[cnczane]

Pacific Scientific PowerMaxII M21NXXA* Nema 23 8-lead stepper motors can be connected as:

Unipolar (100 oz-in at rated 4.0A/phase, 0.46ohm ph resistance)
Bipolar series (142 oz-in at rated 2.8A/phase, 0.92ohm ph resistance)
or Bipolar parallel (142 oz-in at rated 5.6A/phase, 0.23 ph resistance)

I'm happy that the lowest-possible current (2.8A/phase = 5.6A/motor) still gives the highest-possible torque.

I can understand how wiring the half-coils in a phase in parallel doubles the current drawn.

I can't understand why doubling the rated current doesn't increase the torque. I'm led to ask, "Why would anyone wire this thing to draw more current (unipolar or bipolar parallel) for same torque or even less?

I suspect the higher currents might have better speed-torque curves, meaning slower torque falloff at higher speeds?

[Mr.Ed]

A very good question, unfortunately still unanswered. :frown:

I will look around elsewhere and here on the forum, because i'm still confused about a couple of things. The question mentioned above was one of them. Good question! Keep it up! :idea:

Ed.

[ger21]

I can't explain why off hand, but I can tell you that you're right. Bipolar parallel will give much better high speed torque, and sometimes a little less low speed torque, than Bipolar parallel. I believe it has to do with the inductance. RUnning you're motors bipolar parallel gives you a wider operating speed range, as bipolar series wired motors have torque curves that drop very fast.

[fyffe555]

This is a good question. I understand that Unipolar and Bipolar Half Coil uses the least number of coils active so it doesn't give great low speed torque. Because of the short(er) length of wire involved there's relatively low inductance so the torque lasts at higher speeds because the motor is able to turn faster, quicker for a given power.

Bipolar Series uses two coils in series (also the full coil depending on how many wires you've got) and so has very good low speed torque, higher than unipolar or bipolar half coil. Bipolar Series, two coils in series (or the full coil) has higher inductance than unipolar or bipolar half coil. High inductance means the torque drops off rapidly as speed increases.

Bipolar Parallel uses two coils in parallel ( or the full coil - in parallel !) and so has good low speed torque like Bipolar Series. But the inductance is roughly the same as Unipolar or Bipolar Series because the effective coil length is one coil, the same as Unipolar and half that of Bipolar Series. That means the torque is also lasts through to higher speeds.

The current per phase can be confusing and you need to consider how many/much coils are active in each drive method.

Higher inductance basically means the motor is slower to energise each coil. As the step rate increases the motor is required to energise coils faster, the coils inductance increasingly works against this so the average power used over the time the coil is energised is less and so the torque is reduced. Increasing voltages works to counter inductance and increase the average power used by a coil by quickening the time taken to get to the rated amperage. This is why drivers using resistors to limit current, as opposed to choppers are slower than choppers for the same volts and amps.

As a quantitative measure my machine runs far better with bipolar parallel than the other methods, to the extent that microstepping unipolar lost steps badly. Bipolar series ran well on microstepping but only at slower speeds, lost steps at higher step rates. Bipolar parallel works great with microstepping and with about 1.8 times faster g00's.

As you used a Pacsci motor as an example I'll mention I got best response with mine by having as high a voltage as the driver would stand (48v) and not necessarily the all the rated Amps. Not providing the full amps means theoretically that you're not getting the full torque, but in real terms I found that wasn't much of an issue as having the motor respond faster and not loose steps over a wider range made the machine run better. Bipolar parallel gave a much wider 'sweet spot' operating range. Hope that makes sense. YMMV.

[Cliff_J]

Think of power and not just current. Power is like the light bulb ahead, it tells you how much work is being done like a light bulb rating tells you how much light it produces.

Wiring in series/parallel wouldn't matter if you could deliver the same power. However, if you have lower volts then parallel makes more sense and if you have extra voltage then series makes more sense. (this is assuming that you're using a PWM controller to control the current and not resistors although the example still works)

The inductance of the coils will rise with RPM. This rise reduces the current flowing and is why the torque falls off with RPM and why the RPM is limited. All electric motors do this, its just a question of what RPM and steppers are generally lower.

Like everything, counter-examples to generalities exist but its more a matter of matching your power-supply to the wiring than anything else.

