2 May 2012 14:13

## Re: Electronic damping of stepper mid-band resonance, was Re: Drive Tuning

```Better to hear from Mariss himself.

---------------------------------------------------------------------------------
It's a give and take kind of situation:

1) For the same peak current, a microstepped motor will have 71% (1/sqrt 2)
the holding torque of a full-step drive. This is because motor torque is the
vector sum of the phase currents. Advantage: Goes to full-steppers.

2) Most people want motors to turn, not just 'hold'. As soon a full-step
driven motor turns, its torque drops to 65% of its holding torque. Where did
the missing torque go? To resonating the motor is where. Motor mfgs sometimes
specify 'dynamic torque'; this is specified at 5 full steps per second. It is
always between 60 to 65% of holding torque. Not mentioned is the horrible
racket the motor makes at 5 full steps per second.

Microstepped motors do not resonate at low speeds, so no torque is invested in
resonance. Microstepped motors keep all their holding torque while turning
slowly. 65% for full-steppers, 71% for microsteppers. Advantage: By a hair
(6%), goes to microsteppers.

3) Things get a little dicy as speed increases. Microstepping ceases to have
any benefit above 3 to 4 revolutions per second. The motor is now turning fast
enough to not respond to the start-stop nature of full steps. You can say the
step pulse rate is above the mechanical low-pass frequency limit (100Hz or so)
of the motor. Motion becomes smooth either way.

Simple drives persist in microstepping anyway above this speed. This means
they still try to make the motor phase currents sine and cosine past this
speed. A little problem with that and it's called 'area under the curve'. The
area under the sine function (0 to 180 degrees) is only 78% of a square wave
(full-step). Advantage: Goes to full-step again.

More sophisticated drives transition from a sine-cosine currents to
Draw.

4) As speed increases even more, another really big problem crops up; mid-band
resonance. This is the bane of full-steppers and microsteppers alike.

Maybe you have experienced it; the motor is turning 5 to 15 revs per second
when you hear a descending growing sound from the motor and then it stalls for
no good reason at all. Faster it's OK, slower it's OK, but not OK in that
range. All you know is there is a big notch in the speed-torque curve. This
mid-band instability, or parametric resonance.

Simple drives have no defense against this except to try not run the motor in
this speed range. Better drives have circuitry to suppress this phenomena and
it involves rate damping.

This is the equivalent of shock absorbers (rate dampers) on a car, without
them a car bounces repeatedly. Imagine a washboard road surface in sync with
this bounce; there would be sparks flying from the undercarriage in short
order. With rate dampers the 'bounce' is suppressed to a single cycle.
Mid-band compensation does the same with steppers.

5) More than any other type of motor, step motor performance is tied to the
kind of drive connected to it. More than any other type motor, a stepper can
be driven from very simple drives (full-step unipolar L/R) to very complex
ones (microstepping full-bridge bipolar synchronous PWM mid-band compensated).
-----------------------------------------------------------------

Cheers,

Peter

On 2/05/2012 1:28 PM, Jon Elson wrote:
> andy pugh wrote:
>> On 1 May 2012 12:26, cogoman<cogoman@...>  wrote:
>>
>>> I don't see how they could switch from 1/10 to full step without letting
>>> LinuxCNC know, and having LinuxCNC reduce the number of steps being
>>> sent, unless they used a clock multiplier, which would make it look like
>>> full step to the control,
>>>
>>
>> I imagine it is an internal clock-divider, so at high speed it
>> full-steps every N input pulses, and at low speed it microsteps every
>> input pulse.
>>
> I've always been very suspicious of this claim (that Geckos switch from
> microstepping to
> full steps at some speed).  it seems totally unnecessary, and might be
> hard to do without
> causing some manner of glitch.  What I think really happens is that at
> some speed
> the sinusoidal current command gets enough ahead of the motor's
> inductance that
> the winding current never reaches the current setpoint, and so the
> transistors naturally
> switch from chopping mode to regulate current to a mode where they are
> on for the
> half electrical cycle of that winding.  I expect every microstepping
> chopper drive
> will do the same.
>
> In other words, the drive does this naturally due to the lag of the
> motor's inductance,
> and there is no special circuit at all to perform this function.  And,
> the speed
> at which this happens is determined by the DC supply voltage and the motor
> inductance.
>
> Jon
>
> ------------------------------------------------------------------------------
> Live Security Virtual Conference
> Exclusive live event will cover all the ways today's security and
> threat landscape has changed and how IT managers can respond. Discussions
> will include endpoint security, mobile security and the latest in malware
> threats. http://www.accelacomm.com/jaw/sfrnl04242012/114/50122263/
> _______________________________________________
> Emc-users mailing list
> Emc-users@...
> https://lists.sourceforge.net/lists/listinfo/emc-users
>

--

--
-----------------------------------------------------------------------------
eStore: http://www.homanndesigns.com/store
Web   : http://www.homanndesigns.com     ModIO     - Modbus Interface Unit
email : peter@...          DigiSpeed - Isolated 10Vdc I/F
Phone : +61 421 601 665                  TurboTaig - Taig Mill Upgrade board

------------------------------------------------------------------------------
Live Security Virtual Conference
Exclusive live event will cover all the ways today's security and
threat landscape has changed and how IT managers can respond. Discussions
will include endpoint security, mobile security and the latest in malware
threats. http://www.accelacomm.com/jaw/sfrnl04242012/114/50122263/
```

Gmane