How to Control 4-Phase 5-Wire Stepper Motor With ULN2003 and Arduino for EOG Project

Hi,

I am a beginner n00bzorz with circuits, and for this project I'm working on, it was recommended that I use a stepper motor for its precise rotation. So I bought a stepper motor, but I have absolutely no idea how to use it. It's a 4 phase 5 wire lead stepper motor, which looks a lot like this:

http://www.geeetech.com/wiki/index.php/Steppe...e_5-Wire_%26_ULN2003_Driver_Board_for_Arduino

I don't know how to hook it up, and I don't know how to operate it at all. Our project was originally complex, but we have dumbed it down considerably. We are using an electrical signal from the eye (EOG http://en.wikipedia.org/wiki/Electrooculography) to control a prosthetic eye to rotate synchronously. We are only working with left and right movements and it's pretty much three positions for the eye: left, right, and middle. So what I need to do is use the right movement output to rotate the stepper motor let's say 70 degrees and the left 70 degrees the other way. But again, I have no idea what I am doing. If anyone can help, I would greatly appreciate it.

- kavinuma

Or, how about servo motors -- much easier to use -- positioned with pulse width. All Electronics has them cheap:

* http://www.allelectronics.com/make-a-store/category/557/Servos-and-Accessories/1.html

But, if servos are either to jittery or can't drive whatever it is you want it for, then stepper might be the way to go -- but, a downside on stepper motors is a produce a "jumpy" movement -- it's not smooth. A servo would be much smoother (but as your adviser stated, steppers give precise positioning. So:
* Servo: easier to use, smoother turning, but perhaps not as precise positioning.
* Stepper, harder to use (at least until you figure out how to do it--but still requires more support circuitry), bumpy movement, but precise positioning (at least within the angular resolution of the stepper [which can be improved with gears and/or belts, etc, but that add another level of complexity]).

It's been a while since I messed with stepper motors. There are different kinds. BUt, I did it with Google -- there is a wealth of info, just Google "stepper motors" or "stepper motor drive circuits" etc. Then with ohm meter in hand I worked out the coil arrangement, and then gingerly tested the thing by applying voltage through a resistor, holding the shaft so I could feel any activity (because it might not move with the resistor in series, but the resistor is to keep from blowing out the winding while you figure out how much current they can take and what the actual coil configuration is, and what the proper polarities are, etc. Gradually reduce the resistance until you get actual shaft movement. Or, even better, if you have a power supply with a current limit feature, then set a voltage, dial the current way down, hook the power leads across a coil, then see if you can rotate the shaft [of course try rotating the shaft before applying power, to see it can be moved then]. If there is a difference in the ability to rotate the shaft then you getting in the ballpark. Then, let it run for a short while and monitor the temperature of the motor. If it gets hot, the current is too high, etc... This is called "reverse engineering".

Of course, if you have the specs for the motor, then you are all set. Just do a bit of Googling, then try various arrangements until you can pulse the shaft into rotation. In other words, you gotta just dive in!

If you are worried about burning out the motor(s) you have, then buy some cheep ones to play with (just to get up to speed on how they work -- i.e cheap smoke . Again All Electronics, or Electronic Goldmine, or MPJA or Alltronics, or Hosfelt Electronics or Electronics Surplus or Tayda Electronics.

Basically, stepper motors have different coils. Energizing a coil moves the stepper motor to a specific rotation. In your case I think you have 4 coils there are 4 positions, 90 degrees apart. You could use a stepper motor driver chip or drive it directly with four high current outputs. You rotate it by energizing the coils in order. One rotation requires that each of the four coils is driven in the proper order. You can stop it between steps by pulsing two adjacent coils, best done with a micro-stepping controller IC.

A stepper motor IC can sequence the coils in the proper order. If you drive the motor directly you need four high current outputs and have to provide the sequence with software. The sequence rate needs to be slow enough for the motor to move to each step before the next step is output. If the steps are too fast the motor will vibrate.

Most applications require more than one revolution of the motor.

For a simple application I'd go with a stepper motor. When you quite sequencing the coils it stops.

