Hello Guest it is March 28, 2024, 08:48:46 AM

Author Topic: trouble with steps or with stream pulses  (Read 7242 times)

0 Members and 1 Guest are viewing this topic.

Offline nemoc

*
  •  22 22
    • View Profile
trouble with steps or with stream pulses
« on: August 25, 2007, 03:47:56 PM »
Hello, I’m new to Mach3 and English isn’t my natural language therefore I’m sorry for something unintelligible.
      I think that the subject is trouble with steps or pulses.  ???
Before I post I have read everything that I found here about the subject , and I did most of  things that are indicated but don't fix my problem .
I have a little router (a homebrew CNC) that I build four years ago:
Step motors (130N-cm; 2A), ball screw (0.2in/rev), PowerSuply (38V-330W), stepdrivers optoisolated (L297+L6203), breakout board is buffered.
PC – Asus-AMD-Atlhon 1800+, Ram 256Mb
Recent I bought a Mach3 license and after this I’m running another CNC control.
Now I’m running   Mach3VersionR2.00.053.
This is the problem: When I move any axis (jog/MDI/program) I get a cumulative error, something about 0.001 inches although the DROs show the exact displacement programmed. This occurs in one direction but not in the other for the same axis.  The error is consistent and is the same for all axes (x, y, and z). Can someone Help me ?

NEMO
« Last Edit: August 26, 2007, 07:28:43 AM by nemoc »
Re: trouble with steps or with stream pulses
« Reply #1 on: August 26, 2007, 09:14:30 AM »
sounds like you have backlash in the screw... try setting your backlash comp to.001 and see if that is any help..

thanks
Brian
Fixing problems one post at a time ;)

www.newfangledsolutions.com
www.machsupport.com

Offline nemoc

*
  •  22 22
    • View Profile
Re: trouble with steps or with stream pulses
« Reply #2 on: August 26, 2007, 12:14:06 PM »
Thanks Brian.
I have tried with backlash comp. The final numbers are changed but the behavior is the same. Like I said the error is in one direction for the same axis. I 'm running the program below with or without backlash compensation and the behavior (not numbers) is the same for all axes. I have a DTI installed to check this. I have already measured the ball screw backlash and they are something about 0.0004in and they are new.

N002 G90 G61
N003 G01 Y0.0 F20
N004 G01 Y1.0 G4 P2
N005 G01 Y0.0 G4 P2
N006 G01 Y1.0 G4 P2
N007 G01 Y0.0 G4 P2
N008 G01 Y1.0 G4 P2
N009 G01 Y0.0 G4 P2
N010 M05
N011 M30
Thanks

NEMO
« Last Edit: August 26, 2007, 10:31:44 PM by nemoc »

Offline jimpinder

*
  •  1,232 1,232
  • Wakefield, West Yorks, UK
    • View Profile
Re: trouble with steps or with stream pulses
« Reply #3 on: August 27, 2007, 03:57:39 AM »
You say the error is cumulative. Does this mean that you are getting 1/3 thousanth error on each move - or is the error one thousanth each move, but forward and back it just corrects itself.

What if you move to Y3 in one move - do you get the same error.

Are you sure your error is not in the DTI.

The fact you are getting this on all axis - and I assume you are using the same DTI,  really eliminates an error on any one of the axis and it is unlikely you will get the same error on all three.

If you have eliminated backlash i.e. -  Y-1 to Y0 - start -  then Y1, Y-1 and back to Y0 with zero error then this would seem to indicate that all is well with your mechanical set up.
Not me driving the engine - I'm better looking.

