[Rohan2008]

Hi Everyone,

I am an electronics newbie who just learned how solder circuits on a prototype board. I am currently working on a project that requires interfacing stepper motors with my microcontroller. I started looking at different stepper motors and it looks like I am getting lost. I simply don’t know where to start. I have included the details of my project. Any advice regarding the kinds of stepper motors and the direction I need to look at is highly appreciated.

My project

I am designing a solar tracker that needs two stepper motors attached to two gear systems that have a ratio of atleast of about 1:1000. I am anticipating the motors to rotate at very low speeds (probably 1 to 0.5 revolutions per second… atleast for 8hrs a day). I prefer microstepping since accuracy of the solar tracker w.r.t the sun is VERY important.

I want to execute this project in two phases:

Prototype:
The prototype model is going to be a much smaller version. I am looking for NEMA 23 stepper with a torque of about 1Nm. I am assuming that if I attach a gear system of 1:1000 ratio, my torque is going to get multiplied by approx a 1000…? (I am not a mechanical engineer either). I was thinking of Lin Engineering stepper motor (http://www.linengineering.com//site/products/5709.html ). My assumption here is that for every step (1 pulse), the motor is going to rotate by about 0.9 degrees….??? Is there any other high torque, low speed NEMA 23 motor that I should look at? What gear systems should I use for this. As for the electrical ratings, I am expecting that the motor should have about 2-3A rating (probably 30V max). What drivers should I use… I intend to buy a third party driver for my prototype.

Production Version:
I would probably need bigger and tougher stepper motors (preferably NEMA 42) as my production version would be much bigger than my prototype. I am expecting the current rating of the stepper motors to reach 10A. What motors should I look at for this purpose? As for the drivers, I plan to develop my own circuit board that is specific to my application, atleast in the production version. Which driver IC should I look at? Also, I would be happy if I can get much higher ratio… preferably 1: 360000 or so….

I am completely alien to electrical & mechanical engineering. So, any advice is appreciated. Thanks.

[MechMan]

When you get this one figured out please post the results for the rest of us.

[rajiv.tctech]

Why don't you use servo drives and motors? You can get whatever resolution you want by choosing the appropriate encoder and high torque at low speeds and no missing steps... plus you drive them like stepper motors - with step/dir. At the speeds you require, you wouldn't need more than a 400W servomotor to do the job and a garden variety 500 line encoder would give you a resolution of 0.18 degree.

[imobscure]

The easiest way to track the sun is with an equatorial mount. This is a two-axis mount where one axis is parallel to the earth's axis. (In Astro speak this is the Right Ascension axis) To track the sun it's only necessary to drive the polar axis at 1 revolution per day. A small motor can move a heavy load assuming it's well balanced. Astronomers don't usually use steppers because the motion isn't smooth enough. Often a worm gear is used to achieve the reduction required. (No the torque input and output for a worm/worm gear is not the same as its input/output ratio.) Since the earth's axis is tilted relative to the plane of it's orbit you'll need to make an adjustment to the other axis (In Astro speak this is the declination axis) as the path of the sun moves up and down in the sky as the seasons pass. A tangent arm driven by a stepper would work here.

[zeta]

Hi Rohan

Have you had a good look around Mel Bartels site and the links to home-built go-tos etc from there? Everything from small Dobs to professional observatories, motorised focusers, guidecams etc. There is a wealth of stuff relating to what you're looking for:

http://www.bbastrodesigns.com

There's a very active associated Yahoo group also.

And another approach is

http://homepages.accnorwalk.com/tddi/tech2000/

Of course this may be well-known to you, in which case please ignore.

HTH

Z

[Lexx0001]

Actually, this sounds real interesting. the numbers you mention .9** degrees would be for a half step. 200 steps per revolution is 1.8 degrees. The company you refered also sells drivers that just need step and direction for control. Selectible steps via dip switches. A couple other handshaking signals as well.

Do you need such a high gear ratio to prevent wind from blowing it out of position? Maybe a worm gear as a final drive would do. I just threw that out there, it's your design.

I suspect that the big problem is going to be aiming the device. I wonder if 5 photo diodes would do it. 1 as a reference and the other four arranged so that when it's pointed directly at the sun, everything is equal, and when the sun moves, one or two would be more in the shadow. You'd need 3 at least, although the sun travels east to west, it does go up and down. Sorry, I'm just typing out loud.