Cliff

[bunalmis]

Pacific Scientific PowerMaxII M21NXXA* Nema 23 8-lead stepper motors can be connected as:

Unipolar (100 oz-in at rated 4.0A/phase, 0.46ohm ph resistance)
Bipolar series (142 oz-in at rated 2.8A/phase, 0.92ohm ph resistance)
or Bipolar parallel (142 oz-in at rated 5.6A/phase, 0.23 ph resistance)

This motor have 8 coils. One coil inductance is L.

1) Serial Bipolar operations.

One phase inductance Ls=2L and phase current = I = 2.8A

We can find the storaged energy in the inductance.

W = 0.5 * L * i^2 = 0.5 * 2*L * 2.8 * 2.8 = 7.84 * L

2) Parallel Bipolar operations.

One phase inductance Lp=L/2 and phase current = 2*I = 5.6A

We can find the storaged energy in the inductance.

W = 0.5 * L * 0.5 * i^2 = 0.5 * 0.5 * L * 5.6 * 5.6 = 7.84 * L

You can see we find same results.

[ger21]

However, if you have lower volts then parallel makes more sense and if you have extra voltage then series makes more sense.

Cliff

I would think parallel would almost always make more sense for CNC use, as it should always give you a wider usable rpm range an higher rpm. Unless you don't need to spin very fast, or you're driver can't handle the parallel current, I'd always go with parallel if I had the option.

[cbcnc]

I got this graph in an Oriental Motor newsletter. It illustrates it pretty well.

Graph - Speed vs Torque

Chris

[RotarySMP]

Another way to look at it is to think of the magnetic force developed by the coil.

In a motor designed for unipolar use ( a six wire motor) you can saturate the iron core magnetically at rated current with a single coil. This will give you max torque. Any more current gains you nothing, as te iron core can not generate a stronger magnetic field (and thus torque). Could even demagnetise things if taken to extremes.

If you were to wire that six wire motor in bipolar series (isolating the center taps), you have twice the coil windings acting on the core, so you only need 0.7x (1/root 2) the current to saturate the cores.

The doubled number of coil windings increases the inductance, so the coil and iron core will resist changes in current. (Every step is a change in current to the coils). This make for a steep torque curve, with torque reducing rapidly with RPM.

An eight wire bipolar motor is designed to saturate the core with two coils on. That is why you will get a higher max torque rating for Bipolar as opposed to unipolar with an eight wire motor.

In bipolar series it acts just like the six wire with the center taps isolated above.

In bipolar parallel, you have two parallel paths for the current and the inductance is reduced. The coil and core resists changes in current less, meaning the motor can step faster.

At the end of the day the rated torque is holding torque when the machine is stationary. Not very useful. Far more useful is torque at higher speeds to give good rapid traverses.

This is achieved by running the lowest inductance motors for your drives current, at the highest voltage your driver can survive (for a choppering driver).

[cbcnc]

How do the different modes effect holding torque? Ie; when the coil(s) are staticlly energised.

Chris

[Mr.Ed]

I searched a bit through the forum and was pointed towards this link :

http://209.41.165.153/stepper/Tutorials/Tutorials.htm

I found it very helpfull and in fact it answered all my questions. Referring to the very first posting in this thread : Using the centre tap and either one of the two other connections in a 6 wire motor will result in higher speeds but lower torques. Not using the centre tap, but only using the two other leads, results in lower speeds but at the same time higher torques. Users with 8 wire steppers have more wires (configurations) to choose from, but with similar outcome.

Ehh, right...Just read the tutorial mentioned above and you'll know what i mean

Thanks everybody.

Ed.

[RotarySMP]

The torque difference running a unipolar motor bipolar halfwinding or bipolar series is only about 3% of holding torque acording to Marriss. As soon as the thing is moving the halfwinding driven motor gives greater torque.

[Mariss Freimanis]

Easy; it really is. Only a couple of things to know.:-)

Tourqe = ampere-turns. Fancy way of saying multiply the current by the number of turns of wire it passes thru to know your torque. It's easy, a series connection has twice as many turns of wire as parallel (or unipolar for that matter). It takes half as much current in series to get the same torque as in parallel.

