Stepper motor questions

fuentesmartinez wrote:

Are you supplying two phases? For most steppers, the simplest usable
driving waveforms look like

¯¯__¯¯__¯¯__¯¯ and

¯__¯¯__¯¯__¯¯_ . Invert one of the waveforms to reverse direction.

Both of those waveforms are bipolar; they switch between +V and -V. Some
motors have center-tapped coils, often with a common center tap for both
coils. With the center tap(s) grounded, current is applied to one half
of a coil. Switching means applying current to the other half of the
same winding, with the other side always open. 48 steps per turn usually
means 24 steps of each phase. It is usually true that 4 unipolar phases
are simpler to generate than two bipolar phases, so the motors are wound
accordingly. There is a simple circuit that uses two flipflops. which
accepts a single clock and delivers four outputs to drive 4 transistors.
I'll show it if you need it.


That depends on the driver. With simple drivers it is true. The drop-out
speed depends on the nature of the load. By ramping up the speed instead
of trying to start at the highest speed, you can go fairly fast even
with a simple driver.


The motor is rated at 24 V and draws some current there. The winding has
inductance and resistance, so the current builds up according to the L/R
time constant whenever the current is switched. If you increase the
voltage and add external resistance to limit the current to its 24V
amount, the motor won't overheat, but the time constant will be
shortened. For moderately high performance drives of this type, the
external resistor is often double the motor's resistance, so one would
use triple the voltage. I would try double the voltage with a resistor
equal to the motor to start with.


I suspect you're right.

Jerry
--
Engineering is the art of making what you want from things you can get.
¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯

Jerry Avins wrote:

   ...


You are driving an inductive load, so of course you provide damper
diodes to absorb the kickback. There are ways to connect those diodes
that, with a few extra parts, tend to damp the resonance. The motor is
much less likely to shake out of lock as the speed ramps through the
resonance region.

Jerry
--
Engineering is the art of making what you want from things you can get.
¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯

fuentesmartinez wrote:

You can build a chopper drive. Basically, you use a larger voltage and
"chop" it with a 555 and a MOSFET or power bipolar transistor. This is
then used as the power source for the driver you have. The effect is an
increase in current, which increases speed.

Also, as Jerry said, ramping the speed up and down is very effective in
reaching higher speeds. What are you using to control the driver?

-Matt

Thanks for the help, I got more questions,



driving waveforms look like

Some motors have center-tapped coils, often with a common center tap
for both coils. With the center tap(s) grounded, current is applied
to one half of a coil. Switching means applying current to the other
half of the same winding, with the other side always open. 48 steps
per turn usually

phases are simpler to generate than two bipolar phases, so the motors
are wound accordingly. There is a simple circuit that uses two
flipflops. which accepts a single clock and delivers four outputs to
drive 4 transistors.

After some braining I think I understood let see if I got it right. My
motor have 6 wires I found 2 of them are common for 2 coils (2 commons
4 coils), so what I have is centered tap coils and I can either make 2
square waves delayed 90° from each other or 4 unipolar phases. OK,
what I did was to write a program in a GAL16V8 and the outputs
(Q1,Q2,Q3,Q4) polarize transistors (T1, T2, T3, T4).  The program
makes Q1=V+,Q2=V+,Q3=0,Q4=0, then 0V+V+0, 00V+V+,V+00V+, and so on...
(that is because I read that would give more torque).
Then I’m making 4 unipolar phases, right? Are the 4 unipolar phases
less efficient than 2 phases going form V+ to V-?


speed, you can go fairly fast even with a simple driver
Do I have to come up with some kind of acceleration routine? Can I do
it by hand? If I do it by hand would I reach a significative lower
speed? I ask this questions because as soon I saw it was stuck I slow
down and try to speed up, it didn’t make a difference...


resistor is often double the motor's resistance, so one would use
triple the voltage. I would try double the voltage with a resistor
equal to the motor to start with.
I will try it!!


ramps through the resonance region.
Noted.


and "chop" it with a 555 and a MOSFET or power bipolar transistor.
This is then used as the power source for the driver you have. The
effect is an increase in current, which increases speed.
What I understood is apply more juice during less time to each coil,
is that correct? If so, do I still need the resistant that Jerry was
talking about? I think I do.


I made a clock with a 555 and it gives the transitions to the gal so
it switches the outputs. I have some switches to turn it on-off and
direction... very simple.

By the way, I found a similar motor and the label says
Stepper motor ID27
4304 171 90101
made  by Philips
I have guesstimated it will be pulling at max 5 kg... and I'm thinking
whether it will be capable... have you worked with this motor?

Thanks a lot

Juan

fuentesmartinez wrote:

There is something odd about the way  that the high bar (¯) reproduced,
so I'll use 1 1nd 0 instead of ¯ and _.

To get the same flux (hence torque) using 4 unipolar phases as with 2
bipolar ones, you use twice the current through half the number of
turns. If the bipolar power is I^2*R, the 4-phase power is (2I)^2*R/2;
twice as much. So no, it's not as efficient, but the drive is simpler
and the motor is usually specified that way.

