View Full Version : Unipolar vs Bipolar

05-13-2003, 10:55 PM
Well, I've gotten more than enough steppers to build my table...now it's time to design and build a control circuit. (Yes, I want to build my own!)

My steppers are unipolar, 6 pin...which means that there are 2 center tapped coils, making a total of 3 connections per coil. Now, here's my question...

From reading on the web, I've learned that bipolar steppers are more efficient, and provide more torque per unit of power, than unipolar steppers. I've also learned that it's possible to run unipolar steppers in a bipolar mode, by simply not connecting their center taps.

Would this be a good idea to do this? Run them in bipolar mode? Would there be a down side? In theory it should be no problem...in practice, theory is often thrown out the window.

I'll probably end up building 2 control systems, and just try it both ways...but that's double the effort. At the very least, I'd prefer to build the superior system (at least the one expected to be superior) first.

Help? Advice? Comments? I know I won't go wrong by building a unipolar driver...but the bipolar chips also support 3 bit microstepping, while the unipolar chips only support 1 bit microstepping, at most.

-- Chuck Knight

05-14-2003, 08:32 PM
Bipolar drives cost more money and are more complex, but are generally faster. Unipolar drives are cheaper, simpler, and slower, and less efficient, because you need to waste energy on heating the current limiting resistors.

I use a unipolar board. The FET3 from stepperworld. I love it. I use a 27 volt power supply and get over 40" per minute using 100 oz steppers and 1/2-10 acme screws.

If I did it again, I would go bipolar.


05-14-2003, 10:44 PM
Well, thanks to scavenging I have both unipolar and bipolar drives. (More unipolar than bipolar, though) My biggest steppers (identical to yours) are unipolar.

HOWEVER...I understand that the unipolar drives are just bipolar drives with an additional center tap on each coil, hence the 6 connections instead of 4. I also understand that they can be driven in bipolar mode, by simply ignoring the center taps.

Since there are single chip solutions for driving both unipolar and bipolar motors, and I'm going to be building my own driver, I just want to know if I should build a unipolar or bipolar unit, first... Which would be the better approach? Would there be any down side to running the unipolar steppers in bipolar mode with a bipolar driver...that sort of thing.

The up side is greater efficiency, and 3 bit microstepping (the unipolar driver chip supports full and half steps, while the bipolar driver chip supports full, half, quarter, and eigth steps). There is usually a down side in decisions such as this...I seem to be overlooking it, somehow.

-- Chuck Knight

05-14-2003, 11:13 PM
Hi guys,
Sure wish I knew what you guys were talking about. I did how ever understand the word "scavenging". Man do I feel inadequate. Sorry, I'll leave now, tail between legs and all. :p

unipolar or bipolar unit...... say what?

05-15-2003, 12:21 AM
OK, there are many parts to a CNC machine. You have the computer and some software that runs on it, which is called the controller. You also have the stepper motors (or servos...a different kind of motor) which make everything move.

There's also a black box between the two, that actually turns the computer signals into the electrical signals needed to make the motors move. That's the part I'm talking about.

Needless to say, since steppers are used in everything from floppy drives to telescope positioning systems, to xerox machines, there are a lot of chips that are made to control them, and consequently make them move. (You can't just hook it to a battery like a regular motor)

Now...there are 2 kinds of stepper motors. One kind has 4 wires coming out of it, and is called bipolar. The other has more wires (usually 6, but can be 8 or more) and is called unipolar. The difference in these motors, is how many different ways there are to apply power to the magnetic coils inside of them, that make the motor turn.

Think about it this way. You have a coil of wire, with 2 ends. You hook up the + side of the battery to one side, and the - to the other side. This is how it hooks up inside a bipolar motor. If you solder a third lead on, right at the halfway point on the wire, you'd essentially have 2 coils...this is the way a unipolar motor's coils look. The wire in the middle is called the center tap. you hook up the + side of the battery to it, and the - side to one of the other ends.

Now...unipolar motors are easier to control, and are commonly used in consumer electronics -- I got my motors out of some old (and I mean OLD) LASER printers from Hewlett-Packard. The huge old LaserJets.

Bipolar motors offer certain advantages, and are actually simpler in terms of construction. And, by wiring up unipolar motors in a certain way, they can be used as bipolar motors. The down side is that, traditionally, it takes more control circuitry to make them work.

I've managed to find single chips that will drive either kind of stepper motor. And, by applying voltage to more than one winding, the chips can actually make them take a half a step at a time (or as little as 1/8 step...that's the 3 bits...2 raised to the 3rd power = 8) instead of a whole step. Which means that my 200 step per revolution motor could be made, functionally, into a 1600 step per revolution motor.

