Re: Microstepping thru the Tulips

1) It doesn't even take an FPGA; a lowly 64-macrocell CPLD does the
job nicely. No configuration EEPROM and JTAG controller to deal with
that way.

2) 500MHz is like garlic to a vampire in a drive circuit. A 500 MHz
FPGA responds to 1nS-wide noise pulses and a motor drive is very rich
in noise. Get the slowest part you can and wish it was slower yet.

3) Motor power output is proportional to V / SQRT L. MOSFETs, circuit
design and such have no bearing on this identity.

4) Step motors typically have a +/-5% non-accumulative tolerance.
This means the ultimate open-loop accuracy is 1:2,000 for a 1.8-
degree motor or 1/10th of a full step. Using the time-tested idea
that your'e doing OK if your circuit accuracy is one order of better
than what you are controlling, a 7-bit DAC is dandy.

5) PWM fequency and microstep resolution have no relationship at all.
Please see message #89337 that directs you to pictures that prove it.

6) What does it take? Actually a bunch of other stuff, mostly analog.
Mid-band resonance compensation, short-circuit protection, reverse
polarity protection, standby current, current set, sine to full-step
reference morphing, etc, etc.

Why analog? Take current set as an example: We agreed a 7-bit
resolution for a particular current is dandy. It is also reasonable
to have a 0.1A to 7A current set range. This is about a 6-bit range;
to do it directly would require a 13-bit DAC. Not something I'd like
to do without fear.

Mariss

--- In CAD_CAM_EDM_DRO@yahoogroups.com, "Dennis Schmitz"
<denschmitz@...> wrote:

>
> My random thoughts about stepper speed, in case you're interested.
>
> Thinking of the problem from the perspective of requirements, what

is the

> maximum speed you can get out of a stepper? Consider it as a two

phase AC

> servo, driven with sine wave inputs. As the frequency of the input

goes up,

> the current lags and falls in amplitude. Eventually the reactance

of the

> motor can't be compensated to maintain current limited by the

voltage rating

> of the windings.
>
> Clearly low inductance, high current steppers can be operated

faster. Used

> to be that a tradeoff existed in power amps preferring lower

current and

> higher voltages. MOSFETs get better every year and now there's no

real

> reason to design in the low-inductance steppers from the beginning.
>
> In any case, you can calculate your maximum frequency from your

power supply

> voltage and current drive. Just an outside bound. Ideally, you'd

like a PWM

> at least 2x with at least 8 bits resolution. In any case cheap

FPGAs can run

> 500MHz and can be had with built-in multipliers.
>
> So what's to keep an FPGA connected to a power amp from being a

universal

> motor controller? (Other than someone has to write the code.)
>
> On 8/14/06, Steve Stallings <stevesng@...> wrote:
> >
> > violent agreement indeed, except for your assumption
> > of what most folks consider high speed.
> >
> > If we assume Gecko drives microstepping at 10 micro-
> > steps per full step and the known appx. 20KHz PWM
> > rate, then we get a possible 10 rev. per second or
> > 600 RPM with microstepping still nominally functioning.
> >
> > Apply this to a Bridgeport with 5 TPI screws and
> > 2:1 gearing (typical of Boss series) and you get
> > 60 inches per minute. Not blindingly fast, but a
> > long ways from slow.
> >
> > Cheers,
> > Steve Stallings
> > www.PMDX.com
> >
> >
> > --- In CAD_CAM_EDM_DRO@yahoogroups.com, "Phil Mattison"
> > <mattison20@> wrote:
> > >
> > > This diagram illustrates my point nicely. We may be in violent
> > agreement, as
> > > we used to say at Intel. The ramp time at the beginning of each
> > step depends
> > > on motor inductance and power supply voltage. It cannot be

reduced

> > without
> > > changing one of those variables. The reciprocal of that time is

the

> > > frequency at which the chopper (PWM, fixed off-time or whatever)
> > ceases to
> > > have any effect. On this diagram it is a little more than half

the

> > period of
> > > the step frequency, so let's say for the sake of argument the
> > > chopper-irrelevant frequency would be 2*460, or about 920 steps

per

> > second.
> > > That's well below the top speed of most motors I know of. One

thing

> > is
> > > indisputable: there is no microstepping unless the chopper can
> > modulate the
> > > current.
> > >
> > > So the question is, at what step rate can the chopper modulate

the

> > current
> > > in a meaningful way? If you look at a 16x microstepping drive,

the

> > > literature claims to provide 8 different current levels during a
> > single
> > > motor step (8 ramping up, 8 ramping down.) Best case, the

chopper

> > would have
> > > to operate for at least 16 cycles during the motor step. That

