Re: need a transformer

Jon,
The steppers are on a 2:1 ratio with the screws. That works out
too 20,000 steps per inch with microstepping and 40,000 on the Z
cause its geared down even more. The motors when they take off and
run, run smooth as silk. I did all the tuning on the geckos, got
compentation for the bigger motors set, got the trimpot set to run
as smoothly as possable. In mach2 enhanced pulsing is on, motor
ramping is set on the Z as low as possable. On the X&Y it dont have
to be so picky about the ramping but on the Z I can set it to 5IPM
with nothing but ramping and it still stalls. Zaxis also has air
assist.
The computer, AMD athalon XP 2000+ 256megs ram. Kernel speed runs
smoothly at 45K. Also the computer is setup just for mach2, nothing
else installed its on a clean install of XP. I tried slowing it to
25K and still no help. I havnt got the machine up and going at the
moment cause I just moved, need to assemble everything into the new
shop. This computer is an Athalon64 3800+ and it might go into the
shop on the mill if I build another machine for personal use. So far
havnt seen the need for a new computer here.

Jessie

--- In CAD_CAM_EDM_DRO@yahoogroups.com, Jon Elson <elson@...> wrote:
>
> rebel307 wrote:
>
> >Thanks to the 3 who replied.
> > I am useing basically an isolation tranformer then I got a
variac for the power supply. I have been told over and over to get
rid of the variac and get a "real" transformer. Art of mach software
told me that himself. Mariss of gecko told me that. They both also
said to get a transformer that will output the current rated for all
3 steppers put togather and add a tad bit more for a 4th axis.
> > I am useing 1200oz/in steppers for X&Y and one 1780oz/in for
the Z. All of them came from homeshopcnc.com. the 1200s require
6amp, the 1780 requires 6.2amp. You are saying use less amps, I put
smaller resistors on the gecko to see if it would work with less
amps, made the problem worse. I took the resistors out, didnt help
so I put the values in to give me the motors rated current. Cap in
the power supply is a 80,000MFD the caps on the drives are 470uF as
gecko said. I am not an electronics expert, dont claim to be but I
know that my machine isnt working and nothing anybody has said as
helped other then get a real transformer.
> > The problem, the motors insted of moving when a program tells
them to move lock up. Its actually ok while jogging around, throw a
program in and all hell breaks loose, they stall constantly. Make a
rapid feed more then 30IPM on X&Y and more then 10IPM on the Z and
your asking for it. I am beginning to think I should sell off the
control and just go back the reliable way,,, manual.
> >
> >
> OK, this indicates either a problem with the "tuning" of the
software,
> especially acceleration
> parameters, needs to be looked at, or, perhaps, a slower computer
is
> being used, and the
> step pulse train is "ragged". When you jog at close to 30 IPM, do
the
> motors whine smoothly,
> or do they growl, and does the growl get worse as the speed
increases?
> If so, then pulse
> raggedness is highly suspect. Gecko 201 or 202 drives require
2000 step
> pulses per rev.
> Making the assumption you have 5 TPI screws and a 1:1 drive from
the
> motors to the
> leadscrews, then you have 2000 * 5 step pulses/inch = 10,000. At
30
> IPM, that is
> 300,000 steps/minute, or 5000 steps/second. These larger motors
become
> quite sensitive
> to variations in step pulse timing. These are undoubtedly size 42
> motors, there is a jumper
> on the Gecko drives that needs to be changed for size 42 motors.
You
> have to take the case
> off the drive (at least for the 201/210 versions) to get to the
jumper.
> Have you done this?
> If not, then there is no surprise at all that you have performance
problems.
>
> You don't say what speed computer you are using. I kind of assume
you
> are using some version
> of Mach from your mention of Art Fenerty. Getting good pulse
train
> stability up to 5000
> pulses/second shouldn't be too hard with Mach, but it still
requires a
> fairly decent computer.
> Art or other Mach experts should be able to better advise you on
what
> CPU speed is required.
>
> In the long term, getting rid of the variac should be a good thing
to
> do, but I don't think it
> is responsible for your problems here.
>
> Jon
>