Re: problem with gecko 201 dirve and cncpro

Scott,

Couple of things you might want to look at:

(1) If you have a frequency counter, set it to "totalize" and
actually count the steps sent. See if there are indeed only 40,000
sent per inch.

(2) Since the steps taken are more than what you think you commanded,
noise may be an issue. Make sure the motor leads do not share even an
inch of common wiring harness with the step and direction inputs.
Noise problems manifest themselves as a "popcorn" sound while the
motors are running. If the error is precisely repetative, then it is
not noise.

(3) See if you have the step output polarity from the PC set
to "mostly high" and the pulse width set to l uS or more.

(4) Has this problem cropped up recently or was it there from the
very beginning? If recently swap one drive with another. If the
problem follows the drive, then it is suspect. If it stays with the
axis then the wiring particular to that axis is suspect.

(5) Does it affect only one axis or all three? If all 3 then check
what is common to all 3, the PC and the power supply.

(6) If these are supported drives (they came with covers), contact me
off-list at geckodrive@... for further help.

(7) If they are the old drives then they may "squeal" with
particularly "ringy" motors (windings with significant interwinding
capacitance). The blanking period was set to an aggressive 1.5 uS
which in some cases would trip the current servo comparators before
the winding oscillations had died out. This would be irritating to
listen to but have no other effect. If they are current drives, then
your power distribution wiring may be the culprit.

(8) You mentioned the "humming" occurs when you conect the PC. This
may be a grounding problem. The PC should not share the same ground
as the drives. That is the whole purpose of using opto-isolators.

Hope that helps.

Mariss

--- In CAD_CAM_EDM_DRO@y..., "scott_in_tucson2002"
<scott_in_tucson2002@y...> wrote:

> Hello,
>
> I posted a message about this problem a couple of weeks ago, and
> have since had a chance to do some more research. My configuration
> is a sherline 5400 English milling machine, 3 gecko 201 (not

the 'a'

> version) drives, two powermax II stepper motors & one sherline
> stepper, and an XT computer running the latest version of cncpro.
>
> There appears to be two problems:
>
> 1) The steppers are 1.8 degree per step, and I compute this should
> be 40,000 pulses per inch (the gecko is a 10 microstep drive).
> However, it's taking approximately 52000-53000 pulses per inch.
>
> 2) There is some kind of accuracy problem that I'm having trouble
> quantifying.
>
> Since problem #1 seems easier to quantify and more obvious than
> problem #2, I've decided to tackle it first. Yes, my machine _IS_
> the 5400 english model, and not the metric model. I've verified
> there are 20 rotations of the handwheel per inch.
>
> I've tried two different computers, a handful of parallel IO

cables,

> and different parallel IO cards. I've also tried building and
> installing a buffer board to boost the parallel signal for driving
> the gecko's optocouplers. None of these attempts has produced any
> change.
>
> As a side note, I am getting some occasional humming from the
> motors, and only when the computer is connected (this led me to
> build the buffer board I mentioned above, which did reduce the
> humming, but did not effect the positional problems). The humming
> only occurs at certain phases where the motor is stopped.
>
> Does anyone have any clues as to what I should look to next?
>
> Thanks,
> Scott