Re: [CAD_CAM_EDM_DRO] Re:Stepper Motor vs Servo Motor

John: I copy my article on this topic for you. It was written for a
"machinists' group", few of whom are electrically-oriented, so this is
germaine:


Stepper Motors Need Not Be Totally Mysterious! By: Jan Rowland
Keith has asked me to do a kinda follow-up to the November Home-Brew article
about stepper-motor fundamentals and controls-I took that to mean that I
should compose this with an eye toward those bits of such information which
would be 1)interesting to the HMSC membership, and 2)useful to those of us
who brew our own "CNC" gear. Whew! Being academically-challenged, that is a
task! Actually, it is that which caused me to be academically-challenged
which got me on the path of self-made CNC tinkery: I am very, very lazy!
Thus, I couldn't make it in real life via conventional means, so I have
always sought ways to ease the pain of doing real work, and making machines
do mundane, repetitive jobs seemed the way to go. It has always been
entertaining, and continuously very informative; and I am most happy to do my
best to share, although I hasten with an apology to those who already know
way more than I in the following topic: As I have said in my previous HMSC
Newsletter article, there are two routes customarily used to move those
portions in CNC gear which are traditionally moved by hand, say, by turning a
handwheel, etc: "Closed Loop" and "Open Loop". The latter method is the one
which is achieved with stepper-motors, with no feedback to the controlling
computer, which would be "closing the loop". With the Steppers, assuming all
is properly engineered, this feedback is not needed; thus, this system is
considerably simpler to design and build, and much less expensive to build,
oz-in. for oz-in. From the outside, unmounted, most steppers look exactly
like any brushless motor might. (In fact, there are Industry Standard
Case-Sizes which are common to both types). But there is a considerable
difference, and I hope to explain this sufficiently that the reader will at
least have a mental grasp of how a stepper functions: Figure 1 shows a rotor
which is here a simple bar-magnet with both "N" and "S" ends. There are four
electro-magnets about this, on 90° locations, and the polarities of the ends
of those windings are shown, assuming a particular electrical polarity to
each. You can see that the rotor is being repulsed by like
magnetic-polarities at either end, and on one side, giving the rotor no
choice but to try to move to the left, as depicted by the arrows. In Figure
2, you can see that this has happened, and the rotor has continued on due to
inertia, to an "average" position so that its poles are centered between
opposite poles of the electromagnets.


