Messages

51

Mach4 General Discussion / Re: Mach4 Printer Port Discussions

« on: August 08, 2014, 11:49:05 PM »

...
You might want to totally delete the folder to make sure no old file is corrupting things...an reinstall. Might be worth the effort..

...

Art

I deleted my \Mach4Hobby and re-installed everything from scratch - no reuse of anything.  No change in behavior.


-Freeman

52

Mach4 General Discussion / Re: Mach4 Printer Port Discussions

« on: August 08, 2014, 10:44:21 PM »

Freeman:

 Just in case something didnt update, what are the two version numbers reporte din Darwins startup screen in config..

Art

The versions listed on the Darwin startup screen are 1.09 and 1.08.  Just to be through  I deleted and recopied the DLL and plugin before checking.  I still go the same behavior when jogging.


-Freeman

53

Mach4 General Discussion / Re: Mach4 Printer Port Discussions

« on: August 08, 2014, 09:11:08 PM »

Art,

I wish I could say that Mach 4 build 1936 and Darwin 1.11 solved everything, but I'm still getting the same behavior as I videoed last night.  While everything runs fine in Darwin, I'm seeing 100% CPU usage and stuttering when jogging or do a G00 in Mach 4. 


-Freeman

54

Mach4 General Discussion / Re: Mach4 Printer Port Discussions

« on: August 07, 2014, 11:15:57 PM »

Art,

I finally got some time to look at my issues with movement with Mach 4 build 1914 and Darwin 1.10.  I made a video of the behavior so you can see what I'm seeing.  The link for the video is:

https://www.youtube.com/watch?v=qAqcAZq-vPM&list=UU27_FcP3Gq_GKOvqGMof2tg


Going over the finer points... I'm able to set Darwin to any value above 17kHz as you suggested. In Darwin, once I have the port and the charge pump enabled, I can jog smoothly at full speed on X.  Y seizes at the top end of the rate (say above 80%), but runs smoothly up until then.  When I exit out of Darwin and try jogging in Mach 4 (running at 40kHz) the X and Z axis simply sputter and jerk at 28% jog rate.  The CPU in both cases shows Mach 4 bouncing around between 60% and 90%.  The problem doesn't look like it's CPU bound.  I also tried reducing my video interval in the screen set from 50ms to 40ms, with no improvement. I'm running with a very good nVidia GTX 550 TI with at least a 1Gb of RAM on it.


-Freeman

55

Mach4 General Discussion / Re: LUA Script - Please help me

« on: August 03, 2014, 09:57:16 AM »

I did this by manipulating the slider.

I did a GetProperty() on the slider, did the math to add/subtract 5%, and then did a SetProperty() to update it.

I will post the code when I get to my computer.


-Freeman

Sent from my SGH-I337 using Tapatalk

56

Mach4 General Discussion / Re: Mach4 Printer Port Discussions

« on: August 03, 2014, 01:41:04 AM »

Freeman:

   The pop's are the frequency going over the kernal speed.
Those messages are only generated by Darwin if the steps/sec times
the max velocity is greater than the kernal frequency.
  It means you need to lower the speed, or raise the kernal speed.
In the end, the maxvel in units/min must follow the formula

  maxVel / 60 * StepsPerUnit <= KernalFrequency

Art

I have been through the calculations multiple times.  Each time I come up with 17kHz for


    (400 steps * 25.4mm) = 10160 steps/inch               --- I have metric ballscrews

    (100 in/min / 60s) * 10160 steps/inch = 16933.33 Hz   --- this is for X & Y, Z I run at 75 in/min, which is 12700 Hz < 17000 Hz


X and Y axis run smoother now at 100% jog.  Most jogs back and forth on the X & Y axis run fine, but about every third or forth traversal of the axis the motor stalls hard with a high pitch noise.

The Z axis running with 75 in/min acceleration and 17kHz kernel speed is completely useless.  It can't more more than an half a inch before it stalls and squeals.

I noticed my Mach3 motor tuning has 5ms pulses, so I tried setting Darwin's pulse length to 5ms using the register diagnostic tool -- no change.

I'm not sure what I'm doing wrong.  This was all running fine before the last update.  I was using 17kHz and 100in/min which is what I used under Mach 3.


-Freeman

57

Mach4 General Discussion / Re: Mach4 Printer Port Discussions

« on: August 02, 2014, 01:30:27 PM »

Art,

I did a clean install of Mach 4 Build 1914 and Darwin2014-1-10.zip.  I now notice a new behavior that wasn't happening with the previous versions that ran with Mach 4 build 1900 and prior.  Using the same motor tuning settings, I'm now seeing an occasional "pop" in all three configured axis when I jog.  I tried re-tuning the motors, but the problem got worse if I went above or below the original calculation.  It's better at lower jogging speeds, but nearing 100% jog rate I start to see e-stops with messages saying that Darwin is too fast.

Any ideas?


-Freeman

58

Mach4 General Discussion / Re: Mach4 Printer Port Discussions

« on: August 01, 2014, 04:25:25 PM »

Art,

Just to be clear.  This version will only run with Mach 4 build 1914 and later? 


-Freeman

59

Mach4 General Discussion / Re: Re: Screen Set Ideas

« on: July 31, 2014, 12:40:47 AM »

#4012

Ooooo thanks Craig.

I'll have to update my screen set tomorrow. I just got done with a big push on a project at work and I need to bleed the stress off a bit before I try to make progress on my screen set. I don't want playing with my toys to be confused with work stress, takes all the fun out of it.

-Freeman

Sent from my Xoom using Tapatalk

60

Mach4 General Discussion / Re: Homing function - Is this unreasonable?

« on: July 29, 2014, 02:01:12 AM »

Actually Scott, I realized that loop was running in an even more aggressive loop of the PLC script.  No wonder it was locking up.  I modified my approach to remove the while loop and to only issue the gcode to move off the limit switches when there was no movement.  This should avoid stacking gcode calls to move off the limit switches.

I haven't tested this yet.

-Freeman

Code: [Select]

----
--  Test if specified motor is on limit switch (due to homing) and move
--  in appropriate direction to move off of limit switch if needed.
--
--  WARNING : At the moment I am assuming a one-to-one mapping between
--            motors and axis (ie. motor 0 -> axis 0, motor 1 -> axis 1, ...)
--
--  @param motor (number)   motor selector, 0-6
--
function MyStdMill.MoveMotorOffLimitSwitch (motor)
    local inst = mc.mcGetInstance()
    local directionToMoveOffLimit, isOnLimitSwitch;

    if (DSC.GetFunc('ISAXENBL', motor)) then

        if (DSC.IsPlusLim (motor)) then
            directionToMoveOffLimit = MyStdMill.DIR_NEGATIVE;
        elseif (DSC.IsMinusLim (motor)) then
            directionToMoveOffLimit = MyStdMill.DIR_POSITIVE;
        else
            return;
        end

        local axis = { [0]="X", "Y", "Z", "A" };
        local gcode = string.format("G90 G01 %s%d F5", axis[motor], directionToMoveOffLimit);

        mc.mcCntlSetLastError(inst, string.format("Moving motor %d off axis %s, direction = %d",
                                                  motor, axis[motor], directionToMoveOffLimit));

        if (DSC.GetFunc('ISAXSTILL', motor)) then
            mc.mcCntlGcodeExecute (inst, gcode); -- only move if other movement is in process
        end
    end
end