Re: EMC Running My Controller

Dan,

This has been changes since the 08-Jun-1999 release. Here is the release
note on it:

5. The stepper motor task now runs at a configurable pulse rate, using
the [AXIS-0,1,...] CYCLE_TIME value as the period of the pulse. Since
there's one task for all axes, it uses the shortest time. So, if
the shortest CYCLE_TIME among all axes is 0.000500 (500 microseconds),
the pulse rate is 500 microseconds per step, or 2 kilohertz. The
shortest this time can probably be is something like 200
microseconds.

Then it 14-Jun-1999 it was again modified to make it easier to set. Again,
here is the release note:

4. Automatic calculation of the stepper motor task cycle time, based
on the [TRAJ] MAX_VELOCITY and [AXIS_0,1,...] INPUT_SCALE values. The
larger the MAX_VELOCITY and the higher the INPUT_SCALE value, the
faster the stepper task needs to run to output pulses fast enough. The
procedure for setting up steppers is to set the INPUT_SCALE for each
motor to be its counts per unit (e.g., 4000 steps/inch), and then set
the MAX_VELOCITY to be the units/second value you want (e.g., 2
inches/second for 120 IPM rapid). If your motors have different values
for steps/unit, the largest one is the one that counts. Note that the
INPUT_SCALE * MAX_VELOCITY should not exceed 5000 or so, or the
stepper task timing will be too short for RT-Linux to handle it. For
4000 steps/inch, the MAX_VELOCITY should be at most 1.25
inches/second, which is a 75 IPM rapid rate. Keep in mind that your
stepper motor driver may be the limiting factor. You may need to limit
the MAX_VELOCITY to even smaller values if your motors can't run fast
enough.

Tim
[Denver, CO]

----- Original Message -----
From: Dan Mauch <dmauch@...>
To: <CAD_CAM_EDM_DRO@onelist.com>
Sent: Monday, August 16, 1999 8:19 AM
Subject: Re: [CAD_CAM_EDM_DRO] EMC Running My Controller


> From: "Dan Mauch" <dmauch@...>
>
> In the emc doc it points out that the stepper pulses are written to the
> parallel port at "an effective period of 200 microseonds for a full
> up-and -down pulse or a 5 Khz frequency"
> I did the math and they are right.
> The reason that I was interested in the pulses per second was that I have
a
> servo system that reads the signals from the parallel port. It has a 200
> count encoder on the motor . In quadrature it is 800 counts .
5000/800=6.25
> RPS*60= 375 rpms
> I may have to look at using the ISA buss.
> -----Original Message-----
> From: George Potter <gpotter@...>
> To: CAD_CAM_EDM_DRO@onelist.com <CAD_CAM_EDM_DRO@onelist.com>
> Date: Sunday, August 15, 1999 9:47 PM
> Subject: Re: [CAD_CAM_EDM_DRO] EMC Running My Controller
>
>
> >From: George Potter <gpotter@...>
> >
> >I don't know where the figure 5 Khz comes from, as the
> >"normal" parallel port hardware (assuming based on that
> >disclosed in the "original" IBM AT Handbook is only
> >limited by how fast your software can write to it ....
> >(up to the ISA I/O card bus speed). (I believe that
> >even on "slow" machines it should be possible to get
> >around 100 kbytes / second out the port. (The
> >"newer" ECP (etc.) types of ports should support
> >7 or 800 kbyte through put, if I remember correctly).
> >(all of my support books are at work). The capacitors
> >shouldn't be a limiting factor.
> >
> >If someone really needs to know, I can write a few lines
> >of code, and see what I can get out of one (meaning
> >hardware limits), but it will take a few days to get around
> >to it.
> >
> >In a past life, I used the parallel port to run various production
> >test hardware, and it was amazing what could be run that way.
> >
> >Regards,
> >George Potter
> >
> >Jon Anderson wrote:
> >
> >> From: Jon Anderson <janders@...>
> >>
> >> Dan Mauch wrote:
> >>
> >> > Now that I have emc up and running I went back and read the
> >> > docs. I discovered that using the parallel port you are
> >> > limited to 5KHZ thru the parallel port.
> >>
> >> I don't know if this is true, or if it would help, but I was told that
> >> there are small capacitors on the I/O lines of the parallel port.
> >> Clipping these is -supposed- to greatly increase rate of data transfer.
> >> Obviously, it would be better to try this on an add-in card than on the
> >> motherboard...
> >> I don't know enough about electronics to feel confident about what
> >> capacitors to clip, nor measure the effects. Just thought I'd toss it
> >> out for someone else to comment on.
> >>
> >> Jon
> >>