A servo motor has to be controlled continuously at a pretty fast rate so you can ramp it's speed down to the setpoint. If you try to just stop it it will overshoot. A software servo loop requires a position encoder and servo control software or hardware. I've written servo control routines in software for decades and don't recommend it for a novice.

The stepper motor position is determined by the number of steps (from a reference position - home) and the speed is determined by the step rate, so long as the stepper can keep up. If it can't keep up in a multiple revolution application (most applications) you will loose track of where it is.

On power up the stepper motor or servo motor must be moved to a limit switch in at one end of it's travel (home) unless you have an expensive (possibly multi-turn) absolute encoder. If the home switch is in the middle of it's travel then there's no way to know which way to move the motor to reach home when it's first turned on.

If your motion is less than a revolution a single turn potentiometer and an analog input can give the absolute position. If it's more than a single revolution then a more expensive multi-turn potentiometer can be used. Either way you need an analog input and if it's a servo motor you need servo control software or hardware.

A potentiometer and analog input won't provide enough resolution for good servo control unless the travel is very small.

If not you need a digital position encoder. This can be built into the motor but often requires a separate quadrature encoder or absolute encoder.

A quadrature encoder commonly two tracks with a reference pulse. A single track encoder can only read speed, not direction. quadrature encoders usually have a reference output that determines the position within a revolution. If the travel is more than one revolution they need a separate reference switch. In any case the position has to be read continuously (and usually quickly) to close the servo loop.

A simple absolute encoder can locate the position within one revolution. It has parallel output (lots of inputs) or a serial output which must be read and decoded.

If the motion requires more than one revolution then generally an absolute encoder is actually an assembly with two encoders, one of which is geared to make less than a revolution over the positioners full travel - even more expensive.

A servo motor is driven with a high power analog output, or high speed pulse width modulation (PWM) output. The PWM frequency needs to be high enough that the motor is smooth - much faster than a stepper motor rate. Generally this requires a fancy control algorithm to ramp the output up and down to control the acceleration and deceleration. You may be able to just control the output voltage just based on the distance from the target. You may to also add in an integrating output, to move onto the set if it stops with a small drive before it gets to the set. You may even have to add a differential output based on the motor speed so you can slow the motor so it won't overshoot the setpoint. This is particularly important if the load has lots of inertia or the update rate is slow. This is referred to as a PID loop or PID control and is not that easy to implement or adjust.

Reading back through you're application you need less than a revolution of travel and have only three positions . - I got focused on stepper v.s. servo and forgot the details of your application - sorry. Your stepper motor motion would need to be reduced so the 90 degree steps are 70 degrees, probably with a simple linkage.

bq). If the steps are too fast the motor will vibrate.

But, even at any rate, there will be "bumps" (another form of _vibration_)

bq). When you quite sequencing the coils it stops.

Unless there is a great enough rotational force on the shaft. In that case, the shaft will turn, when power is removed.

bq). If you drive the motor directly you need four high current outputs and have to provide the sequence with software.

Microcontrollers (such as the PIC by Microchip, or an Arduino, etc.) are excellent for doing this.

bq). Your stepper motor motion would need to be reduced so the 90 degree steps are 70 degrees, probably with a simple linkage.

Or you could just step it within the 70° range, with a limit switch at 0° or 70° or even at any point within that range. For instance, if the limit switch is at 19.8°, and each step is 1.8°, that's 11 steps to get to 0°, 27 steps in the other direction [from 19.8°] to get to 68.4° and 28 steps to get to 70.2° -- that's 38 steps to sweep 68.4° and 39 steps to sweep the whole 70.2°)

Please be aware that some of the previous answers are conflating RC type servos with industrial type servos. They are quite different animals. It's pretty clear from your statement that a stepper would be an overkill for your project, and an industrial servo would be way overkill. Both would require an excessive amount of control electronics, software, and power.

An RC servo (the kind used in Remote Control model airplanes) is very easy to control and can be driven directly from a microcontroller (like an Arduino) with a PWM output. The position of the servo is proportional to the width of the pulses sent to it. Most microcontrollers will have a library of routines or examples describing how to control an RC servo. Try Sparkfun.com