means

> > the
> > > chopper is relevant only at full-step rates at or below 1/16th

of

> > the
> > > chopper frequency, which commonly seems to be 20 or 30 KHz. In
> > other words,
> > > maybe 1000 to 2000 steps per second, depending on the motor. As

we

> > have
> > > seen, beyond that it doesn't matter anyway because the chopper

does

> > nothing.
> > >
> > > I would suggest that what microstepping really buys you is the
> > ability to
> > > operate the motor in the resonance frequency range with less

chance

> > of
> > > stalling, and that contrary to many comments I have seen, it has
> > little or
> > > no effect on high-speed performance. I suspect that drivers
> > offering the
> > > best high-speed performance simply have better control of their
> > switching
> > > times.
> > > --
> > > Phil Mattison
> > > http://www.ohmikron.com/
> > >
> > > ----- Original Message -----
> > > From: John Dammeyer <johnd@>
> > > To: <CAD_CAM_EDM_DRO@yahoogroups.com>
> > > Sent: Sunday, August 13, 2006 11:54 PM
> > > Subject: RE: [CAD_CAM_EDM_DRO] Re: Microstepping thru the Tulips
> > >
> > >
> > > > Hi all,
> > > >
> > > > Text descriptions can be difficult to follow if you haven't
> > played with
> > > > chopping stepper drives.
> > > >
> > > > I tried placing this picture in the Yahoo group but can't

find a

> > button
> > > that
> > > > lets me do that so instead here's a link:
> > > >
> > > > http://www.autoartisans.com/images/stepcurrent.jpg
> > > >
> > > > The motor driver is an LMD18245 micro-stepping driver. The

top

> > scope
> > > trace
> > > > shows the current sensed by the driver. The bottom trace

shows

> > the
> > > voltage
> > > > across the winding which makes the current flow through the
> > winding.
> > > >
> > > > The step rate is about 460 Hz and the motor is turning so it
> > creates both
> > > > back emf as the armature works through the windings and also

as

> > the
> > > magnetic
> > > > field from the previous step collapses.
> > > >
> > > > Hopefully the labels on the photo show everything clearly but
> > I'll add
> > > less
> > > > than a thousand words anyway. Forgive me if I'm not totally
> > clear on this
> > > > as I did this back in Nov. 2005 and didn't keep any notes on

it.

> > > >
> > > > 1. When the signal to push current through a coil is turn on,

the

> > voltage
> > > to
> > > > the winding is applied and current is supposed to flow.
> > > >
> > > > 2. But because it's an inductor current doesn't start right
> > away. It is
> > > > also prevented because initially there is an opposite voltage
> > bucking the
> > > > applied voltage (back emf). Until this back emf decays no
> > current will
> > > flow
> > > > through the winding.
> > > >
> > > > 3. Finally, the back emf is lower than the applied voltage

(and

> > still
> > > > falling) and we start to see current flow through the winding.
> > > >
> > > > 4. When the maximum current is reached the applied voltage is
> > removed and
> > > > the current starts to decay. When the current is below the

set

> > point, the
> > > > applied voltage is there again and the current climbs.
> > > >
> > > > 5. This happens forever if the motor is stopped; the current
> > averages out
> > > > to the setpoint value.
> > > >
> > > > 6. Finally since the motor is turning the whole things starts
> > again with a
> > > > different winding and the current builds up again.
> > > >
> > > > So where does micro-stepping enter into this you might ask?

If

> > you look
> > > at
> > > > the scope photo the yellow trace stops rising at about 2V.

This

> > may be
> > > > equivalent to 3 amperes of current through the winding. If

the

> > reference
> > > is
> > > > changed so that the yellow trace stops at 1V then there would

be

> > only 1.5
> > > > amperes of current flowing through the winding.
> > > >
> > > > Where 3 amperes in one winding and 0 amperes in the other

pulls

> > the
> > > armature
> > > > around to a magnetic pole 1.5 amperes in one and 1.5 in the

other

> > holds
> > > the
> > > > armature between the poles. Various ratios between the two

hold

> > the
> > > > armature at other places. The high quality micro-steppers
> > simulate a two
> > > > sine waves offset by 90 degrees and the armature smoothly

moves

> > between
> > > the
> > > > detenting magnetic poles. (There's more to this but this is

the

> > simple
> > > > explanation).
> > > >
> > > >
> > > > Another interesting phenomenon that the scope photo shows

clearly

> > is why
> > > the
> > > > torque drops off as the motor turns faster. If the step

pulses

> > were
> > > > happening twice as fast, the photo would show the current

rising

> > to about
> > > > half of the ramp before the direction of the current was
> > changed. Less
> > > > current, less torque. In fact it wouldn't even start

chopping.

