Stepper Motor System / Advice

[Spanrz]

I have got the itch to get a setup for controlling the Az/Alt of the 12" by the stepper motors.
I've read here, there and everywhere about all the systems but still a little dazed at what system will do me, an off the shelf system or a custom built setup.

I have a plan / project to make a system up, but need some advice on my methods of my madness. The plan isn't defined, but I am not at all concerned if it takes a few months to figure out.

Now, I am very noob-ish, to these systems, though I am electronically minded to make a system, but this is what I want to do :
Have a stepper motor / control system to guide on DSO's / Planets etc, you know, the normal stuff.
Also want to have the ability to speed up or slow down both axis's (er or is it axi ?? :shrug:), as I want to have the ability to follow a satellite or other feature that goes slower / faster than sideral time. (e.g follow the ISS or shuttle in orbit or eventually show the shuttle's entry to earth's atmosphere)Watched a few vids on youtube too, just to get an idea as well.
Stepper motor system - http://www.youtube.com/watch?v=DYG8SlrwGHM
Moon guiding - http://www.youtube.com/watch?v=k-O5H2drEsw )

Now please correct my noobishness, but is this what I might need?

2 x stepper motors, a variable controller (either controlled standalone or by PC), wiring, gearing system (or a method to drive the mount), batteries.

Do I need a PC / Laptop (as I already have both)? Will the PC / Laptop take away the complexity of making a variable box on it's own or is it better to use both in conjunction?

Now the variable controller, I want the adaptibilty to switch between sidereal time to whatever speed I choose.
Like using 2 potentiometers and stuff.

Now if all else fails, and I choose to throw it all away, that I might choose an Argo Navis or similar system.
Is the Argo system encoder a motor? Or just a digitial sensor (where you still have to move the mount manually?)
I'm just confused at exactly what the argo navis system is. (automatic or manual)

Thanks.

[Barrykgerdes]

Hi

Stepper motors are usually limited in what they can do. They don't have the speed range of the more common electric motors with encoders. If you make the steps small enough for smooth tracking the slew speed will probably not be fast enough for satellites although they can be quite precise in their positioning.

There was a stepper motor drive system quite popular 12 years ago called the DOB 11 driver system or Roboscope. Used for driving Dobsonian mounts I believe they are still available but quite expensive. I had one and did a lot of experimenting with it until I got a LX200 which was far superior in its control.

I wrote a program in Qbasic to drive it complete with goto, two star alignment and catalogues of ngc, IC an Messier objects and a selection of stars for alignment. The method of imparting drive to the axes was by belts or friction drive and was suseptible to slipping. I also had a version of this program that used to drive stepper motors salvaged from old hard MFM drives. I found one variety that would accept up to 3000 half-steps/sec from a stepper drive PC board kit from Silicon Chip. Incidently the steppers that came with the DOB 11 ran at up to 5000 half-steps/sec.

I am not very familiar with the Argo Narvis system but If you put decent ball or roller bearings with some preload on all the rotating sections and a positive method of drive by toothed belts or gears I think it will turn out more practical and cheaper in the long run.

On my web site there is some old files of the work I did years ago in beefing up a Dobsonian Mount and fitting of a DOB 11

Barry
www.geocities.com/barrykgerdes (http://www.geocities.com/barrykgerdes)

PS. By the way the plural of axis is axes

[bojan]

Have a look at Mel Bartel's system (100% freeware if you know how to solder).
I build it and it works perfectly - Goto, Push-to (with encoder from mouse or better ones, with David Ek box - also freeware in principle), data base with objects.. you name it.
http://www.bbastrodesigns.com/BBAstroDesigns.html#Computer_Operat ed_Telescopes

We live close it seems so we can even arrange a demo :-)

[Rod]

Hi Barry,

Are you aware of Mel Bartels stepper drive system? You can find details here:

http://www.bbastrodesigns.com/BBAstroDesigns.html#Computer_Operat ed_Telescopes

I built the stepper drive for my 12 inch dob and it works well, although I am still ironing out some issues with goto. You can get excellent tracking, goto and slew speeds are usually between 1.5 and 3 degrees per second. You do need an old laptop for this system but the software is free and there is a very good yahoo group that can assist you. Mel is very active in his support of those who build his system.

A friend did the electronics for mine and I made all the mechanical parts. I just made moulded epoxy gears for the reduction system. They work very well visually.

Hope that helps,

Rod.

[bojan]

BTW, it seems the interest for gathering in Melbourne is again on (well, one person for now :-)
http://www.iceinspace.com.au/forum/showthread.php?t=26200&page=4
So, perhaps, this may also be the opportunity to meet and have a look :-)
Rod, are there any plans for gathering at place near you in 3 weeks ?

[Rod]

Hi Bojan,

I am sure we could arrange to catch up at the Briars (in Mount Martha) on Friday Feb 27. On the 28th is the Messier star party which the ASV is holding at Heathcote. You would be welcome to meet up there too - great skies but more traveling. It would be a good chance to look at a lot of scopes too.

Rod.

[Barrykgerdes]

I had a quick look at that site and the method appears to be very similar to what I did in 1994. My main difference was that my stepper drives used a computer program to drive discrete components rather than those special chips. My budget was a lot less than the prices listed. Driving the steppers is a function of the sequence of how you excite the fields and was done via the printer port with two four bit words.

This needs a faster computer than the old PC I was using as the control loop was quite large. Later With 386 and 486 computers speed was no problem.

The program to drive the DOB 11 originally only sent commands to the controller at one per second which at high resolutions made the scope move up a stair case. I developed a method of sending commands every 100ms which was a great improvement.

I used parts from the junk box and salvaged stepper motors. I went through lots of motors looking for the ones that could drive the fastest and at one half step for each second of rotation I managed to get the fastest slew to just under 1 degree/second.

I wrote my own software and the source is free to anyone who wants it although it is written oin Qbasic and is a bit archaic by modern standards.

My main problem was getting suitable mechanics to drive the scope. I experimented with all sorts of devices using parts mainly salvaged from old computers and printers.

I spent many months refining software and mechanics but before I got everything fully working. I had a loan repaid and bought a LX200 10" classic. The stepper drive system and 8" Dob was given to my daughter.

The system worked OK from a laptop although at the time the only hand controller I had was via the games port which left a lot to be desired in speed of operation.

I occasionally get the old test board and motors out and hook it up to the computer and do dummy alignments and gotos. The motors look good slewing and tracking. Oh did I say how I tracked. I re-calculated all the coords every 100ms in the program loop at sidereal rate and sent them to the motors.

It was a lot of fun but I prefer my LX200 these days.

Barry

[Spanrz]

Thanks for the replies. Wow. :)

Barry, thanks for the advice. I've programmed in Basic a little myself on the C64, but it's been a long time between drinks.
Hehe, I was being a bit jovial with the plural meaning of "axis", I just thought it sounded well, calling it "Axi".

Bojan and Rod, That link is fantastic. (Mel Bartel's). Now why didn't I just search on Mel.:doh:

I just needed to quell my curiosity, just in case I was going to make an idea that was simple, but in practice, got a little too complex for my liking.
Again, the idea, is sort of to do what Nasa does when they film the shuttle going up (how they track the shuttle in the middle of the view). I know how complex that can get, so I'm trying not to aim too high in a build aspect.

I'll have a good read over that link. Great starting point. Thanks.

[gary]

Hi Brett,

Gary Kopff from Wildcard Innovations here.

On its own, Argo Navis does not directly drive stepper or servo motors.

Instead, it interfaces to a pair of devices known as optical encoders which
are fitted to each axis of the mount.

As the mount rotates, the encoders provide a stream of electrical pulses
to the Argo Navis unit such that it can keep track of the position of the mount.

In order to use Argo Navis, you align on any two objects you know, typically
a couple of bright stars. Argo Navis then can keep track of where you
are pointing in the actual sky. In typical use, one then dials up an object
from the in-built catalog of objects and the Argo Navis display shows you
the amount in degrees by which you need to push the scope in both axes in
order to acquire the target. The display updates in real-time as you move the
scope so that when the numbers read 'zero' you can look through the eyepiece
and there is your object.

So, in a nutshell, one can use Argo Navis in conjunction with a pair of encoders
and then manually push the scope, using what John Dobson would refer to
as 'yoghurt power'. Some refer to this mode of operation as PUSHTO.

