resolution encoder to the RA axis. Some mounts are easier, perhaps the Losmandy G8 and G11, since much of them are exposed. For other mounts, not so easy.

Guiding with a camera seems expensive and overly complicated for such slow feedback.
I agree, but you and me are a minority.

For me, it is just another component that can go wrong in an already complex system. Astronomy mounts intersect too many disciplines: optics, mechanics, electronics, ...etc. How many times do you go out on a rare clear night, only for something to go wrong? Be that power issues (e.g. GFI tripping for no apparent reason), a stiff cable in the cold, a software or firmware update that introduces a bug, ...etc.

So, for me, I don't want more components (camera, telescope, adapters, software, settings, ...etc.) that can go wrong.

I dont think of the problem as trying to track a star, the problem is moving an object at a constant rate regardless of outside conditions, there are lots of highly accurate robotics projects dealing with the same sorts of problems and they solve it without a guide star and machine vision. I just feel like Orion selling a motorized mount for $500 dollars that cant track well enough for a 1 minute exposure without adding guiding is a bit ridiculous with all the cheap technology's available today.
One minute exposures with my OnStep mount and 1310 mm is not a problem. Even 120 seconds is often not a problem. You can see some such single exposures that I made in this album. It is 180 seconds where issues are showing for me.

Thanks for all the help, time to start gear shopping i guess.. the way my mount is made i think a belt drive is my only option on the dec axis, using a 1:5 gearhead and coupler on the RA is probably going to be the easiest thing to fabricate a mount for. I just want to get the dang gears and belts on the thing so i can be finished designing the motor mounts and welding them and move on to the acuall fun stuff
Here you can find pictures of how I did it, including the pulleys and belts.

The showcase page has links to many other conversion where you can get ideas.


Robert Gast
 

those pics are great for unguided unprocessed at 1000mm+ focal length especially orion! I saw your mount, all you have to do to add the type of encoders im talking about (not huge optical ones) is stick a round diametric neodymium magnet at the tip of your dec shaft, on the piece that stops your counter weight from sliding off and then mount a small pcb about 2mm away from the pcb, same procedure for however you access the metal rod and bolt through your RA shaft!.

Im finding out 5:1 gearheads cost more than i thought unless you buy with the motor for same price, and going the 4:1 pully route isnt going to be easy on my DEC shaft either there just isnt space for anything more than a 36T pulley unless i use a u joint and some other ridiculous mechanical trickery ARGHHHH this mount positioned the DEC shaft in the worst place ever, i wish id stuck with an eq5 but i had to be different!


Howard Dutton
 
Edited

On Fri, May 3, 2019 at 06:57 PM, Robert Gast wrote:
Like i said the encoders are basically $3.50 for 12 bit absolute or $7 for 16 bit absolute at digikey.com, there just chips with hall effect sensor arrays in them combined in a way that you only have to center a diametrically polarized two pole magnet (90 cents also at digikey) over them, you can probably just stick them straight on the bottom of the shafts. Guiding with a camera seems expensive and overly complicated for such slow feedback. I dont think of the problem as trying to track a star, the problem is moving an object at a constant rate regardless of outside conditions, there are lots of highly accurate robotics projects dealing with the same sorts of problems and they solve it without a guide star and machine vision. I just feel like Orion selling a motorized mount for $500 dollars that cant track well enough for a 1 minute exposure without adding guiding is a bit ridiculous with all the cheap technology's available today.
Yea using an on-axis (RA axis) encoder to correct tracking is a long-term pet project of mine too and a 16bit encoder isn't going to cut it (even if its accuracy were at that level which I seriously doubt)...

https://groups.io/g/onstep/topic/26142320#5715


Khalid Baheyeldin
 

On Sat, May 4, 2019 at 12:34 AM, Robert Gast wrote:
those pics are great for unguided unprocessed at 1000mm+ focal length especially orion!
The Orion Nebula is easy. The sweet spot for it (without stacking) is around 90 to 100 seconds.

I saw your mount, all you have to do to add the type of encoders im talking about (not huge optical ones) is stick a round diametric neodymium magnet at the tip of your dec shaft, on the piece that stops your counter weight from sliding off and then mount a small pcb about 2mm away from the pcb, same procedure for however you access the metal rod and bolt through your RA shaft!.
The problem is the RA, which is fully enclosed and barely any room to add anything. Even if it was 'easy' for regular people, I am not that handy so more of a challenge.

Im finding out 5:1 gearheads cost more than i thought unless you buy with the motor for same price, and going the 4:1 pully route isnt going to be easy on my DEC shaft either there just isnt space for anything more than a 36T pulley unless i use a u joint and some other ridiculous mechanical trickery ARGHHHH this mount positioned the DEC shaft in the worst place ever, i wish id stuck with an eq5 but i had to be different!
How about one of these precision gearbox steppers? Someone posted in the group that they have zero backlash. You can do the motion transfer with 1:1 pulleys, and even a 15:1 will give you 0.19"/step resolution (at 1/16 microstep).


simingx@...
 