Why would I want to do this? Finer detail on my work. My intention is to make clock parts -- certain gears in a clock need to be very precisely made, or it'll stop ticking.

Does this help explain things, at least a little bit?

-- Chuck Knight

05-15-2003, 12:46 AM
I have a XYLOTEX (www.xylotex.com) bipoler board. I Also have 3 motors like the biggest one you have. I have mine wired using a center tap wire.
1. find the 2 center tap wires
2. match the other 2 wires up by there coil.
3. take a center tap wire and a regular wire from the same coil.
4. do the same to the other coil.

That is how I have mine wires.

Bipolar half winding.

Bipolar half-winding can give better torque performance at higher speeds

05-15-2003, 08:55 AM
For a given 5- 6- or 8-wire (unipolar) motor used with a bipolar drive you can get higher step rate by using only 1/2 of each winding. But you will have lower torque at low to medium step rates which is where the cutting is done.

So... you have to trade higher inches per minute (ipm) and lighter cuts at the tool against lower ipm and heavier cuts at the tool.

In most jobs a LOT of time is spent moving the tool from place to place on the workpiece, so you are better off with the higher speed / lighter cut approach. In other words bipolar is better - most of the time :) .

robotic regards,


05-15-2003, 09:20 AM
I suppose that if I was to do heavyer cuts and wanted that torque on the lower end with slower speed, I would just wire my motors to the board so they are Bipolar series. That would give me full torque at lower step rate.
Again I think it is what you plan to do with it.
Thats what I like about this board. Very versitile!

05-15-2003, 12:34 PM
Thanks Chuck and others,
For the quick lesson on this interesting subject.
That clears things up somewhat.

06-19-2004, 12:52 AM
from the discission I deduce that this is a unipolar motor.
How do I wire it for bipolar. Since the bibolar schematic I have does not match the wire colours.

06-19-2004, 01:06 AM
schematic that I have.

06-19-2004, 01:09 AM

06-23-2004, 12:18 PM
Sixpence: You will know what wires are center taps with a ohmmeter. The coils 'end to end' will have double resistance that the coils 'end to center tap', and not same coil ends will show infinite resistance. Just make a table with the colors in each axis on your workbook and try all the combinations. The table itself will be very explanatory when finised.

Ynneb ed alters: Just remember that if you use a unipolar wired as bipolar you need to feed the coils with half the current of the faceplate. It is not obvious but true, and will help a lot in order to make a cheaper driver circuit.


06-23-2004, 02:24 PM
I may be wrong, but I believe that when you drive both outside ends of a unipolar motor with a bipolar driver, you only need to drop the current to 0.707 times the unipolar current. You want to keep the power dissipation(ie heat) the same, and power is equal to resistance times the square of the current. Therefore if you double the resistance by driving both halfs of the coil in series, you only need to drop the current by the square root of 2 to keep the power dissipation the same. That is why you can get more torque by driving a unipolar motor in bipolar mode. Someone please correct me if I am wrong, so I don't burn out my motors.

06-24-2004, 02:00 PM
yes, could somebody please help with answerin the above.
I too suffer form burnout fear

10-29-2004, 10:37 AM
Ok, so now this begs to be asked .... can I effectively then drive bipolar steppers with a unipolar driver board ? I imagine yes you'd lose some of the torque that a bipolar driver would give you.... so, would it then use the current rating per phase listed on the motor, or draw more from the unipolar driver because it's not switching the poles at the same frequency .....?
Is this possible with the FET3 which is what I've got.... and I'm becoming more and more frustrated with because it doesn't seem to drive the unipolars I've got now......
Such a tease when the whole machine is done, but you can't make it work happily.......



10-29-2004, 11:04 AM
You can use 6 or 8 wire motors with your FET3, but not 4 wire motors. If you decide to dump that and get say, a Xylotex, you could then use 4,6,or 8 wire motors. And you'd also probably double your speed. :)

10-29-2004, 12:06 PM
Chuch mentioned in a previous post "...Why would I want to do this? Finer detail on my work. My intention is to make clock parts -- certain gears in a clock need to be very precisely made, or it'll stop ticking....."

My interest in clockmaking is also the motivation behind my interest in getting a CNC machine running.

I ended up putting the CNC router on the backburner while I made the parts for a wooden clock using a scrollsaw and other manual tools. I found that sufficient precision can be achieved by manual means, it's just so much work and takes an enormous amout of time.

Good luck with your CNC machine and clocks.

Regards, Tim.