Having said that, Argo Navis can interface to a couple of the popular third
party servo motor solutions on the market, namely the ServoCAT by
StellarCAT (http://stellarcat.com) in Arizona and the Sidereal Technology (Sitech) (http://www.siderealtechnology.com/) controller by Dan Gray in Oregon.
The Sitech Controller is the one on the left-hand column of the cited web page -
http://www.bbastrodesigns.com/BBAstroDesigns.html#Computer_Operat ed_Telescopes

Keep in mind that the Sidereal Technology controller, despite some confusion
out there, *is not* what is referred to as the 'Mel Bartel's system', which is the
one that appears in the right-hand column of the aforementioned web page.
It just happens that BBAstroDesigns Inc, which is a company owned and
operated by Mel in the United States, is a dealer for the Sidereal
Technology controller as well.

Sidereal Technology's own web site is at -
http://www.siderealtechnology.com/

These days, the better commercial telescope control systems employ servo motors
rather than stepper motors owing to the considerable advantages servos
have, including a higher dynamic torque range.

The older Bartel's system is a stepper-motor based system.

When interfaced to either the ServoCAT or Sidereal Technology controller,
Argo Navis performs all the object offset and tracking rate calculations on
their behalf. Simply dial up an object on the Argo Navis and then press the
GOTO button on the motor controller hand-pad and you have full slew and
track capability.

Keep in mind that on an Alt/Az mount, when tracking, both axes
are moving simultaneously and the tracking rates are continually changing as
you track the object.

An example of a Sidereal Technology controller used in conjunction with the
Argo Navis is the SpicaEyes scope by Tom Osypowski of Equatorial
Platforms. See http://www.equatorialplatforms.com/spica.eyes.slipstream.html

An example of a ServoCAT controller used in conjunction with the
Argo Navis are many of the scopes coming out of Obsession in Wisconsin
where both are optional accessories.
See http://www.obsessiontelescopes.com/accessories/index.html

SDM Telescopes in Victoria also routinely install Argo Navis and ServoCAT
systems on their range of premium Dobs.
See http://www.sdmtelescopes.com.au/

The ServoCAT tends to be optimized for scopes in the 14" to 48"+ class and
its price/performance point tends to reflect this and I note you currently
have a 12" Bintel Dob.

Argo Navis itself is deployed on all ranges of apertures and mounts and
there are many users with 12" Bintel Dobs. The unit is completely standalone,
doesn't need a PC or laptop operate, will run off a set of AA batteries so it
is easy to throw in the car, drive to a dark sky site and begin observing.

We provide full assistance to many ATM'ers and see units deployed
on all sorts of mounts, from the conventional to the exotic.
For example see http://www.wildcard-innovations.com.au/images/BINO460_argo.jpg

Argo Navis is designed and manufactured in Sydney and has been a U.S.
Sky & Telescope Hot Product.

Hope this answers your question.

Best Regards

Gary Kopff
Managing Director
Wildcard Innovations Pty. Ltd.
20 Kilmory Place, Mount Kuring-Gai
NSW. 2080. Australia
Phone +61-2-9457-9049
Phone +61-2-9457-9593
sales@wildcard-innovations.com.au
http://www.wildcard-innovations.com.au

[bojan]

Brett,
Argo Navis is much too expensive for what it does.
If you have any DIY skills, you should seriously consider Mel's free software, as it will give you the same functionality but for much less money.
Also, as time goes by, you can add necessary mechanical parts + stepper drivers (they are not specialized in any way, it is a design using discrete components) to have full-blown Goto system (including field rotation, focus control etc), something not possible with AN only (which is a good system, but it has its limitations, one of them being a price range, as mentioned before).

And, I am sure Mel Bartels solution will give you much more satisfaction - after all, this is DIY area of the forum.
Being a freeware, you can even write your own routines if you want, as the source code is made available under the public license, and contributions are actually encouraged.

EDIT:
Another alternative to AN is David Ek's box: a small interface between optical encoders and computer (via RS232) which can be interfaced with your favorite planetarium software via ASCOM platform.
Again, free if you know how to use soldering iron.

Link to David's website is below:
http://digicircles.eksfiles.net/software.php

[Spanrz]

Gary,
That's the info I was after (and the detail of it). I just found it hard to get a definition of what the Argo was. Thank you.

Bojan, yeah pretty handy with the soldering iron. :) I am a Diesel Fitter by trade, which is now becoming more of an Auto Elec job. So, well into electronics.
The interesting this is, I know of a system with encoders (PWM sensor) coupled with a driver box and a small ECM, but it's used in a hydraulics circuit to slow the hydraulics down with precise movements. It's quite a Japanese system, built and made in Japan, had to read Japanese manuals to figure out how it works. Problem is, I can't just figure an easy way to implement it as it will need to be modified quite a fair bit. Frustrating, that the infrastructure is there, it's just not easy to install. So that idea is out.

I even got thinking about using wiper motor units, or planetary drives for the mechanical motor parts.
I just have to think what is best for what I want to do.
Lots to think about, but my curiosity is a lot better now.:thumbsup: Thanks.

[bojan]

Brett,
The encoders you are mentioning are not suitable for telescopes.
Telescopes are using incremental encoders (or rarely, absolute - but they are VERY exepsive compared to incremental ones), which is not PWM sensor.
Incremental encoder outputs a certain number of pulses per full revolution of shaft. and it has two channels (signal is delayed by 90°), so the processor can determine the direction of motion, and can account for position/direction of where telescope is pointed at.

To (quickly) track satellites and that sort of stuff with telescope, most likely steppers are not the right answer (they will be too slow), you need a fast servo mechanism, and you will not find this sort of thing used for astronomical telescopes very often, as we are here more concerned with positional accuracy and smooth tracking without periodic errors and much less concerned with speed.

Perhaps you should look into military surplus for something to implement here for your project?
In the past I was working with Marconi artillery radars (from WWII), they were using powerful servos to drive radar antennas very quickly, as they needed to track aeroplanes and calculate the parameters for aiming guns.
Maybe something similar can be found somewhere on some military crap yard..
Also, so called RAWIN receivers (used for tracking balloons with meteorological sondes, again in US artillery units), they also had suitable servos.
Anyway, good luck with your project.. :-)

[netwolf]

Well this brings back memories of my journey from dob to goto telescopes.

Most of the usual options have been covered of here but there is one that has been left out. Many people have retrofitted a Meade Autostar system to there Dob and various other mounts. This is commonly refered to as a "roboscope". There is a yahoo group called roboscope where people discuss such conversions.
http://tech.groups.yahoo.com/group/roboscope/

This is a good site with various conversions done.
http://bedair.org/ScopeStuff.html

I am actually using the Siderealtech Controller with Losmandy Servo motors on my G-11 mount. Dan Grays system is very robust and Dan actively participates with users why the yahoo group.
http://tech.groups.yahoo.com/group/sitechservo/

The Mel Bartel system can be built if you have the skill from ground up relatively cheaply. Mel also has yahoo group through which he interacts with the user base and is quiet responsive and very helpfull.
http://tech.groups.yahoo.com/group/scope-drive/

Another good resources for kits for either Mels Stepper system or Dan's Servo system is
http://gototelescopes.com

Gary's AN system while expensive comes with awesome support. And Garry is always available on the phone and willing to talk and help you with anything astronomy related. I have the AN and will be using it with my Siderealtech system to give me a portable system without the need for laptop.

Plenty of options out there. And also plenty of great people to meet along the journey, that perhaps is the best of all things.

Regards
Fahim

[Barrykgerdes]

Not something like a 262 by any chance.

Barry

[Spanrz]

Update.
Well I have put myself to shame (to a degree). I spent a bit of money. :scared:
I have now a stepper control system that can be computer controlled (like a CNC machine can)

After many hours of contemplating, pulling apart a dead fax machine for it's steppers, reading IIS for stepper system ideas, countless hours watching youtube on concepts/designs, reading Mel Bartel's Website for ideas, researching and understanding what IC's do, decide what system do I need, designing and redesigning a circuit board for my needs, I thought it was all too much for what is supposed to be a simple system.

So, after many ummm's and arghs, took a venture down to Ocean Controls in Seaford (Melb) and asked what's on offer for my needs.
Figured that a system that closely matched a CNC design is exactly what I need.

So after my efforts, I bought 2 stepper motors (biploar), 2 microstepping drivers (digital ones), 1 parallel port breakout board (to connect the drivers to).
Ok, it cost a small fortune, but after seeing what it can do, I think this best suits my needs of different speeds of tracking stars to all satellites. (original concept).

