Before I explain the problem, here is what i'm using.

Leadshine DCS303 DC Servo Drive                              http://www.leadshine.com/UploadFile/200910162592891415.pdf
Pittman 9234 Brushed 24V DC Servo
US Digital HEDS-9100 512cpr Optical Encoder

Using the software that came with the drives the servo's work great, i can jog them all over and have no issues, they are fast and very quiet.
With mach3 i hardly get them to work at all. The Drive has PUL/DIR inputs, i have them hooked up to the para port and the pins are setup correctly in Mach. Using unmodified tuning values in mach, When i try to manually jog in mach the servo spins but it looks like its pulsing back and fourth while spinning, and almost sounds like a stepper motor, and it only turns in one direction, its just alot slower in one direction than the other.

If i go into the motor tuning window, and put 250 in the "Counts per" and 500 in the field right next to it (can't think of what it is right now, no mach on this computer) i can get it to jog smoothly and its quiet, but it goes faster one way than the other, and if i try to run a program its right back to vibrating and jerking around. If i set the feed rate just right and its on a long straight part of the program, it will turn smooth, but thats the only time..

If i set the "Counts per" or the one next to it to anything +- 20 or so then it will go back to running bad, the more i stray from these numbers the worse it gets.

If anyone has any ideas at all as to what could be the problem here i'd really like to hear them, i've been trying to get it working for days now and i'm about to blow my top.

First thing is work out your steps per unit and put them in correctly, next thing is the Vel which is the max speed of your motors and lastly the acceleration, this will depend on your machine/motors/drives etc and best to start low and work up.
 

Hood

i worked out the steps per several times, and even tried mixing things up to get a different number and it still would not work right. Half the time the motor spins the same direction when you try to jog either direction..

I downloaded linux emc2, and it worked right after installed. I did some tuning with the electronic gear ratio on the driver and now its even working inch for inch. With the current setup i've obtained 2.2inch/second (132in/min) feed rate and it is very smooth and almost silent.

I'm still not sure what configuration was wrong in mach, i've tried about everything i could think of to make it work and it just wouldn't work.

All i did with emc2 was load the configuration program, it had a list of different pre-configured setups, just for the heck of it i chose shereline and then corrected the pin out to match what i had wired up, and it worked fine, just very slow. So i used my drive tuning software to change the electronic gear ratio to 25:1 and i guess just out of luck i told it to move the X axis -10 inches, and it moved it -10 inches. I don't know how it was that easy but it just worked. I'm not preaching saying emc2 is better but in my case it just worked.

Just for the sake of it could anyone here interpret the emc2 config file i'm using, and see if they can't figure out what settings need to be put in place in mach for it to work properly? This may help others in the future using the same setup that i am.

I plan to go ahead and use emc2 since it has all the functionality i need and it was easier to work with (not to mention its free, and runs on a free operating system, and i do happen to be a linux fan).

You use Sherline in EMC, did you try that with Mach? Sounds like your drives are needing a large pulse width.

Hood

Hmm.. yes i will definitely try that and see. Here is a little info out of my drive manual:

"    Pulse signal: This input represents pulse signal; 4-5V when PUL-HIGH,
    0-0.5V when PUL-LOW. For reliable response, pulse width should be longer
    than 2.5μs. Series connect resistors for current-limiting when using +12V or
    +24V control signals (OPTO). The same as DIR and ENA signals."

and:

"    DIR signal: This signal has LOW/HIGH voltage levels, 4-5V when
    DIR-HIGH, 0-0.5V when DIR-LOW. For reliable motion response, DIR
    signal should be ahead of PUL signal by 5μs at least. Please note that motion
    direction is also related to motor-driver wirings and encoder feedback A/B
    phase signals. Exchanging the connections of two wires of the motor to the
    driver and A/B phase signals of encoder will reverse motion direction."

I wish i just had an oscilloscope so i could compare the two signals (mach vs. emc2) i'm getting and compare the difference..

I am not sure if you mentioned the specs of your computer but even high spec computers can be a bit dodgy with jitter and need a wider pulse than would be normal. I think the default in Mach is Zero (obviously not zero in real terms )
  I dont know what EMC has as a default when running Sherline but Mach is 40uS if I recall, that would seem too high for your drives but setting the pulse width to 5uS  and Dir Prechange the same, without using Sherline may help a lot.

Hood

i tried messing with the pulse width paramaters in the motor tuning window, didn't matter what i put in they nothing changed. And yes i did save it lol

What are the specs of the computer? Have you done the drivertest to see what the pulse is like?
Hood

the driver test looked good, a perfectly flat line with a very small blip every now and then.

The computer is a 2.53ghz Celeron D with a gig of 400mhz DDR.
My other test computer has a 2.0GHZ p4 w/ 1gb ram and it does the exact same thing.

Both of them have fresh windows installs, only drivers loaded, no crapware running in the background to mess things up. There both running in tip top shape.

Seems strange then, maybe for some reason your drives do require Sherline.

Hood