Suitable Cap for Gecko's was Re: 8-wire 1/2coil stepper?

Alan,

Good afternoon! A full step sequence is where A and B are the
windings, + and - is the "direction" of the current thru the winding.

+A +B, +A -B, -A -B, -A +B

After this, the sequence repeats.

For microstepping, pretend "A" is sine and "B" is cosine magnitude.
Then "+" and "-" are the "signs" for the sine and cosine functions.

For the full step sequence (nothing more than a single-bit
approximation of the sine and cosine functions), the motor moves 1.8
degrees on each "sign" change. Being sine and cosine approximations,
the sign changes occur every 90 "electrical" degrees. 360 degrees in
a cycle, therefore 4 full steps per cycle.

You can use a bridge to measure inductance but your reading will be
high because it is taken at zero DC current. Magnetic saturation
effects (where you are on the B-H curve) will cause the working
inductance to be somewhat lower.

I prefer to measure working inductance directly from the V/L slope
measured with a 'scope. Read the "delta I" over a "delta t" period on
the 'scope. As an example, say you measure a 300mA change in winding
current over a 12uS time span and you are using a 72VDC power supply.

The inductance will be 72V times 12uS / 300mA, or 2.88mH if the slope
was linear.

The cap is fine for power supply voltages up to 50VDC.

Mariss

--- In CAD_CAM_EDM_DRO@y..., Alan Marconett KM6VV <KM6VV@a...> wrote:
> Hi Mariss,
>
> Very informative! Another "keeper"! Thanks for the
clarification. Why
> is it 4 steps/cycle? Seems like something is changing (a current
> through a winding) EVERY step?
>
> I see a "Fl = 2 PI F L" inductive reactance calculation! Can I
measure
> the inductance of "unknown" stepper motors with an L/C meter (kit)
such
> as:
>
> http://www.aade.com/lcmeter.htm
>
> I think the operating frequency is 1K. I don't have an HP4275A
digital
> L/C meter, or or a B&K 878 digital LCR meter!
>
> Also, is this 470uf cap suitable for the Gecko's (only 63v, couldn't
> find higher)?
>
> P10351-ND EEU-FC1J471L CAP 470UF 63V ELECT FC RADIAL Panasonic -
ECG
> 470µF 63V ±20%
>
> Found in DigiKey's catalog (only a quick online search).
>
> Alan KM6VV
>
>
> mariss92705 wrote:
> >
> > Hi,
> >
> > I think the thing being missed here full-stepping vs.
microstepping.
> >
> > If you have a 5V 1A per phase unipolar rated, 8-wire motor, then
you
> > would run it at 1A unipolar (half-winding) and it would dissipate
5W
> > (1A squared times 5 ohms).
> >
> > If you connect the windings in parallel, then the winding
resistance
> > is 2.5 ohms (1/2 what it was in half-winding). You can now
increase
> > the current to 1.41A and still dissipate 5W (1.41A squared times
2.5
> > ohms = 5W).
> >
> > What matters here is your torque (from the URL you referenced)
> > increased 30%. It should have increased 41%; the reason it didn't
is
> > because the iron has magnetically saturated.
> >
> > If you are full stepping, you don't care. There is no negative
> > positioning effect from saturating the iron except for a drop in
> > inductance.
> >
> > If you are microstepping you do care because saturated iron means
a
> > non-linear motor and the microstep positions will begin to bunch-
up
> > at the full step locations.
> >
> > An analogy is an audio amplifier. If you overdrive it with a sine
> > wave, the output will limit or "clip" and produce distortion. If
all
> > you are amplifying is a square wave, then who cares if it clips?
> >
> > Second point. Take the same motor and drive it at 1.41A half-
winding.
> > You will not be able to tell it apart from the parallel connected
one
> > at 1.41A torque-wise. The only difference is it will dissipate 10W
> > (1.41A squared times 5 ohms).
> >
> > Third point. This 30% increase in torque occurs at low speeds
where
> > step motors usually have surplus torque anyway. It does not get
you a
> > single extra oz-in of torque at high speed, where improvement
would
> > be appreciated.
> >
> > Inductive reactance becomes the bottleneck that limits the winding
> > current. The drive may be set at 1A or 1.41A, it will make no
> > difference. 10 mH of inductance will have 62.8 ohms of inductive
> > reactance at 4,000 full steps per second and will limit winding
> > current to 0.5A for a power supply voltage of 31.4VDC.
> >
> > I = 31.4V / 2 pi times 1kHz times 0.01 Henries
> > 4 full steps is one electrical cycle, so 4,000 full steps/sec =
1kHz
> >
> > Mariss