I'll post up more details soon, with pics, but the drivers blew me away with the beautiful control at slow speeds, with low vibration from the motors.
The standard driver (assuming it's analog??) does not control well (vibration and jerkiness wise) at slow speeds with somewhat of a microstep, motor gets all noisy and vibrates a lot.

OK, so now I have to come up with the goods, so happy :) :thumbsup:
Going to drive my wife nuts, but it's all about me, not her ;):lol:

http://farm4.static.flickr.com/3550/3631419697_1438b24d01_b.jpg

[netwolf]

Brett, you might want to ask Mel Bartel if his DOS based stepper software can be made to work with your hardware. I would have thought Mel's hardware would be cheaper to buy rather than microstep controllers.

[Bassnut]

Its great that your stepper motor set up produces the results you are after, and it suits low cost astro mounts in general. But very few true industrial CNCs use steppers, they are mostly servo drive, as do high performace astro mounts, just to prevent possible confusion.

[Spanrz]

Netwolf, I was stuck bewteen a rock and a hard place about using DOS/Mel's system. It was just too old, for my liking. What works is great (Mel's System), but me being tech savy and all, I just couldn't bear with older technology.
Don't get me wrong, I did venture down this path, just I didn't feel comfortable being left in the 70's and 80's technology era.
The only bonus about this type is cost savings and basic funtionality.
The major issue I had with a stepper system, was of course the jitter at low speeds. Now seeing what a digital driver does vs the older drivers, there is no comparison to smoothness.
I was floored and stoked about how it performed over the older drivers.
Unfortunately my mind overtook my wallet and parted with a few hundred dollars as soon as I seen it. I was silly, hehe.

I found this website as a bit of relief about the decison of the Digital Driver.
http://www.cnczone.com/forums/showthread.php?t=82317
(I made a posting about some wiring, but it's all good now)

Bassnut, I did contemplate a servo drive, where I work, I can get servo drives and a few types of gearboxes (reductions). I looked at buying a premade system (ServoCat, ETC) and the cost was just not to my liking for my functionality.
For some reason, the servo method just didn't grow on me.

I've prepared a small vid, to show the stepper in it's first ever movements tonight. Ok, it's fast, but it's being controlled by preprogrammed software, that I have to get used to. Hehe.

http://www.youtube.com/watch?v=7fOY-9HJJ_g

I just have to setup the parameters now, that's the fun part.:scared:

[Barrykgerdes]

Hi

I have been reading back through the posts on this topic and have read some interesting findings. However my experimenting with dob driving using friction drives and stepper motors always gave poor results due to slippage and poor speed ratios for the stepper motors. I could not find any steppers easily salvaged that would spin under computer control faster than 3000 steps/sec.

The only really satisfactory method is a servo system based on the feed back from an encoder rigidly fastened to the rotating axes of the telescope. I believe Gary's encoder system meets these requirements in the Argo Navis. Although I have not had the oportunity to test it, in the long run I believe the price would be well worth it.

I have given up playing with stepper motors as in the long term they are never going to be as good as a full servo system.

My next experiment will involve gear drives (or toothed belts) to the axes of my Lightbridge using the salvaged motors and encoders from a Meade Autostar system and use a 497 controller.

Barry

[bojan]

It all depend on do you REALLY need such fast slewing speeds.
What is the rush?
If you do something serious, it is of no importance (unless you want to check hundreds of objects over the night).
Just fiddling with the computer (inputting the coordinates etc) takes significant time and it is not very practical for occasions such star parties.. I found that star-hopping method is much better (of course, you need to know how to find a way among stars). In that case, Bartel's system tracks the object when it is already found (by star hopping or Go_To... or Push-To, via cheap encoders of course.. 7000 ticks per shaft rotation is enough).
Also, you need to consider the rotating momentum of the telescope.. no-one wants to swing it so fast that things are flying of in all directions in the night.

Stepper motors are OK for tracking and short jumps to objects that are not easily visible. They are also much more mechanically reliable (no brushes etc) and simpler to drive.
Slippage in the wheel system is not the problem, if the wheels are done right (brass or steel wheel on steel shaft for example.. and no grease of course), and if the pressure is sufficient.
Of course, wheels are just one way to do it.. I have seen Bartel's system done with worm gears, so any possibility of slippage is removed.
Also people are using timing belts. No slippage there.

"Old" computer technology is OK, it works and it is cheap. $50 for old computer (that will be used outside in the cold, dew, prone to fall in the mud in the dark from a table and so on) is justification enough, I would think.

I wish you guys all possible luck with Vista and such modern products, you will need every bit of it :-)

[netwolf]

Brett, you should take a look at the Sitech system I mentioned above, I know i do harp on about it in multiple threads but it really is developing into a bees knees system. The Ascom driver now includes Platesolving system, a Mount Modeling system, all this in the Ascom driver itself. Most other systems relly on external software for this intelegence.
http://www.siderealtechnology.com/
Its reltively inexpesnive esepcially if as you say you can get your hands on Servo motors through work. Or you get them of ebay.

Dont get me wrong the Stepper systems are very good for what they do, and yours sound like its quiet a work horse. But what are you going to use to drive it? Do you have some plan in mind for this? I am not sure if there are any generic programs that you can use, but you may have to write your own. I would like to see something like this develop here, becuase there are many dob users out there that want an affordable local solution that can be made from of the shelf parts.

There are a few Stepper systems out there in Europe that are quiet good also. Like the LittleFoot and BigFoot http://littlefoot.rajiva.de/
The MicroGigga system, this is available via Telescope and Astronomy in SA. http://www.micgiga.com/
The FS2, etc etc.

All of these perform well based on user reports.

Regards
Fahim

[Bassnut]

Bojan, I agree with you that an Astromount is not in a hurry, so no need for a powerfull, fast servo. But I have found the holding torque and acceleration of a stepper problematic on a system with high inertia. One of the reasons for useing a stepper is its often not necessary to use a tacho or encoder feed back, just count the motor pulses, so cheaper, but electromagnetic slipping in that case (even a bit,without an encoder) leads to rapid loss of position accurracy, or oscillation with an encoder. As with Barrys experience, I think steppers are a bit of a dead end these days, and not much cheaper than servos. Mount motors do so little work generally, that bushes are unlikely to wear out for many years. A servo will always give better performance.

Im not saying you are dead wrong Bojan, am interested in futher thoughts from you on this.

[bojan]

Fred,
Steppers can not slip a bit.. they can only miss a whole step. And when this starts to happen, motor stalls completely, so the user is made aware of it immediatelly. The reason is mechanical overload (very rare), or too high driving frequency (very often) - thus low torque due to coil inductance which does not allow current to ramp up sufficiently.
But all this can be dealt with by proper (and not expensive) design of the motor driver, with appropriate parameters (low enough max speed, high enough driving current/power supply voltage etc). Also, as you said, encoders on the stepper shaft can be used, so missing pulses can be accounted for and corrected as necessary. This however adds to a cost of the system (and with servos, they have to be there).
As to moving heavy telescopes, there is no reason why appropriate acceleration should not be used. And Bartel's program supports it, the practical ratio of speeds is around 1:10 and it is of course configurable. The only problem with this particular software is, slewing is done in half-step mode, so at lower speeds motors are a bit loud (every step is audible). However, this software is supplied with source so someone who knows how to do it, can change it as necessary.
As I ponted out before, I am using this system for years on my dob and it is good enough for visual on star parties and allows for finding difficult objects by short jumps from easier and brighter objects.
Usually I do not need to reset coordinates for the whole evening, which means there is no significant mechanical slippage present. Of course, if I deliberately jump to and from between two far away objects, probably there will be some noticeable slippage, but I never had the urge to test my mount to such extent.. I rather spend this time observing :-)
True, there are better systems.. Ferrrari versus Toyoyta argument was presented on this forum countless of times.
And I am sure there are people on this forum and elsewhere who prefer Ferrari (and perhaps some of them can afford it) over Toyota (or Datsun 180B) but I am not one of them. For me it is important that with my mount I can do what I want it to do, and this is more or less so.
I see no reason to spend k$ for something that may be better, but what will almost never be used (other that need to posses something exceptional and valuable). Being a technical mind and engineer by profession, I always prefer to be practical and extract maximum value from what I already have (or from what I can afford to have).

Anyway.. each one of us is different..