What kind of micro? Motorola makes several nice ones. Very basic software (but useable) is free. Basic, Assembler, and C. Again, I just threw that out there, you should use what you know.

I'm interested in following the progress. Good Luck.

Bill K.

[RBrandes]

Rohan,
I don't know much about steppers. I bought a system from Leo in NJ and it works great. Two small steppers move my big mirror array. Here is a link to my pages showing it and there is a link to a diary of progress. In that link you will find links to Leo's site and Justin's site.
<http://www.foundry.ray-vin.com/fusion/fusion.htm>
Regards, Ray

[pminmo]

A geared DC motor is more appropriate for solar tracking than steppers. #1 steppers consume virtually the same power staionary as they do moving. #2 DC motors are easier to drive than steppers. #3 they are more economical than steppers per HP.

Horizon to horizon at .1 anglular degress only requires movement apporximately every 30 seconds. A simple encoder can give you the position feedback of the dc motor movement if you want. Most solar trackers utilize sun's output and photo sensitive devices off axis to determine position.

[MPCRice]

Sounds to me like you are biting off more than you might be able to chew. Since you said "I am completely alien to electrical & mechanical engineering" I would suggest you use previously developed circuits, structures, gearing, drivers, etc unless you have infinite time available.

While I have never tried to make a system for following the sun, I have designed a microscope focus system using stepper motors. I too am a novice at EE, but there are many drivers and control circuits available to get the prototype running. After that is working the way you want, then would be the time to decide which of the parts you should spend your time to redesign for the production machine. I am sure your uniqueness in the market place will be based more on the task your machine will accomplish than the great circuits you designed or the few dollars saved in the process.

There are many threads on the pros and cons of steppers vs servos on CNCzone and they will guide you to the best choice. I read several of them and decided that steppers would better for my slow speed focus. The choppiness doesn't matter to me as the motor is still during observation.

If you do decide on a stepper system. I would recommend the Gecko G203V driver (it is the one I picked). They have built that driver with both input and output protection so no matter how you screw up the wiring, the driver is well protected. My daughter is part of a robotics team at her school with a number of fine engineers as mentors, but they still manage to destroy several drivers each year because of faulty wiring and other mistakes.

As to the size of the motors, you will probably surprised at the amount of mass the motors can move with the proper bearings and gear reductions. You may not need as big of motors for the full sized system. The first thing is to get the prototype working.

John

[pminmo]

Take a look at the installation guide for the AZ-125 here: http://www.wattsun.com/residential_trackers.html

It will give you some idea of the size/mechanical of a commercial tracker for a residential panel. They use a DC geared motor as well as virtually all solar trackers that you can buy for that purpose.

The larger the array the more wind load potential. It's not just the weight of the panel which can be balanced.

Another source of info: http://www.redrok.com/main.htm and http://www.heliotrack.com/Products.html

[kreutz]

A "few" years ago we used to have big satellite dishes (6-10' diameter) driven by a low voltage DC motor with encoder (for positioning). You might be able to still get one of those mechanisms to move your solar tracker, or use a similar solution if you plan to manufacture them.

[Motorcoach1]

Heres a link that deals with solor trackers lot O info http://midlandsastronomyclub.org/pro...orm_info.shtml

[Rohan2008]

Thanks so much for your time and advice. So, here are some specs and directions that I am looking at:

In general, solar trackers can be classified into to two informal groups:
Solar trackers for PV, which can tolerate tracking inaccuracies and Solar trackers for Solar focusing devices that cannot tolerate even an inaccuracy of about 0.25 degrees (elevation or the azimuth angles).

The device I am working on is a tracker for a solar focusing application. Due to its idiosyncratic specific, I much make sure that the inaccuracy is no more than about 0.01 degree. Yes, this might sound aggressive, but there are a few commercial trackers that achieve this.

Most trackers either use the mathematical modeling or photodiodes/phototransistors to accurately locate the sun. Both these approaches have their own pros and cons. See Links:
http://www.tapthesun.com/PDF/EEInc%2...0precision.pdf

http://www.tapthesun.com/technical.html

I want to use both these methods to eliminate each others’ individual drawbacks.
My tracking mechanism will be having two modules:
Mathematical module that computes the location of the sun
Phototransistors module that detect the sun location (just like: http://www.redrok.com/main.htm , http://www.heliotrack.com/Products.html )

The math module is initially going to compute the position of the sun and dive the motors accordingly until the phototransistors lock onto the sun. Once the sun is locked, the phototransistors drive the motors consistently while the math module keeps on monitoring the motor movement… correcting/directing the movement incase of clouds, light reflections, rain/wind etc.

