Re: EMC freqmod.o question

Ray, Thanks for the help. With your guidance I am finally
understanding more and more of the .ini files. Believe it of not,
the solution ended up being a faster PC. (from 400 Mhx to 1.5 ghz)
Everything works as expected now.


--- In CAD_CAM_EDM_DRO@yahoogroups.com, Ray Henry <rehenry@u...>
wrote:

>
> > Date: Mon, 01 Dec 2003 02:44:17 -0000
> > From: "earla13" <earla13@h...>
> >
> > I have been using EMC steppermod.o for some time to control a
> > Sherline Mill with a simple Stepper Word step direction

controller.

> > When I try using freqmod.o the motors run much smoother, but I
> > eventually loose control of the Y and Z axis expecially when

homing

> > or sometimes when starting up EMC. The parallel port appears to

be

> > locked up and completely shutting down and turning off the PC

power

> > corrects the problem. It is not the parallel port...changing it to
> > another brand has no effect. Could it be that my PC is not fast
> > enough for freqmod.o ? (450 Mhz Pentium, Intel BX Motherboard,

512

> > meg RAM.) I have experimented with the Sherline Mods and played

with

> > their freqmod.ini files and get exactly the same result. Does

anyone

> > have any ideas?
>
> The first batch of Sherline PCs used a 1.2 GHz processor and the

INI file

> for frequency stepping is set up for that. It is also set up for

1/4

> microstepping and the standard direct connection between the 200

steps

> per revolution motor and the 20 tpi screw. This means a value of

16000

> steps per inch of travel.
>
> The first thing that you will need to do is set INPUT_SCALE and
> OUTPUT_SCALE so that the distance you command with the EMC matches

the

> actual distance traveled by an axis. If the commanded distance is

the

> same as actual distance moved then a value of 16000 is okay. If

not you

> will have to change these scale values to match your setup.
>
> You'll need to increase the time allowed for one period and adjust

the

> trajectory max velocity variable. The Sherline values are;
>
> PERIOD = 0.000016
> ...
> MAX_VELOCITY = .36
>
> As a first guess, I'd try
>
> PERIOD = 0.00004
> ...
> MAX_VELOCITY = .2
>
> That will reduce your rapid travel to about 12 ipm but should let

it run.

> You may gradually reduce period until you see noticable delay in

the way

> that the mouse pointer moves over the mini interface.
>
> From there you can gradually increase max_velocity until you get

following

> errors on long moves. When you've found a good max_velocity you

should

> change that parameter in each of the axis definitions as well as in

the

> traj section. You don't need to have the motors powered up when

you do

> these tests and adjustments -- except for the actual vs commanded
> distance.
>
> The home thing is a puzzle, do you have real home switches? If

yes, you

> can NOT use the "Joe" or "All Zero" button. You'll have to use the
> individual home button or keyboard <home> instead.
>
> Hope this helps.
>
> Ray