[richardo]

Brett,
if you are wanting a Bartels system already built and ready for you to mount,
I still have mine for sale....
See ad..
http://www.iceinspace.com.au/forum/showthread.php?t=44874

These motors have heaps of tork...
Sounds like you'll get some help from Bojan which wil be good... to get you up and running.

Can sell it without lap top for $380.00...


PM to let me know if your keen..

Rich

[Rod]

I can't add much to the technical discussion here but thought it might be useful to let you know I can run my Bartels stepper system comfortably at 3 degrees per second when slewing. In fact I can run azimuth at 4 degrees per second. The alt drive won't go that fast - I assume slight imabalances in the tube prevent that as well as a slightly finer gear resolution. I use large expoxy worm gears so they are very cheap and work well visually.

I am used to running the system at 2 degrees per second for slews and find that more than adequate. Adding flywheels recently increased the speed. Slewing is much quieter than the commercial SCT systems I have seen. Steppers are noisy when tracking and I find that annoying but the noise can be reduced substantially by using plastic mounting bolts etc.

Thought I would add this for sake of completeness when discussing / comparing the Bartels drive.

Rod.

[Spanrz]

You know, lots of good valid points. Something I have taken on board.
And gladly accept your views. :)

Rich, thanks for the offer, but I think I've committed myself now. :)

Bojan, the concept is to be diversified. Meaning, not only track stars, but also Satellites, Space Shuttles, ISS and a challenge to possibly track other airborne stuff (i.e For a bit of fun, Jet Planes).
It wasn't designed to track so fast that it's like a radar dish whizzing by. It's the variance of speeds with steady motion.

Now I looked at all possibilities, but the cost factor put me off (Netwolf, I did consider Sidereal a while ago). Not only that, the function (IMHO) of speed varience was limited (locked). Shoot me down, if I'm wrong??
It was designed to track stars, with a very low error rate.
With Argo and (we'll say) a Servocat or to the likes of a high end GOTO, the $$$$ vs Functionality just didn't gel with what I envisaged/dreamed of.
I wanted to push the envelope, than just tracking stars. Blame NASA, hehe. I have just got the want to track the ISS and the Shuttle.

What your views are on stepper systems Bojan and the issue about steppers at low speed are an issue which I took great recommendation to (your views). That was one of my concerns and I teetered to a Servo System.
I have read entireless threads on Mel's system and agree wholeheartedly on what you guys think of "If it works, leave it alone" or "stick to what's best".
I grew up with DOS, win 3.1, win95, win98, win2000, win xp and now vista. I am accustomed to changes with the software (and the sometimes lack of tech support/redundancy when all else fails), but IMHO, it's time to get a project in today's world.
I had made myself sucumb to making a system from scratch, you know, on the cheap (what some of your views are), keep it basic and all, but the effort that I have already put in wasn't giving me results on how I wanted it to be.
I just got tired of trying to find something that was "proven" but wanted a view on todays world that used today's technology, like using a USB port, not parallel or serial (etc)
Yes, as you guys have voiced good opinions about what is best, I haven't forgot that.

Now what I bought, cost $600. Gave me a power supply, parallel board, 2 drivers and 2 motors.
I have had a play around with it the last few days. Have an issue with the PC talking through to the Parallel board and making that tell the drivers what to do.
I can connect to the driver itself through the serial port/cable and program and test off that. That works fine.

Though, I've got a case of "IKEA" at the moment. Got it all together, but the instructions say confusing stuff.
The Software for all the talking is working, but not correctly. The board instructions are very very basic, I know I got the board on and functioning, but the software to drive the board isn't moving the drivers.
I've done as the instructions say, but nothing moves. It's a trial and error. I've been told to have a crack at VisBasic, if all else fails.
It's just that I don't know (I can't see it - the signals), that where the communication is failing (i.e signals at the board but no further, the board is faulty, the software is setup wrong, cable wired incorrectly etc etc.).

It will be something stupid :screwy: I haven't setup yet. Plenty of settings to get through.

But it does work and to show, I have got the speed down again, to a point where it still makes no noise or vibration.
I must make a point, that "all" stepper systems I have seen thus far, whether on youtube or what some of the IIS users have got or experienced, do make noises, vibrates and jitters quite a lot at this level of speed. And IMHO, they use older technology.
But what I have achieved (to show myself), that at slow speeds that it shows no signs of vibration (resonance) and noise. Even no jitter.
And heaps of torque.
I've had this system a lot slower than the link/vid below with outstanding results, but forgot to take a vid of it. I had it turning about 1 rev a minute, with no vibration / noise / jitter.
It has settings in the software for the driver, to adjust for that. (phasing etc etc)

http://www.youtube.com/watch?v=TYTS4-YgaXw

It's one of the reasons why I bought it. As the non digital driver models could not perform at this level. I was shown with the same model motor on both devices and the digital one blew the other one away.
Now, I'm only assuming (voice your opinions if you want), that most models of what Mel's system or most other people use on a stepper control systems, use the non digital driver one with older tech.
:)

The problem is, there is hardly anyone here or around the world, that have tried to use the Digital driver for a Stepper system for a Telsecope.
Someone (1 person in the states) has tried it and also gave absolute rave reviews on how it's handles the steppers.

Me think's, I'm a guinea pig :eyepop:

[MrB]

I'm a little unsure what you mean by digital/non-digital stepper drivers?
Could you elaborate please?

[kinetic]

I made the Datsun 180B from the ground up too :)
I cannot comment on the pros and cons of the servo system
as I have no experience of it.
Feedback on Mel's site about the Servo system is very positive.

I couldn't add anything more about the stepper system except
maybe about positional feedback and stepper stalls during slews.

If the system has been designed and refined well, as Bojan refers
to in his post, the slews can ramp up in speed, slew at a linear
rate and ramp down....all without a stall or problem due to inertial
load. On my system the slew is about 2 degrees per second.
But I hardly ever use it...it's mainly push-to locally or a correction
using the ASCOM interface in microsteps.

Also, positional feedback can be made to be quite accurate.
And quite cheaply too. 4000+ count rotary encoders can be a few
hundred dollars each US.
Or you can rip apart a cheap serial mouse and use the photo
interrupters out of it. Your imagination is your only limitation as
to the practical implementation.
Here (http://www.iceinspace.com.au/forum/showthread.php?t=42151) I recently upgraded my positional accuracy on the homemade
GEM to be 8500 counts per 360 degrees on DEC and 25,000 counts per 360
deg on the RA. All using cheap recycled bits and pieces.
FWIW

regards,

Steve

[Spanrz]

Digital :
http://www.americanmotiontech.com/products/SteppingDrives/DMSeries/DM556.htm

Analogue :
http://www.americanmotiontech.com/products/SteppingDrives/MSeries/M542.htm

Look the same, but perform way different.

The Digital is promoted "as Servo like".

The analogue talks a lot about "the driven motors can run with smaller noise, lower heating, smoother movement and have better performances at higher speed"
Which tells me, they still have a choppy jittery feel at low speeds, but better than most on the market.

Steve, loving your (Datsun 180B) project -(worm gear stuff). :thumbsup:
Engineering greatness!

My head (mind) is somewhat in the Datsun 180B, but my heart is in the Ferrari. I have the hands to design to build that 180B, but my heart won't let me...haha.
The Datsun 180B will be in the gearing and mounting. :)

[netwolf]

In essence I think the diffrence is that a DSP microprocessor is used on the controller to provide intelegent microstepping over a larger variance in speed. Things like natural resonance are automatically calculated and dampning is provided. Also there is more microstep options available than on a standard analalogue controller, in which the microsteps available are pre-defined.

I think however you will still need to program code to drive it for Alt/Az tracking. Genrally the software provided is for testing and demo use. There is probably some basic code out there that you can use for some intial simple tracking programs and build on that. In essence you are going to need something like Mel's software Scope for dos or Scope II for windows. Or even simplar to a AScom driver like what Dan has developed for his Sitech controller. I am not sure of the full capability of your controller, but you might be able to program it to do just basic slew and track without need for computer, and use the computer to provide extra control for Goto, alignment, pointing etc.

You may also consider something Like Labview to create your computer interface. Labview alows you to create realtime instruments on your computer that you can then interconnect with external controllers to provide a total control system.

[MrB]

Thanks guys, was just a little confused as all stepper drivers are essentially digital.

[Spanrz]

An update.
It's taken a bit of time to get used to all of this system.