Now, if the electrical polarities of these windings are inverted, the rotor
will snap again, to the next position. Thus, by alternately inverting the
polarities of the windings, the rotor can be made to rotate at a constant
rate. But this rotor has, as you can see, only two possible steady-state
positions, NE-SW, or NW-SE. Not very fine resolution! But if the rotor is
made with many teeth like a pinion-gear might be, each with alternate
magnetic poles, and the windings are likewise made with teeth, but the total
number of these teeth is not exactly the same as the number of teeth on the
rotor, then, that rotor will find a point at rest where as many teeth as
possible face opposing polarities on the stator around it, and as few as
possible are not directly-facing opposing polarities. This way, when the
polarities of the stator-windings are electrically changed, the rotor steps
to the next-most-satisfactory position. Before you stop reading in
frustration, I will hasten to admit that, no, there are not 200 different
windings in a 200-step-per-turn stepper-motor! These windings are common to
groups of five teeth (I think!), and are arranged in a parallel-series
combination so that, electrically-speaking, there are actually only two
different windings, such as you have when windings in Fig. 1 (or 2) A and B
are in series, and form one winding, and C and D are in series, forming the
second winding. Thus, you have only two circuits for motor-current. But let's
go back a couple of decades in electrical State of the Art before we go on:
Refer to Fig. 3. Here, you see a unipolar DC supply, that is, a single
battery. The two double-throw switches inside the heavy boxes are actually
solid-state circuitry in real life, but are so drawn here for
explanatory-simplicity. You can see that there are effectively two windings
in the motor, but these are center-tapped, and the center-taps are the (-)
common, via the white wire, through the resistor "R". Thus, when a switch is
changed, the magnetic-polarity changes in that center-tapped winding. Note,
this motor has 6 leads. Also note that only one-half of either winding is
being used to carry current, switches in either position. In other words,
one-half of the copper is unused at any time. But you have only one DC
power-supply to build or buy. Now, look at Fig. 4: Here, you have two DC
supplies ("batteries" in the schematic). But you have no center-taps, and
only three wires to the motor-windings (in real-life, both ends of both
windings, a total of 4 wires, are actually brought out to the
driver-circuitry; here, the draftsman was lazy, so, to save time, drew the
concept-circuit as simple as possible for himself). By tracing the circuit
with a pencil, you can see that a similar swapping of magnetic-polarity of
either winding happens when either switch 1 or switch 2 is operated. Now, as
you can see, all the copper is conducting current at any instant. There is
less wasted-heat, less wasted copper, less wasted physical space, less wiring
from the motor, all at the expense of a somewhat more complex electronic
driver circuit, here, shown as double-throw switches for simplicity. For a
time, a little over twenty years back, I requested all the literature I could
get manufacturers of stepper-motors and their drivers to send me, so long as
I didn't have to pay the exorbitant postage. At first, I read it all, cover
to cover, and was arrogant enough early-on to believe I had taught myself
enough about stepper-technology to dare to build some home-brew CNC gear. I
did, and it worked. Didn't I say, before, that the reason it did was simply
that I was too stupid to realize an ignorant fool such as I couldn't do that,
so I went ahead and did, and it worked, anyway? It can happen! Back then,
driver-circuitry for size-34 and smaller stepper-motors was simple enough
that I could design better, myself, and did-and I built my own drivers for
some of the first little machines I built. Only once, did I have opportunity
to build something using size-42 and a MO-172 motor (big as a 1-hp. 3-Ø
tool-motor!), and the driver for those was a bit too serious for me, so I
bought them, ready-made. Then, just 15 years back, those very-high-power
stepper drivers were very expensive! Since the customer was not fiscally
disadvantaged, this worked OK. But lately, I have had occasion to build two
new CNC lathes, and I have decided to beef up some parts, including the
size-34 motors for the lead-screws. Fortunately, Superior Electric, now
"Motors and Motor-Controls Division of Warner/Dana", has begun to make a new
series "KM" of size-34 motors which have over twice the torque of the old
standard MO-92 motors. Those older MO-92 motors had tapped windings for use
with unipolar drivers, whereas the KML-92 motors have untapped windings, and
require bipolar drives. Oh, and the old unipolar drives were the "R/L type",
that is, in a compromise to keep the inductive time-constant down
sufficiently that the stepping-rate of those old unipolar designs was
acceptable, a power-resistor was wired in series with the +commons to the
motor-windings (center-taps). (R=Resistance, and L=Inductance) Due to
electrical-engineering details such as that might begin to imply, and which I
need not nor want to get into here, more energy is wasted as heat from those
series-resistors than the motors actually convert into useful mechanical
energy! But the new Bipolar Chopper drives chop the DC to the motor-coils,
and thereby modulate the duty-cycle, so that the average current is no higher
than the nameplate rating. But such a drive is a tad more electrically
complex, and it makes no more sense for me to design/build my own bipolar
chopper drivers than for a new-car buyer to build his own engine in his
home-shop, and buy that new car sans motor. So, for a couple $hundred more, I
buy myself more time to spend doing the mechanical work. And, to be
realistic, even if I thought I could build my own chopper drives efficiently,
the parts-cost, alone, would probably exceed the unit-price of the
ready-made/tested/warranted drives! To recap a bit: I feel it is worth the
extra money you might imagine you thus-spend to buy something that
works-for-sure, and for which you spend no time building anything electrical.
DC Power Supply? With either the power-wasting R/L driver or the more-modern
Bipolar Chopper driver, you still have to supply amperes of current at 24 to
40 VDC for two size-34 steppers, or even two size-23 steppers such as MO-62.
Little things, only 21/4" square flange, and maybe 3" long, but they want
some amperes! A DC supply which can supply 24 VDC at, say, 8 or 10 amps, is a
serious piece of electrical hardware! For stepper-motor drivers, either kind,
well-regulated DC is not particularly necessary. In fact, I recently noted in
the specs for a new drive I was perusing that "DC Required" was "24-40 VDC".
That's another advantage to the new chopper designs: They automatically sense
the average-current through their output-terminals to the motor, and, so long
as the incoming DC will supply that average current, the chopper-drive
automatically "chops" at the required duty-cycle so that desired average
current flows, with 24 VDC in , or 40 VDC. That is, it would do you no good
at all to supply regulated 24 VDC to the chopper drivers. For the reader not
"into" electronic details: A DC power-supply can be as simple as a
transformer connected to the AC line, followed by one, two, or four diodes,
and a filter-capacitor. The resulting output would be "DC", but the actual
voltage would depend upon the instantaneous AC-line voltage into it, the
load-current, the condition of the parts and the temperature of the
transformer, etc. But if you "followed" that DC with a regulator, you could
keep this DC-out within a fraction of a percent, so long as the "raw DC-in"
was at least 2 volts, or more, than the regulated output voltage. Such
details are trivial to a person into electrical stuff, but such a detail can
have a significant effect upon the reliability, actual function, and
cost-to-build of such a project as discussed here. Warner/Dana has been
introducing new stepper-drives and DC supplies faster than the post-office
has been delivering their junk-mail to me! For the two new CNC Lathes I am
building at the moment, I have ordered and received two new units from them
which contain power-supplies and two motor-drives, each. Oh, and they have
opto-isolated inputs, meaning that they should be about as
electrically-immune from EMI (electrical noise) as practical. Electrical
noise? Yes -- with circuitry involving computers and like electronic gear,
you can have "spurious inputs" from such sources as a large machine with a
nasty magnetic contactor coming on, across the room; an AC unit with a
poorly-wired AC-line-in, or even something as simple as a Mr. Coffee machine
switching to the "keep it warm" mode, in the same room! Surely the reader has
noticed those little "sparkles" on the TV-screen, now and then, even when the
lights don't blink, and there is no lightening-storm in visible proximity?
That is "electrical noise"! It can be harmless, or it can cause several steps
(0.0005" each!) into that nearly-finished arbor you are turning, which has an
OD tolerance measured in tenths! Some further chat on that point might be of
interest: The little CNC lathe I built at home in '84 for a very esoteric and
specific task, turning pipe-organ drawknobs, has, by now, made about 18,000
knobs of "exotics" such as ebonies and rosewoods of several kinds, and even
some Continental European Boxwood, which had been stored in an attic in
England for the last 127 years-long very illegal to harvest, transport, sell,
export, mention, or show!- In all that time, whenever I hear thunder, I
switch off and go inside and hassle the wife. Because a lightning-strike on a
power-line within miles can cause a "glitch" which can cause an unwanted
step, or stall. This latter might not "ruin" a work-piece, but then, I have
to "find" the 0,0 point again, once the lightning has gone away, and "cut
air" until the cutting-tool gets back down to where it left-off. Oh, that is
very, very rare, but it has, more than once, cause the utterance of untoward
syntax!


Jan Rowland copied here 14.6.02


[Non-text portions of this message have been removed]