I played with Arduino and a few phototransistors (4 with shading is ideal) and I think I got all the basics right. As far as I can see, its just the implementation that’s in my way. On one hand, I am learning machine drawing and on the other hand, I am getting my hands dirty with all these electronics stuff. As far as I can see, I don’t need any ingenious electronic circuits.

I need VERY high gear ratios (1:1000 to 1:36000) because I want to drive my solar tracker at 0.01 degrees per every 0.5 - 1 motor revolutions/sec. Further, my design requirements need VERY high torque. In fact, I don’t have an upper torque limit for one of my motors. The more torque, the better it is.

I need this sort of gear ratio because I prefer the motor to constantly track the sun rather than lock, wait until sun of off track & relocking once every 30secs as pminmo mentioned. I want my motors to constantly move all through the day. On an average, the Sun moves about 1 degree every 4 minutes and so we are looking at dividing 1 degree with 240. So, the gear ratio has to roughly match this sort of requirement.
Can you please point me to some good DC motor vendors, drivers & models that you definitely recommend? I can probably start my search from there?

There are tons of drivers, motors and circuits all around the web and so any advice will help me understand what I should look for. Thanks.

[TCarey]

Having read the other replies I would add a few comments to the other good ideas.

1. 5 1/4 disk drives are a great source for small steppers and the associated electronics. The controller board expects a direction bit and a step bit. The board does the rest for you.

2. If you use a worm/gear arrangement there is no requirement for holding torque between steps so the power requirements drop a lot. Depending on how big your mirror is, with the gear ratio you are talking about, a quite small stepper will do the job.

3. Look at Redrock for ideas and he sells a cheap solar traker that uses leds as sensors. http://www.redrok.com/main.htm

Using his sensor to give you a "It's time to step" signal driving a 555 timer circuit and the 5 1/4 disk drive electronics, you can bypass the microcontroller entirely.

How big is your mirror array going to be?

[pminmo]

Heres a link that deals with solor trackers lot O info http://midlandsastronomyclub.org/pro...orm_info.shtml

That astronomy, different issues than sun tracking.

[secreteagle]

I hope this link to the CoSaTrak tracking project helps, found when i initially started out looking at steppers, http://www.saao.ac.za/~wpk/CoSaTrak/cosatrak.html they have working mechanical & electronics examples and some software going for them. I've successfully built PMINMO's 3-AXIS single-sided board, controlling 3 steppers just fine. It looked like the simplest way to start and it was.

[secreteagle]

(My 1st post on here, lets try again)
This tracking project site might help, http://www.saao.ac.za/~wpk/CoSaTrak/cosatrak.html
has pictures of electronics, mechanics & they have software too.
I've built PMINMO's single sided 3-AXIS board successfully, was the easiest way to start & works fine for any such control project, thx pminmo

[Rohan2008]

Never mind... my message got posted anyway....

[pminmo]

Solar trackers for Solar focusing devices that cannot tolerate even an inaccuracy of about 0.25 degrees (elevation or the azimuth angles).

The device I am working on is a tracker for a solar focusing application. Due to its idiosyncratic specific, I much make sure that the inaccuracy is no more than about 0.01 degree. Yes, this might sound aggressive, but there are a few commercial trackers that achieve this.

I played with Arduino and a few phototransistors (4 with shading is ideal) and I think I got all the basics right. As far as I can see, its just the implementation that’s in my way. On one hand, I am learning machine drawing and on the other hand, I am getting my hands dirty with all these electronics stuff. As far as I can see, I don’t need any ingenious electronic circuits.

To do .01 degree, you have to more around every two to three seconds. And to do .01 degree accuracy, you won't get it with a couple of phototransistors even with shading....

Commercial trackers for concentrators top out about .05 degrees right now.

This is more of a technical challenge than "no electronics or mechanical engineering" would typically venture into. Frame load and stress, mechanical backlash, motor control methods and measurement encoding, wind loads. I started down that path..............then realized the company didn't have a clue as to what they wanted for the price, compensation and system resources.....

[rdpzycho]

good luck in your venture...

if you have to gear it down that high using series of gears, you need to consider a lot of backlash...if for example you need 1:36000, using 3 gears with ratios of 1:30 - 1:30 - 1:40, you may need at most 4 encoders rather than just one on the final output shaft if you cannot 'totally' eliminate backlash in each gear...