> > > >
> > > > If the step rate increases to a point where the current

hasn't a

> > chance to
> > > > begin to flow, the armature just locks up and doesn't move.

No

> > current,
> > > no
> > > > torque.
> > > >
> > > > Finally, if the applied voltage was changed to a higher value
> > then the
> > > slope
> > > > of the ramp would be steeper and maximum current would be

reached

> > faster.
> > > > The back emf value doesn't change because the back emf is

based

> > on the
> > > > current flowing through the windings when it's interrupted at

the

> > motor
> > > > speed; not the applied voltage.
> > > >
> > > > And, if I used a low inductance motor here instead of a

surplus

> > high
> > > > inductance motor, the slope of the current curve would also be
> > > dramatically
> > > > steeper. That also results in more torque at medium RPM and a
> > higher top
> > > > speed.
> > > >
> > > > Hope that helps a bit.
> > > >
> > > > John Dammeyer
> > > >
> > > >
> > > >
> > > >
> > > > Addresses:
> > > > FAQ: http://www.ktmarketing.com/faq.html
> > > > FILES: http://groups.yahoo.com/group/CAD_CAM_EDM_DRO/files/
> > > > Post Messages: CAD_CAM_EDM_DRO@yahoogroups.com
> > > >
> > > > Subscribe: CAD_CAM_EDM_DRO-subscribe@yahoogroups.com
> > > > Unsubscribe: CAD_CAM_EDM_DRO-unsubscribe@yahoogroups.com
> > > > List owner: CAD_CAM_EDM_DRO-owner@yahoogroups.com,

wanliker@,

> > > timg@
> > > > Moderator: pentam@ indigo_red@ davemucha@
> > > [Moderators]
> > > > URL to this group:

http://groups.yahoo.com/group/CAD_CAM_EDM_DRO

> > > >
> > > > OFF Topic POSTS: General Machining
> > > > If you wish to post on unlimited OT subjects goto:
> > > aol://5863:126/rec.crafts.metalworking or go thru Google.com to
> > reach it if
> > > you have trouble.
> > > > http://www.metalworking.com/news_servers.html
> > > >
> > > > http://groups.yahoo.com/group/jobshophomeshop I consider

this

> > to be a
> > > sister site to the CCED group, as many of the same members are
> > there, for OT
> > > subjects, that are not allowed on the CCED list.
> > > >
> > > > NOTICE: ALL POSTINGS TO THIS GROUP BECOME PUBLIC DOMAIN BY
> > POSTING THEM.
> > > DON'T POST IF YOU CAN NOT ACCEPT THIS.....NO EXCEPTIONS........
> > > > bill
> > > > List Mom
> > > > List Owner
> > > >
> > > >
> > > > Yahoo! Groups Links
> > > >
> > > >
> > > >
> > > >
> > > >
> > > >
> > >
> >
> >
> >
> >
> >
> >
> >
> > Addresses:
> > FAQ: http://www.ktmarketing.com/faq.html
> > FILES: http://groups.yahoo.com/group/CAD_CAM_EDM_DRO/files/
> > Post Messages: CAD_CAM_EDM_DRO@yahoogroups.com
> >
> > Subscribe: CAD_CAM_EDM_DRO-subscribe@yahoogroups.com
> > Unsubscribe: CAD_CAM_EDM_DRO-unsubscribe@yahoogroups.com
> > List owner: CAD_CAM_EDM_DRO-owner@yahoogroups.com, wanliker@...,
> > timg@...
> > Moderator: pentam@... indigo_red@... davemucha@...
> > [Moderators]
> > URL to this group: http://groups.yahoo.com/group/CAD_CAM_EDM_DRO
> >
> > OFF Topic POSTS: General Machining
> > If you wish to post on unlimited OT subjects goto:
> > aol://5863:126/rec.crafts.metalworking or go thru Google.com to

reach it

> > if you have trouble.
> > http://www.metalworking.com/news_servers.html
> >
> > http://groups.yahoo.com/group/jobshophomeshop I consider this

to be a

> > sister site to the CCED group, as many of the same members are

there, for OT

> > subjects, that are not allowed on the CCED list.
> >
> > NOTICE: ALL POSTINGS TO THIS GROUP BECOME PUBLIC DOMAIN BY POSTING
> > THEM. DON'T POST IF YOU CAN NOT ACCEPT THIS.....NO

EXCEPTIONS........

> > bill
> > List Mom
> > List Owner
> >
> >
> > Yahoo! Groups Links
> >
> >
> >
> >
> >
> >
> >
> >
>
>
> [Non-text portions of this message have been removed]
>