Offline nemoc

*
  •  22 22
    • View Profile
Re: trouble with steps or with stream pulses
« Reply #4 on: August 27, 2007, 03:16:29 PM »
Thanks jimpinder.
Below answer for your questions.
You say the error is cumulative. Does this mean that you are getting 1/3 thousanth error on each move - or is the error one thousanth each move, but forward and back it just corrects itself
       - Actually, we have one thousanth each move.
What if you move to Y3 in one move - do you get the same error.
       - Yes correct.
Are you sure your error is not in the DTI.
      - I borrow one from my friend and check it . It's OK!
The fact you are getting this on all axis - and I assume you are using the same DTI,  really eliminates an error on any one of the axis and it is unlikely you will get the same error on all three.
      - Sorry , I don't understand but  I test one axis each time not all at the same time .
If you have eliminated backlash i.e. -  Y-1 to Y0 - start -  then Y1, Y-1 and back to Y0 with zero error then this would seem to indicate that all is well with your mechanical set up.
      - I do a test as follow:
      - Move Y-1 to Y0 to eliminate backlash then install DTI at X0 named hear "P0".
      - Run Gs below.
            G20 G90
            M98 P123 L50
            M30
            O123
            G1X1
            G1X0
            M99
      - After many repetition of subroutine  the called "P0" is far front DTI and don't tap it .
      - I think that the system lose pulses/steps in one direction but not in the other . That is corret?

Thanks

NEMO

Offline jimpinder

*
  •  1,232 1,232
  • Wakefield, West Yorks, UK
    • View Profile
Re: trouble with steps or with stream pulses
« Reply #5 on: August 27, 2007, 06:34:40 PM »
You must eliminate your backlash first.

Do not run a G code program - this introduces too many problems.

Run the axis in one direction by jogging, then set up your DTI

On the Midi Alt 2 Page, manually put in G code . First zero the DROs then type in the input G0 X1 (making sure the axis moves in the same direction as it did when jogging). If your DTI is still in contact - did the axis move 1 inch. If NOT your number of pulses per inch on the motor is not correct.

Without moving anything, type in G0 X0 - does the DTI return to zero.

Have you got dials on your machine - do they return to where they started???
Have a go at that then come back.
Not me driving the engine - I'm better looking.

Offline nemoc

*
  •  22 22
    • View Profile
Re: trouble with steps or with stream pulses
« Reply #6 on: August 27, 2007, 11:34:27 PM »
Thanks for your post jimpinder.
Today I do all that you suggested , many times. Yesterday I did the same test.
- Without a conclusion.
Now, I did something radical.
I uninstall Mach3 and install my old cnc control and I was running with it now.
The error (0.001 inch) disappears. CNC runs stable and fine.
I measured the backlash that is 0.0002 inch or 0.005 millimeters. Very good .
Now I can eliminate backlash when I move in the same direction.
I’m sure that the error (0.001”) isn’t a backlash.

Jimpinder
This sounds like an electronic problem.
Thaks
Nemo

Offline jimpinder

*
  •  1,232 1,232
  • Wakefield, West Yorks, UK
    • View Profile
Re: trouble with steps or with stream pulses
« Reply #7 on: August 30, 2007, 01:10:44 PM »
Yes - I would agree. If you now know your backlash you can enter this in Mach3, and that should be that - all moves shold be precise.

I do not know anything about the way Mach 3 puts out pulses to control the stepper motors, so I must pass to someone else.

Jim
Not me driving the engine - I'm better looking.

Offline nemoc

*
  •  22 22
    • View Profile
Re: trouble with steps or with stream pulses
« Reply #8 on: August 30, 2007, 04:50:05 PM »
Thanks Jim for your help.
I install Mach3 again and do many experiments.
I put a reduction between Y ball screw and step motor ratio 2:1 slow down more accel and speed. I was thinking about inertia then a reduction may be help.
Changed a setup of “Pulses per” from 2000 to 4000 (ratio 2:1).
The error now (eliminate backlash) is 50% of the past 0.001”. The error is 0.0005” now.
Well the system loses pulses.
Someone know how Mach3 work with Dir Pulse and Stream of Step Pulses?
I need a help with this.

Nemo

Offline jimpinder

*
  •  1,232 1,232
  • Wakefield, West Yorks, UK
    • View Profile
Re: trouble with steps or with stream pulses
« Reply #9 on: August 31, 2007, 02:33:17 PM »
If my maths is correct, you are loosing 2 pulses per inch. It might be the first two pulses, whilst your card is sorting out direction. Try adding 2 to the 4000 - make it 4002 - see if that makes any difference,

Jim
Not me driving the engine - I'm better looking.