As to your sequence, I'm not completely sure I follow (especially the
"and so on" part, so I'll write it down. Before I do, the wiring is
worth discussing. Assuming you use a 24V supply for the 24V motor,
connect the center taps to +24, and each coil end through a power
transistor to ground. Remember to connect a diode from each transistor's
collector to +24V, cathode to + and anode to the collector. The diode
will be back biased when the transistor is on, and provide a path for
the coil current (it's an inductor, right?) when the transistor turns
off. For faster switching, use a higher supply voltage, with a separate
resistor in each center tap to limit the current. Both the current
build-up and decay will be faster.

I label the phases in sets. A and A' are one coil and B and B' are the
other. For the regime called "full stepping" (half stepping is more
complicated, has less holding torque, but is smother) this is the
sequence table:

Phase   A A'  B B'
   1     1 0   1 0
   2     1 0   0 1
   3     0 1   0 1
   4     0 1   1 0
       repeat

For completeness, this is half stepping:

Phase   A A'  B B'
   1     1 0   1 0
   2     1 0   0 0
   3     1 0   0 1
   4     0 0   0 1
   5     0 1   0 1
   6     0 1   0 0
   7     0 1   1 0
   8     0 0   1 0
       repeat

Half stepping at double the single-step clock speed may avoid problems
with resonance. Stopping on a phase that is also in the full-step
sequence maintains holding torque.

   ...

Jerry
--
Engineering is the art of making what you want from things you can get.
¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯

fuentesmartinez wrote:


In regards to the chopper drive, the point that if you want higher
speeds, you need to get the coils to charge faster. The increased
voltage gets this to happen. The power must then be switched off in
order to avoid applying too much current and blowing out the motor.

The resistors Jerry mentioned are to get the "impedance" of the
electronic part of the motor to act more like a resistor and less like a
coil. In this way, the coils can be switched faster. This is a much
simpler way to get similar results. The only downside is the increased
power consumption. If this is not an issue, go with this. If you do use
this, just make sure the resistors you use are rated for the power you
are going to draw.

Hope this helps.

Matt

Matthew Douglas Rogge wrote:

   ...


I'm sure it will help eventually, but I feel that it's premature now.
First, lets get the motor running smoothly with a simple drive.

A straight chopper drive -- essentially a class-D current source --
presents a high impedance to the load. When the motor is driven from a
low impedance source, resonance effects are greatly reduced. In
troubling cases, I have used a negative-resistance source to offset most
of the motor's internal resistance; then resonance goes away. (Thump the
cone of a loudspeaker while you listen; you can hear the motional
resonance. Thump it again with the terminals shorted, you hear only the
paper.) With sophisticated drivers like these, you can expect to achieve
speeds at least five times the manufacturer's spec and drive inertial
loads considered impossible. "Stutter stepping" is another technique for
dealing with mechanical inertia. It's nice to know all this, but foolish
to use it just because it's there.

About the resistors: the windings have resistance and inductance. When
connected to a voltage source, the current rises asymptotically ti its
final value with an L/R time constant. An external series resistor Rext
changes that to L/(R+Rext). The larger denominator means a smaller time
constant and a faster rise. You pay for this two ways: the supply
voltage has to be increased to compensate for the voltage drop in Rext,
and the larger resistance undamps the resonance even further.

Jerry
--
Engineering is the art of making what you want from things you can get.
¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯

It does RUN!  I still have some doubts

(2I)^2*R/2;

I’m sorry, I meant "and it repeats".

I used to think I knew everything to be known about stepper motor and
I know now that I don’t so yep let’s review the wiring thing...

taps to +24, and each coil end through a power transistor to ground.
Remember to connect a diode from each transistor's collector to +24V,
cathode to + and anode to the collector. The diode will be back biased
when the transistor is on, and provide a path for the coil current
(it's an inductor, right?) when the transistor turns off. For faster
switching, use a higher supply voltage, with a separate resistor in
each center tap to limit the current.
I follow the instruction and it worked!!!
My connection was quite different I had collector to V+, coil to
emitter,  center tap to ground no diode at all... let me chew the
differences and if I can't grasp them firmly I'll come back...
(before I was very confused about the transistors getting real hot
jejeje)

In the sequence thing what you mention as full stepping I used to know
it as "high torque stepping" and in the full stepping the order will
be

     A  B  A'  B
1  1  0   0   0
2  0  1   0   0
3  0  0   1   0
4  0  0   0   1

and it repeats, I don’t think it matters but I just wanted to mention
it... I am using  "high torque" anyway...

what I am using it for is to move a platform in a x-axis that was
originally designed to be operated by hand and it has some points
with lots of friction... the motor gets stuck there... if there is
enough money I’m going for a more powerful source (right now I have
24v and 1 A) or a gear box... but of course first I’m going to take a
part the whole thing and put some oil and stuff ... then surf the net
and get some info to see how to make a "chopper".

Do you think it would be more expensive to make my own driver or to
buy it? What do you recommend? I need to modify 4 units.


Jerry, Matt thanks a lot.