I used a CNC software called Mach3 to drive the motors.
After many hours, attempts, trial and error and plenty of coffee, I got the parallel port to talk to the drivers. I think I had the steps/settings a bit messed up.
Still a few adjustments on the drivers, but it is now a fully functional 2 x stepper control system.
The motors have a huge amount of torque, but as some of you have said about steppers, there can be an issue of dropped steps.
I've had that issue, but because there are so many types of settings, you can accomodate for that.
Wrong step setting/amperage/phase on this system, does produce drop steps and no/reduced torque, but with a bit of paitence, it all comes together.

Vid of what it can do now.
http://www.youtube.com/watch?v=bex-VJLsgeM

Now, it's just making a G-code up, to track.
Going to see if I can control a g-code as slow as I can with no jitter/dropped steps. Less gearwork, less backlash.

Then comes the fun part of mounting and geartrain system.

[rajiva]

Hi Fahim,


yes, but the BigFoot Project was replaced by the new LittleFoot Elegance Photo (http://elegance.photo.rajiva.de/elegance/photo/indexs.html) Controller. The LFE Photo is also a self construction project but available as ready made controller (~450 Euro) over several companys as well.

CS
Rajiva

[MrB]

Hi rajiva, welcome to IIS :)

Thanks for the link. Looks like a great little unit, very well thought-out.
I particularly like the heater for the LCD panel! A very nice touch. ;)

[rajiva]

Hi Simon,


thanks.


Yes, we put a lot knowledge in this controller. The LittleFoot story is now running for many years and the community around the project is growing and growing. :)


We put the heater in as with very low temperatures the LCD gets slow. The temperature is measured by a NTC that is connected to the ADC of the MCU. The MCU is triggering the heater depending on the temperature.

We also put a lot features in the controller that we see outside as extra equipment. EPEC is a similar feature as the Meade Telescope Drive Master, the SkyAgent is something like Celestrons SkyScout but useing our huge object database, the Robofocus controller is just that :lol: and so on. I like the idea to just have one small box that includes as much as possible.

At the moment we are trying to get the stand alone autoguider running which uses the build in video controller. I do not know where we end up with this project in the future. :lol:

CS
Rajiva

[MrB]

Hehe, I discovered this problem when I visited Germany mid-winter, my mobile phone screen was un-readable most of the time!
Australians don't normaly have this problem ;)

[rajiva]

Hi Simon,


yes, LCDs do not like the cold, me neither. :lol:

CS
Rajiva

[Spanrz]

Bit of an update.
Been a slow progression here, namely costs and priorites got the better of me.



Since the last update the following has happened :

I spiked the breakout board a couple of weeks ago, with one of the loose motor wires (doh!) And it still worked, but one motor would turn only one way. I figured I spiked the IC. Just re-purchased another board, no worries now... Motor wires are taped up :)
Bought a USB to Parallel adapter, on the onus that I would try to atleast attempt a connection to the controller. Nada, tried so many ways of drivers and all, I gave up as there is virtually no software or extremely hard to find someone who makes a type of usb to parallel port connection, so you can talk to the controller. These devices are "strictly" Printer only. I did read on IIS about the issues, but had to make it clear to myself that it could or couldn't be done. (It cannot be done)
Looked at various ways of getting a parallel port connection using the new laptop. Found that you can buy an "expresscard" to use with the late laptops that can be built with parallel port connection.
This option is the "only" and best alternative, than to buy an older Laptop PC with a dedicated printer port.
Finally found a expresscard with a printer port, however this was an arduous procedure. No one stocked them. Went to about 7 stores (DSE, Jaycar, Officeworks, HN, etc etc). You could buy a Serial Port version, but not the Printer Port one.
The only way was to purchase an expresscard via Ebay. That is, a PCI-E version, not the USB one. Had to buy it from China (Taiwan) for $59 AUD, including airmail shipping.
So far, as everything was a guess that it would work, has now come to a point that everything does infact work.

I now have succesfully tested a late model laptop, fitted with an expresscard (PCI-e) parallel port connected to a cnc breakout board (controller) that supplies two drivers and motors with a function to work.

Because there is such a finite control, I still haven't figured out the stepping yet, as it works so finitely, it's so hard to see it work.

I am still awaiting the delivery of my worm gear set from HES (Smallparts in QLD), it's due in around a week or so. But just realised, I have a package at the Post Office waiting for me. It might be the gears, or it might be some AC/DC tix.

Anyways, i'm just so relieved that now I can get somewhere with this stepper system with the newish laptop.

The hardest part now, is setting up the software (configs). I had it working really well on the Desktop ages ago, but wiped all my settings due to windows being corrupt and reloading it.
So now, I have to "trial and error" things to get what I used to have.

[Spanrz]

Just got the gears in the mail. Happy with what I got.

Now comes the engineering of the brackets and the fitment of it all.
Not going to rush this part.

Glad I got the parallel port yesterday.

Here's the gear pics.

http://farm3.static.flickr.com/2576/4094983996_c8a3736692.jpg

http://farm3.static.flickr.com/2573/4094984040_895da330c7.jpg

http://farm3.static.flickr.com/2787/4094984198_97fc696221_o.jpg

[bojan]

That looks good.
Good luck with the rest :-)

[kinetic]

I agree....sweet.
Certainly looks the business!

Now get that baby well centred :)

Steve

[Spanrz]

Stepper system working off the laptop. Phew! :doh:
After months of research (and a bit of guesswork, with a lot of dreaming), the baby has come alive. :evil:

http://www.youtube.com/watch?v=nncvO3ZgguM

It took a bit of fiddling to get it to move. All of the initial settings had it moving so little, you couldn't see it move.
These movements are amplified to show the functionality of it.

Still deciding if Mach3 is "go" for using it to control the scope, as they say Mach3 doesn't really work well on laptops (due to the power saving features).

This is using an eBay bought expresscard (PCI-e) parallel port.

I know the system is portable now, but I tried the power inverter the other day and the power supply unit didn't like it.
Unsure if it was the Wattage rating, or if the modified sine wave had something to do with it. The inverter was rated (continuously) just above what the power supply was rated at.
So still have to think of another alternative for power.

[bojan]

hmmm
Looks good, but too much hardware for what it is supposed to do, IMO (for telescope control, of course).

Bartels works on DOS machine, with only 8 transistors (4 per motor) to drive 2 steppers (in micro-stepping mode), plus it supports additional steppers for focus and field rotation.
However, you mentioned earlier you want much more than that (tracking satellites etc) so all this is probably justified for you.

Your problem with PS is most likely rating. But if inverter is not sinusoidal (cheaper models output square wave) then the rectified voltage from them is too low.

[Spanrz]

Have done some thinking.
I finally bought the mount hardware (aluminium bar) for the gear drive and also bought a few little electric pieces too.
Had a play with 3DMAX to get an idea how the ALT drive should look like.
You know when I wanted to do this mount, I had writers block and couldn't find the materials. 2 days ago, it just came to me and just made it in 3DMAX in 1 day. Materials were also found today as well.
Do you know how hard it is to find thick aluminium (or flat bar)? Finally, the supplier was just around the corner from me, so now if I stuff up, I got a very local supplier. Usually they are the other side of town. I had to buy 4 metres of the stuff though.

Taken me 6 months (money and time) to get things together, but decided to take a slightly different route.
Whilst the breakout board will be used for testing, I've decided to utilise a 555 timer circuit. Just went and bought a couple of kits today from jaycar and bought a few little extras if I need to modify the circuit if needed.

Going to manually control pulses to the driver for the stepper. Which is always the way I wanted to do it, just my brain couldn't get around "how to do it" the way I wanted. Auto driving it (computer software driven) was a great learning tool. How and what to adjust and how it affected the stepper.
The computer software gave me a problem, if I need to slow down or speed up or just reverse it made it cumbersome to edit finitely instantly.
No software was out there for just what I wanted.

Doing the kit, just so I can learn about the timer circuit and stuff.

Hopefully I can come up with a product soon.

3DMAX concept : http://www.youtube.com/watch?v=yRMDifNM9kc

[Spanrz]

Had the best ever success tonight.

I built the Pulse Generator and it worked first crack. Then had the thought to test the output, via a speaker, that worked.

Ok, next step the stepper driver.

Unbelievable success.....The pulse generator powered the driver for the stepper and the stepper turned accordingly to the frequency.
It is a little bit dirty on it's operation, but the test was a little bit past expectation.
The power I was using was 12V with 300ma. Unsure if that was enough current, but it seemed to work fine.