Hi,
The planetary gearboxes don't have "zero" backlash, I think the specification is <1 arcmin. 


George Cushing
 

This mount is very similar to the generic EQ-3M and the Celestron CG-4. There have been a few such conversions discussed here. There are more examples that used Meade servos on the Yahoo Roboscope group. These systems are proven and relatively inexpensive. Your proposal seems overly complicated and at least in the time involved quite expensive particularly when one considers limited capability of the EQ-3 mount. 

Go2 robotics have been developed over the last 50 years and at this point are fairly highly developed. It took about ten years from the development of the first microprocessors until the first commercial Go2 mount was introduced. It cost $20000-30000 in constant dollars. Today much more capable commercial products sell for under $1000. 

There are physical constraints that limit the performance of any telescope system. For example, there is a limit to the resolution of any optical system, due to the physics of diffraction. The physics of the atmospheric causes distortion or turbulence that distorts telescopic images by causing random phase shifts in the incoming light. These distortions often reduce the resolution of a telescope far below its diffraction limit. Mounts built with the OnStep system can have specifications that exceed those imposed by these constraints. While you may be able to build an exquisite system, there is no way around the reality of physics. 


Robert Gast
 

@Howard, disregarding accuracy specs, you dont think 16bit absolute encoders are enough on the shaft to correct for periodic and non periodic error? That allows for a comparison against expected and actual postion every 18 arc seconds. What resolution would you say is needed, i would think any system above 16 bits is going to start having issues with signal to noise ratio.

@Khalid, Those motors you linked to are the exact gearboxes i was talking about, i had seen the gearboxes themselves for like $15 dollars from a US shipper but can not manager to find the seller again. I looked closer at my mount and cut some circles out the size of gt2 pullys turns out i can fit the 60T ones on my dec axis after all i just have to clever about installing it with the worm removed from the mount. So for now im going to go 4:1 belt drive since i can easily and cheaply get the pullys from a US seller and already have belts I can splice on hand. Maybe later I will experiment with gears and diy gearboxes or different pully systems where i can get a higher ratio. It would acually be even cheaper and easier to get 10:1 worm sets but they will just take much longer to acually get installed and will still end up probably needing a 1:1 pully or u joints and shaft to get the correct spacing.

I looked at TDM (Telescope Drive Master) mentioned earlier and it looks like they are using a lot of the same ideas i planed to implement long term, For the life of me though I can not seem to find the specs on that expensive encoder they use, is it optical, whats its res etc. I am pretty sure using onstep as a software base (im interested in using onstep for everything but guiding basically), it should be more than possible to develop a system as good as if not better than TDM that anyone with soldering skills can put together for less than $150 dollar depending on how carefully you source your motors/gears/belts. My end goal is to get a decent 10 minute exposure unguided.

As far as modifying an EQ3 mount ill tell why I went down this path.. an Exos Nano can be picked up for $99, this was the cheapest mount i could find new (easily accessible to anyone wanting to copy my work from scratch) that is not a hunk of crap. I want my system to work on eq2 decently in the end to tell you the truth. The nano mount is actually very stable and well built for loads of 15 to 20 lbs, the tripod not so much... 10lbs at most and could be more stable. The point is you can take that eq3 head and mount it on a pipe and have a very sturdy mount that can support a smaller newt astrograph or in my a case small sct and small refactor with dslr and other accessories you may use, and to boot will be absolutely excellent for a ccd/dslr with a nice heavy telephoto even on supplied tripod. 

I think commercial manufactures of GoTo mounts like Meade, Orion, Celestron etc are selling people stuff that cost more than it should because they are still thinking of things in a traditional sense. Obviously a mount will cost more when you build it using precision mechanics and go beyond the techniques of mass production to ensure it is of a higher standard. Quality expensive mounts are GREAT and there pretty, I most defiantly would not turn down a 5000 dollar Meade or a Paramount! I believe that these days you can skimp on the mount and don't have to pay 4Gs for a byers worm wheel and clutch, you can pretty much substitute quality machining for cheap to moderately priced sensors, good software, and the right microprocessor architecture for the job. I ask myself if I can build a good GPS .5ppm disciplined frequency standard using a 3 dollar e-bay gps module and a 2 dollar SI5351 chip that generates square wave clocks, why are people still clocking processors used in telescopes with 50ppm crystals... probably because they spent to much trying to machine the perfect worm drive. Why do we care if a gearbox on the mount has backlash? Its pretty darn easy to run a motor both ways real quick and measure backlash with encoders the compensate for it in software. I am coming at this project with a background in electronics and hobbys of robotics and and radio astronomy (actually i got in to optical to eventually overlay it with radio data for wide spectrum pics), so when i look at a mount and throw electronics at it im thinking of it more as a real robot than the more traditional motorized mount paradigm.

Sorry guess I kind of went off on a tangent