Juan

fuentesmartinez wrote:

Me too. Now we both know better. :-)


Great! That's what it's all about.


The idea is to use them as switches, so either the current is zero, or
they're in saturation (less than half a volt across them).


There must be typos above. In this mode, there is always one side of
each coil that is turned on.


I don't use choppers. Instead, I use two supplies and the same current
sensor that a chopper would need. I use PNP transistors at power
supplies to turn the power on and off, and NPNs to ground to complete an
H-bridge. (Using the entire winding produces more torque for the same
heat, as I explained earlier.) The low-voltage supply connects to the
motor through a diode that is back biased when the high-voltage supply
is on. When the motor current rises to the design value, the
high-voltage transistor is turned off, and the coil voltage is sustained
through the diode and low-voltage transistor. This is a low-impudence
connection (a bit more elaboration can remove the diode's impedance, but
it doesn't seem to be worth it) that applies strong one-sided damping to
the motor. I've never had resonance problems with this drive even
driving a fly wheel. 100 volts on a 5-volt motor ramps the current up
real quick, and 5 volts at low impedance damps it real good. You
shouldn't need anything like that. The motor smokes if the current
sensor fails to shut off the high-voltage transistor.

Do you know anything about lapping? Use a little 400-grit emery or
norbide or carborundum in the oil, then work the slide back and forth
past the tight spot until it loosens. When it's good, take it apart and
wash it with a brush and plenty of soap and water, then dry it (a hair
dryer is good, spray it with WD-40 (sold as KantRust before WW II) to be
sure, wipe that down, re-oil and that's it. (I just did that with a
cheap pan-tilt head for my tripod. Now it's inexpensive, but not cheap.)


Ho do you value your time? It might be best to buy surplus
200-step-per-turn motors that do an adequate job with the kind of driver
you just built. Way back when, I made a circuit-board driller that way.
 
http://www.herbach.com/Merchant2/merchant.mv?Screen=PROD&Store_Code=HAR&Product_Code=TM93KIT2421

Jerry
--
Engineering is the art of making what you want from things you can get.
¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯

fuentesmartinez wrote:

Wonderful!


Taking care of the mechanical problems will definitley make your job
easier. As Jerry suggested, lapping the ways is a good idea, also, if
they are too irregular, you can read up on scraping them, but this is
not nearly as easy.


Are you modifying the units for sale or personal use? If you're selling
them, I'd be sure your design will last. If it's for home use, then
you'll probably be making improvements over time anyway. :)

Matthew Douglas Rogge wrote:

   ...


It's about time I scrape the bed of my lathe. If I take up all the play
with the gibs near the headstock, the carriage binds as I approach the
other end. The tailstock, which bears inside rather than outside the
bed, has a similar problem. Taking off a few mils is easy enough.
Keeping it straight is hard.


--
Engineering is the art of making what you want from things you can get.
¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯

Right. I visualize them like 4 coils being A-B-A'-B'--A-B-A'-B'--....
In this way  I turn on one at the time, that is why it is
1000-0100-0010-0001-1000-0100-...


low-impudence

It useful to know these things... but probably I am gonna review the
200 step-per-turn and the driver kit you suggested before I’m move
take a decision...


selling them, I'd be sure your design will last. If it's for home
use, then

I work as a teaching assistant in a lab at tecnologico de monterrey in
mexico. I talked to the guy who built them and he mentioned a lot of
short cuts that were taken to built the units because the tight
budget the school has... so the professor assign the units to me to
see whether I could improve them getting the stepper motor form old
printers and building a driver form thing we already have... before I
post the question I just follow one of the practices we teach the
students and it worked fine the problems started when I connected the
shaft to the all-tread that moves the platform... I didn’t take the
unit apart because even in the parts with "normal friction" it
performed poorly and I decided to post the question and start looking
for another solutions... now the motor performs very good and there
are few spots where it just can not do the job...

I think I can do the lapping without having to get everything
disassembled... actually I will put it in practice in 2 weeks since
the lab will be closed due to summer vacation.
I will post the end of the story by the 20-th :)

Later
Juan

fuentesmartinez wrote:

Not right. There are always two coils on, one associated with each
center tap. Putting it differently, each center tap always carries
current and except for brief transients, the current never varies and is
double what you show. Your sequence develops 71% of the sequence I
described. If total motor dissipation were the limiting factor, you
could raise the voltage to get the torque back, but with your way, the
dissipation is concentrated in 1/4 of the copper instead of 1/2. Since
the windings are bifilar, the heat distribution is the same at all and
1/2, but there's no way to get back all of the loss from going to 1/4.


There's certainly no point to going a high-performance drive before you
squeeze all you can from an L/R drive.


   ...


You can start lapping by putting the right grit into the lubricant;
that's easy. The hard part is stopping before things get worn out. You
have to remove every last trace of grit, and that usually requires
disassembly.

Jerry
--
Engineering is the art of making what you want from things you can get.
¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯

i got the error  about the transistors figured out... I understood why
my configuration is not good... and...

The project  worked... the whole thing is working now... i did the
lapping and its working...

thanks a lot Jerry and Matthew...


later

juan carlos