The stepper drove fast and slow enough, but there was a slight issue with the graduations of it. It ramped up and down quite fast, it was a little hard to control the speed. The finite control was missing.

But all in all, I just didn't believe it to work. Pretty stoked. :D

Now to modify the smoothness of it.
Also not sure how to get "reverse" yet, it's seeming to show a double pulse to turn it in reverse. I won't do that for a bit, as changing speed in one direction is ok, as long as the object is moving.

[Spanrz]

I now have a working Pulse Generator, and a stepper motor system working in both Forward/Backward directions (CW/CCW).
Took a few brainstorms and a few forum postings later (electronics forum), that gave me the concept for the other direction I needed.
I already had CCW direction, but failed to get CW.
I needed a 5V (with a 2.6K resistor) constant signal on the DIR port of the driver, to get the direction change.
I thought I had to have a second pulse for that, but obviously not.

So it looks like now, 2 x pulse generators will drive 2 x drivers independant of one another.
It's more the calibration of it now. (how to setup the slowness of it all and how much speed it up by manual means via a potentiometer).

[alistairsam]

Hi All,

I've been avidly following this thread as I've been working on a stepper control system without computer control for a month, and thought I'd share my findings.

Below is a control system for a forkmount as I converted my 8" dob to a forkmount. Pics to come soon.

First of all, I'm not sure if its been mentioned before, but i found an excellent stepper driver from Allegro Systems on this aussie web site http://www.littlebirdelectronics.com.au (A4983)
it is amongst the best value for money i've seen so far and works a treat. (I had a look at ocean controls as well, but that was a bit too much for my budget)
The A4983 has built in chopping circuitry to limit current draw when idle, microstelling upto 1/16th step (3200 steps/rev for a 1.8deg motor), drives upto 3A, has a current limiting pot), extremely small footprint, digital logic inputs for EN, SLEEP, DIR, microstep, reset, etc and the list goes on.

I wanted to start motorizing my scope using discrete control and not software, so I worked on a control system using digital logic.
I came across some really good electronics and hardware suppliers (australia-rsonline.com , farnell.com.au, littlebirdelectronics.com.au) for all rare components and handy kits.

I started with a Silicon oscillator and found the LTC6900 to be perfect. It is a 5 pin SOT-23 IC, has 3 settable divisions for its master freq, uses just one external resistor to set the frequency, and is precise, extremely small (SOT-23, had to buy a breakout board from littlebird).

My Concept was simple. the Silicon oscillator generates square waves at a fixed freq (variable with a trimpot to arrive at final freq), this is fed to a freq divider CD4060, three frequencies (1/5, 1/6, 1/8th) are fed to a multiplexer (CD4051) that has three channel selection control inputs and 8 channel inputs. (i used only two bits as I needed only one of three frequencies).

I planned to use a hand held controller that has 5 buttons
CW Hi Speed (high speed slew)
CW Lo Speed (micro slew)
CCW Hi Speed
CCW Lo Speed
Tracking (fixed dir)

(only 3 frequencies are reqd as CCW requires a change only in DIR bit)

Dilemna was switch control as I had to control five different variables with one spst button each.
I therefore built a truth table to see what 5 bit nibble I needed for each of the above buttons.

the 5 bit nibble was
X,Y - Mux channel selection
EN - Stepper driver enable (when not enabled, driver disables output circuitry internally thereby conserving power)
INHIBIT - inhibit / strobe input for Mux
DIR - Simple High/Low at driver's DIR input for CW/CCW

Therefore, I settled on a 20key encoder (MM74C923) that I bought from Farnell. This has 5 bit outputs which is what I wanted, built in debounce circuitry, built in multiple key control, (if you press two buttons ).
So from the datasheet, i just chose the 5 bit outputs that I wanted as per my truth table, found the corresponding key number, wired it up to spst switches and it worked.
My criteria was that when no key is pressed, the stepper and Mux enable inputs are disabled. I achieved this by setting the correct high/low states on the EN / INH pins using the encoder.

I salvaged a 1.8deg hybrid unipolar motor from a fax machine which proved to be perfect as it came with a timing driver and belt. just wired as a Bipolar. still testing on whether to use one half winding for higher torque instead of full winding.

a general purpose PCB from jaycar, and I got the whole controller and driver working.

I use a 9V battery with a 7805 regulator for logic supply, and a 78T12 3A fixed 12v regulator for the motor supply (2 x 9V batt at this stage)

I made the fork mount entirely out of wood ( i know its not the best choice), but I'm handy with wood, so it worked well.
I got two self aligning bearings from a bearing supplier in dandenong, a 20mm steel shaft ($12) from CSC in Clayton.
I used 35mmx90mm pine sleepers (bunnings) for the fork, but had double sided cladding with 8mm ply to reinforce it, and its extremely rigid with no flex under full load.
I bought two 20mm flanges from australia-rsonline.com on either side of the fork that slides onto the shaft so that minimizes perpendicular deflection due to load.
I'm just waiting to get my timing wheel from small parts and have to figure out how to mate it to the steel shaft.

I setup jumpers for microstepping selection on the PCB and can easily switch between full, half, 1/4, 1/8th and 1/16th microstepping depending on what speed I need.

I had a few posts on Pololu's web site as I was using their driver, and they gave the idea of using a PICAXE microcontroller, and programming it with the free software or BASIC.
Thats my next step to computerize it or atleast add position sensing, gradual acceleration for slew, etc.

I have learnt a lot from this exercise that took me about a month, and although there are fully functional systems out there like the Bartels, I thought I'd share my ideas for someone who might be looking for a system like this.

To achieve accurate tracking, I need to ensure there are no missed steps, the motor is not overloaded, clock is stable, and adjust the master freq until the tracking rate is achieved.

One more bit that I wanted to share is that I built a Digital Oscilloscope from a Kit sold at littlebirdelectronics. I bought the $55 kit that comes with everything with the MCU already soldered, but all else including around 40 SMD parts need to be soldered.
after 6 hours of soldering the 40 odd SMD's, (desoldering braid was invaluable), it worked straight away, and was well worth the effort and price.
It has a built in frequency counter so I can see what my clock and divider frequencies are, and more oscilloscope features that I don't really use. they sell a no soldering required kit for 70 odd.

I'd welcome feedback on my ideas and criticism as well.

Thanks

[hikerbob]

Alistairsam is the stepper controller you are using the Easydriver? If so the price from littlebird seems fairly good. If you have not found them already have a look at Sparkfun (www.sparkfun.com (http://www.sparkfun.com)) for lot's of odd bit's and links to the homepage for the easydriver etc. I use the easydrivers drivers driven by an Arduino, I've not yet tried Pic's. The Arduino is a bit awkward for this as it does neatly do variable frequency but relies on configuring a timer based interrupt - simple enough but I've not yet worked through the consequences of using multiple interupts (eg running a second driver, hardware interupt's to read optical encoders etc). I'm waiting for a board to arrive to make up a DAC based stepper driver which should allow me to feed relatively clean sine waves into the steppers which should give very good control. I've chickened out from doing those on veroboard.

I've got a long term project in the corner at home to make a horseshoe fork for a 12" skywatcher. Some parts have gone well but it's not as smooth as I want in the horseshoe bearing and too much flex in the gearing mount's so I've put it aside while I learn some more.

Bob

[alistairsam]

Hi Bob,

the stepper controller i'm using is not the easydriver. I had a look at that one, but it can only microstep till 1/8th. It also uses a chip from Allegro, the A3967.
The A4983 is at http://www.littlebirdelectronics.com/products/A4983-Stepper-Motor-Driver-Carrier.html and is $15. Delivery is from Sydney, so is quick.

the A4983 has 1/6th stepping which in tests has easily noticeable smoothness compared to 1/8th. So I might use that for tracking and micro slew and digitally switch to 1/4th, 1/2 or full stepping for high speed slew which should give me a broad enough range.

I'm pretty sure the PICAXE's have build in DAC's, and from what I've seen, they look very easy to work with.
Voltage regulation made a significant difference with the motor's rotational smoothness, and if you have a look at the decay modes at 1/16th step in the A4983 datasheet, its quite close to a pure sine wave.
I'm not sure if sparkfun has an australian site, but I noticed they were US based.

[hikerbob]

alistairsam thanks for that info. Something else for me to look at. 1/16 stepping would be nice.