2 Questions then: 1) Why not keep the same current and get twice the torque? 2) Why ever would you want to run the motor in parallel?

1) If iron didn't saturate, you could. Iron is magnetic. You can think of it as being made of many little bar magnets called magnetic dipoles. These are randomly oriented when no current is passing thru a coil and the iron has no bulk magnetism. It just sits there.

When you pass some current thru a coil, some of the dipoles line up and the iron develops a magnetic field. Pass even more current thru a winding and even more dipoles line up. Sounds good so far.

Problem is, there is not an infinite number of dipoles. At some current, all will be lined up; past that point increasing the current won't increase the magnetic field (and torque) any higher. That is called magnetic saturation.

Motor iron is about 80% saturated at its rated phase current. That is why current must be halved in series. That last 20% is very elusive because it is non-linear, meaning it might take 10 times the rated current and a burning motor to get it.

2) Any winding has an electrical characteristic called inductance. Inductance increases as the square of the turns of wire; a series connection has twice the turns and therefore 4 times the inductance of a parallel connection.

Inductance limits motor power. Inductance has a property called inductive reactance and it is measured in Ohms. Unlike resistance, it increases proportionately with frequency. Double the speed (frequency) and you double the inductive reactance.

According to Ohm's law, doubling the reactance halves the current (I = V / R). Torque is proportional to current so doubling the speed cuts torque in half.

Power on the otherhand is speed times torque. Doubling the speed while cutting torque in half leaves power unchanged. The step motor has a perfectly flat power vs. RPM curve.

The only way to increase power is decrease inductance or to increase the power supply voltage. Look at Ohm's law: I = V / R.

Holding and low speed torque stays the same. The drive's job is to see to that by current limiting the winding to its rated value. What changes is the speed to which this torque stays constant. Doubling the voltage doubles this speed. Cutting inductance 4-fold also doubles the speed (and power) if the voltage is left unchanged.

I leave it as a puzzle for someone to solve why output power is proportional to the inverse of the square root of the inductance. It's not that hard.

Mariss

[bunalmis]

This motor have 8 coils. One coil inductance is L.

1) Serial Bipolar operations.

One phase inductance Ls=2L and phase current = I = 2.8A

We can find the storaged energy in the inductance.

W = 0.5 * L * i^2 = 0.5 * 2*L * 2.8 * 2.8 = 7.84 * L

2) Parallel Bipolar operations.

One phase inductance Lp=L/2 and phase current = 2*I = 5.6A

We can find the storaged energy in the inductance.

W = 0.5 * L * 0.5 * i^2 = 0.5 * 0.5 * L * 5.6 * 5.6 = 7.84 * L

You can see we find same results.

I found an error in my text.

If one coil inductance is L. Serial inductance is not 2L.
Because two inductance magnetic coupled therefore Ls=4L

Lp=L (two coils use same magnetic core)

Ws=(1/2)*4L*i*i
Wp=(1/2)*L*2i*2i

Ws=Wp

[Novec]

All this parallell/series talk made me kinda confused, but I was already confused, so who cares The thing is I'm using five of the PowerMax II motors mentioned in the first post, but it has different specs:

Parallel: 3.50A 0.53ohms 2.0mH
Series: 1.75A 2.12ohms 8.0mH

I've been told to use 35V power supplies, but that isn't necessarily correct... The main problem is the amperage, which I honestly don't know how to figure out, and I'll probably end up making this PSU myself. First of all - do these motors draw 3.5/1,75A continously, regardless of speed, even at stand still? Considering it's a mill, only two of them will be active at once, sometimes three.

I'm using bipolar drivers from Quasar Electronics. They're rated at 50V/5A, but with some extra cooling, it shouldn't be a problem to go a bit further. There's also a chance I'll get a hold of some Pacific 5410 chopper drivers, but that depends on whether they drive more than one motor per driver.

Also, there's the issue of parallel vs. series, which I'm too much of a newbie to figure out from this thread. My mill is for metal working, so it won't be moving very fast, only a couple of revolutions pr. second. If the 140 oz-in for some reason isn't enough for driving the screws directly, I have 1:4 gears to help it. Does that mean series is the way to go, needing only half the amperage?

Bear with me guys, you were all newbies once