Sparkfun is US based but orders generally arrive with 5 or 6 business day's (far better than some local suppliers). Not so fast on their custom PCB service but that's the nature of the way it works.

Bob

[alistairsam]

Hi,

I've got a video showing the stepper controller and the 5 buttons that I wanted in my scope control.

its at http://www.youtube.com/watch?v=ldKXewyt7X8

motor jitter is because the clock pulses are not synced with button depresses. will be fixed later. the 5 button functions are
high speed CW slew, high speed CCW slew, micro adj CW, micro adj CCW, tracking.
does'nt incorporate encoders at this stage. something to be added later on.
I wanted an autonomous simple but accurate hand held controller, hence this design.

[alistairsam]

Hi Bob,

could you tell me how long they take for delivering custom PCB's once you submit the files and roughly how much it costs.
are there similar priced sources in aus for one-off pcb's?
did you use Eagle pcb s/w?

[alistairsam]

replaced my previous discrete digital component based stepper controller with a picaxe microcontroller to perform same functions high speed, low speed and tracking.
have included acceleration for high speed slews to 24000Hz at 1/16 microstepping, thats around 450 rpm.

the picaxe is an extremely easy to use microcontroller, uses basic to program, does not need a programmer, just uses 3 pins from the serial port and very inexpensive. recommended for anyone looking at microcontrollers. the one used here is a picaxe 28X2 running at 16Mhz clock and stepper is driven with the a4983 stepper driver.

http://www.youtube.com/watch?v=oO68oe5R-Ws

extremely easy to adapt for a motorized focusser as well.

[Barrykgerdes]

Hi All

I have re-visited this thread after about 12 months to see what everyone is doing with stepper motors.

It prompted me to get out my old Silicon Chip motor drive board from 1997 (mentioned in an earlier post) and see if it still worked. Well I plugged it into the computer and supplied power. Then ran the old programs that once worked quite well. Absolutely nothing worked! The program locked up and needed a hard reset.

I checked out the board as well as possible. I had lost all the drawings, article and circuits. And replaced every component. Still no go.

After scratching my head I booted the computer up in DOS 7.0 from a DVD and retried. Yes everything worked again as before.

I checked out my Basic derived software and did some research and discovered the BASIC commands "OUT" and "IN" to directly address the computer ports only work on DOS based systems. I think it worked on Windows 98 last time I checked but I do not have a Windows 98 installation any more.

This means that I need to do away with parallel port control and find another way to talk to the board with my XP system and Windows 7 64 bit. Further the none of the old compiled BASIC programs I have run in a 64 bit system.

Has anyone used Windows XP to drive equipment from a parallel port?

Barry

[Spanrz]

Loving the system. That's what I had sorta had in mind, but a little beyond me.



Ahhh the new tech vs old tech...... You can still get parallel port infrastructure, but it's very limited now.
There is software to use parallel ports in XP, but you would have to use something like a CNC style of system, or think "industrial".
I have used MACH3 with great success, bit fiddly to setup, it works they way I wanted. However, I haven't got it to allow for variances, so if you needed to move it (manually adjust), it is a little tricky. But the basis of using it, to see if it would drive a stepper system off the PC and it did. Very happy, just the finer adjustments are a bit of pain.

I have a new PC (i7), which does not have a parallel interface. You can buy a Expresscard with parallel port (which I have), but you then have to buy a expresscard slot interface for the PC.
The cheapie USB to parallel ports don't work, unless you have USB software to drive it. I have tried tirelessly to find out about this, not many people produce software on what we want. And not many produce software other than "printer driver".
There is one local guy who sells stuff on eBay (usb to parallel), but his interface is somewhat cumbersome and not supported all that well. His software is used for driving stepper motors, there isn't much adjustment in his code. Just that it works (bare bones)
I think you would have to be a software engineer to code up a software interface. Bit like ASCOM.

Sometimes, I surf youtube insight of inspiration, to see how others have done it. Moreso, if people have done it and how cumbersome the system works.
I know some people hate youtube, but I use it like a tool for everything and anything. It is a great learning tool.

[Bassnut]

Yes, I found picaxe insanely cheap and easy to use too, the 28X2 at 16mhz ($20 odd), despite the (free) interpreted basic is so fast, it doesnt matter.

They have a USB interface cable too. I cant believe the grief ppl go through with C and generic PICs.

[alistairsam]

Hi,

I picked up some second hand stepper motors from a sunday market at wantirna, vic for $2 each. pics attached.
the guy had 20 or so motors and has heaps more.

my question was regarding current rating and motor drivers.
i'm using a TI DRV8824 driver thats rated for 24V/2A max.

as in the pics, below are the motor ratings for the ones i got
6.0V/phase - 0.85A / phase
1.8V/phase - 3A / phase
5.1V/phase - 1A / phase
3V , 2.3ohm coil Res
3v, 3.5ohm coil res.

as my driver is rated 2A, i should be able to drive the 1st and 3rd motor in the list.
but how about the rest?
the minimum voltage for my driver is 8V. so does that mean I can't use the 1.8V/3A motor?
is the current rating calculated as I=E/R for the last two? so 1.3A and 850mA for the last two?

thanks

[bojan]

All those are low voltage motors.
However, PWM block with current sensing should work as a current source.. so as long as the driver is not overloaded (1.6A) you should be OK.

[snowyskiesau]

In most cases, you can go well over the rated or stated voltage of a stepper, it's the maximum current that you can't exceed.

As an example. I have some steppers that have a stated maximum voltage of 5.1 volts but they run on a stepper controller that provides a maximum of 50V.

If you know the inductance of the stepper as you can then calculate the maximum voltage it will handle. The formula is:

max voltage = 32 * sqrt (inductance).

For the motors I have, the inductance is 2.8mh so maximum voltage is 53V.

[bojan]

Be careful with this ...
The maximum voltage that can be applied to motor coil is dependent not only on inductance but also on PWM (or stepping) frequency.
The max coil current shouldn't be exceeded, though.

[alistairsam]

thanks, not sure if i can find the inductance.
so if i used say 18v supply on the motor rail, and set my Itrip chopping current to around 1A max, will I be able to drive the 1.8V/3A motor without overheating the driver?
i'm using a variable resistor for the chopper reference voltage so i can vary max current upto 2A (or 1.6A).

[bojan]

Yes, that is correct.
Chopper should generate the saw-tooth voltage on motor coil, dependent on coil inductance.

[snowyskiesau]

Check out this (http://geckodrive.net/images/fck_uploads/Step%20Motor%20Basics%20Guide.pdf) stepper motor guide from Gecko (http://geckodrive.net/).
The drives I use are Gecko 250's and G540.

[Bassnut]

Yes, no problem driving them at higher voltages with current limiting, tried it, works well. I like the TI DRV8824 microstepping driver your using, very interesting. Is it easy to interface to with a PIC?.

[Bassnut]

Funny you should say that, I noticed the average current is higher at very low stepping frequencies, the motor heats up quick (5v motor at 12v current limited). Was overcome by limiting lower freq and disabling driver close to stop.

[alistairsam]

hi,

yes, very easy to use. just the usual interface pins
step, enable, sleep, reset, 3 microstep mode pins.
doesnt need a separate logic supply.
only issue is soldering. if you can solder smd's then this is easy.
it also has fault and home pins that go low if there is a fault and home when at home state of the step table.


vid of a position test at 1/32 microstepping. just simple of loop of 6400 pulses each direction.

http://www.youtube.com/watch?v=53tgJ4pwn1w

[Bassnut]

Excellent, thanks.

[alistairsam]

not sure if you have heard of Trinamic's motion control chips.
they have stepper controllers and drivers.

the motion control chips are far more complex but have heaps of very cool features.
you dont send pulses, instead send hex commands from the mcu like goto this position, accelerate to this speed, change direction, etc.

it has internal algorithms that calculate a trapezoidal acceleration or deceleration profile on the fly.
also has encoder inputs for closed loop positioning.

also has anti stall features, accurate position, power saving features etc.

the chips are only $7 or $10 in the US, but from element14, its around $60 odd. very good value for money, but bit more complex.

not sure if anyone's used these.



http://www.youtube.com/watch?v=xaCzpRhMlLI&feature=mfu_in_order&list=UL

http://www.trinamic.com/tmc/render.php?sess_pid=394
http://www.trinamic.com/tmc/render.php?sess_pid=152


been thinking of getting these chips.

[hikerbob]

The trinamic IC look very interesting. It's a shame the eval board is so expensive.

If the stall detection is good it would be a big step forward for small stepper systems.

Bob

[hikerbob]

I'm working my way through trying to get some of the TMC223 IC's. The one one-line supplier I found only seem to use very expensive shipping options and I'm still waiting on a price from Trinamic. I'm swapping emails with a Chinese supplier to try and work out what it will cost. If it all works I'll see if it's Ok to pass on the details.

I recently picked up a box of 5V 750mA Nema17 steppers from Ebay so the IC's look like a great option for controlling them. Fingers crossed.

Bob

[hikerbob]

I've ended up using the expensive shipping option and ordered some TMC223's. I'd negotiated with two other suppliers but after a number of exchanged emails (including links to spec sheets etc) neither actually had the right IC in the SI package.

Now I need to get my head around the I2C instructions to talk to them.

Bob

[alistairsam]

Hi

thats cool. I'll be following this to see what you've managed to do .

the TMC222's are similar and they're available through element14

http://au.element14.com/trinamic/tmc222-si/motor-cntrl-i2c-interface-soic20/dp/1214195

did you get the eval board or just the chip?
which mcu are you planning to use to interface?

I2C is easy as its just the master/slave config and addressing that needs to be right. more crucial is sending the right commands and getting the ack back.
also, i think one of these don't do constant velocity, you've got to send a max position, reset and send the max position again and repeat if you want continuous motion.
oh and don't forget the pullup resistors on the I2C bus.

[hikerbob]

alistairsam, I've just gone for the chip's (amd I also got elsewhere a bunch of SMD to Dil boards). I'll drive them from an Arduino, intitally a mega.

As I understand it you are correct about the lack of a constant velocity instruction (it also takes the shortest path to the specified position) and I think that the 222 does not have the stall detection.

I've had some experience with the I2C working with temp humidity sensors in my dew controller but that was mostly minor tweaks to other peoples code.

Bob

[alistairsam]

yes, the 222 does not have stall detection, but if the application is for driving a telescope and it has a slip clutch, will you still need stall detection?

I was planning on getting one of the motor drivers driven by the motion controllers.

I'm eager to know how the acceleration profile works in practice as thats the toughest bit i had to deal with to get the max speed out of a stepper and avoid resonance effects.

I managed to do it with pwm out of a picaxe. works fine to drive a stepper at around 25Khz and upto 1/8 microstepping.

i'll be ordering the 246 driver which has stallguard and the 429 motion controller as it has constant velocity modes.

[hikerbob]

alistairsam I'm thinking of focusser control. If I'm understanding it right the 223 could be instructed to drive the focusser in to a stall (fully in) to establish it's position and then could position the focusser exactly at a nominated point. Perhaps temp compensation could be built in as well as focus points for camera and different eyepieces.

Bob

[alistairsam]

Hi Bob,

I don't think you need to use stall detection for this although its a good application. you could easily use a limit switch to indicate when the focuser has reached the end of the tube. the mcu can poll that switch or
you could have a simple optical limit switch as well which can raise an interrupt to the mcu.

so on power up, the stepper rotates in one direction until an interrupt is raised, when it stops and sets the step counter to zero.
and then depending on what switch you press, it could go to a preset step number that could be the position of different eyepieces or cameras.

i'll be building this soon as well.

[Bassnut]

Carefull what sort of limit switch you use. Needs to have accurate repeatibility since focusing can be measured in microns. There are special homing switches, some with diamond tips. Normal optical limit switch accurracy can be affected by ambient light level.

[bojan]

Ditto.

BTW, there is no need for this procedure (limit switch than back to predetermined position).

If stepper motor is de-energised from full step mode, it will stay there (if not moved by external force).
Your software just have to remember the last position (phase) it was in before power down. and it has to restore it before moving to a new position.

[Bassnut]

That would be a logical way. My stupid Pyxis focuser (supposedly one of the best) homes on everypower up, increadibly slow, noisy and timesome. The Optec does too, dont understand that.

[alistairsam]

Hi,

that is possible, but there is always the chance that it gets moved slightly when powered off or at transport, especially if it is friction driven like a 10:1 microfocuser.
so a proper initializing routine will ensure repeatable accuracy.
else other option is on power off, it parks at the home position and then turns off like inkjet printers.

I tested focusing using a bahtinov mask and a 10:1 microfocuser. i took images everytime i turned the microfocusing knob, and noticed that a half turn of the microfocuser was the minimum movement after which a change in focus is discernible. I didnt measure the linear travel though.

this switch comes with 0.25 to 0.5mm sensitivity and has different types of actuators like roller actuators.

http://australia.rs-online.com/web/search/searchBrowseAction.html?method=getP roduct&R=4235301&cm_vc=prev_au

else this reflective optical switch would be better as the electrical ones may have issues with conductivity over a period of time.

http://australia.rs-online.com/web/search/searchBrowseAction.html?method=getP roduct&R=2192634

[alistairsam]

other option is to keep the homing as an optional feature you can invoke on demand.
so the mcu remembers the last position, but if you notice that the position you want it to go to is not correct, you can do an init routine on demand with a switch. that'd be the best of both i guess.

[Bassnut]

Yes, and home on demand, perfect.

A home switch of 0.2 - 0.5mm accurracy isnt nearly good enough IMO, but I could be wrong.

[alistairsam]

Hi Bob,

any progress with the stall detection? or have you thought of a limit switch? either way, let us know how you go with the 223.
how exactly would temp compensation work?

[hikerbob]

alistairsam the first challenge is to get a chip soldered neatly to a board :(

I've been distracted by my play with Bluetooth so that's off to the side for a bit.

As far as I can tell stall detection will be a matter of working out the right settings (rotate a motor externally and take some measurements. Might be some trial and error.

I've looked into opto limit switches on two scopes and don't like the mechanics of fitting them. The idea's I've some u[p with so far need to have some kind of fitting on the focusser tube as well as the sensor on the body of the focusser. Both look awkward.

I've not reached the stage yet in my astrophotography yet where temp compensation is likely to make a difference but have seen mention of it in discussions about focussing. Some material at http://www.engineeringtoolbox.com/thermal-expansion-pipes-d_283.html and other places that give some formula's which should be possible to use. It may not be worth the bother but the idea is interesting.

My thinking is an temp sensor connected to controlling micro and minor tweaks to the step position based on changes to ambient temp. If you are at focus at 15 degrees and the temp drops 5 degree's then the focusser would need to move by some number of steps.

Bob

[hikerbob]

alistairsamI just did a quick play with Excel and the formula listed on that page. Assuming a path of 1.5 Meters, mild steel and a temp shift of 10 degree's C and assuming that I've done the maths right the change in length is 0.195 mm.

That's a basic check assuming that the only factor is the change in the length of the steel tube which is probably not a good assumption but it's a start.

Anybody got any comments on the impacts of thermal contraction/expansion on focus for a 12" Skywatcher primary?

Again not a practical issue for me yet but I enjoy the learning.

Bob

[alistairsam]

Hi Bob,

is that the total contraction in your OTA tube?
interesting, will be good to see a comparison of different material including aluminium pipes if it is a truss dob.
but can you visually or through imags, detect change in focus of a star if shift is by 0.2mm?

[hikerbob]

alistairsam I don't have my spreadsheet with me but the page linked to earlier includes a link to a page with the coefficient of expansion for a lot of materials. Steel was 13 x 10(-6), aluminium was 22 x 10(-6) I think. Plate glass 9 and pyrex 5.5 (I think). Diamond is the material we should be working with, from memory it was at 1 x 10(-6).

Bob

[hikerbob]

If anybody else is playing with the TMC22# IC's and an Arduino I've found what looks like a useful library at http://translate.google.com.au/translate?hl=en&sl=de&u=http://2nd.homelinux.com/blog/%3Fp%3D343&ei=PY_QTdenCqfWiALVxNWXBg&sa=X&oi=translate&ct=result&resnum=6&ved=0CF4Q7gEwBQ&prev=/search%3Fq%3DTMC222%2Barduino%26hl% 3Den%26biw%3D1015%26bih%3D367%26prm d%3Divns

The code comments, the PI doco etc are in German and I need to do some translation work to understand it better. Once I get some of taht to a usable format I'll try and post some relevant material.

I've not tested any of that code yet, my own attempts to communicate with the chip have been less than inspiring but a skim of the functions and procedures looks promising.

Bob