Hello Guest it is July 12, 2020, 12:18:07 AM

Author Topic: Mach 3 - driving beyond soft limits ?  (Read 1282 times)

0 Members and 1 Guest are viewing this topic.

Offline Filou

*
  •  44 44
    • View Profile
Mach 3 - driving beyond soft limits ?
« on: April 07, 2018, 03:25:35 AM »
Dear Forum Readers,

Has anyone else experienced something similar ???

I have been using Mach 3 for about a year for prototyping work on a 1 meter x 1.5 meter 3 axis router system, without major problems.

Two recent events are suspicious and potentially dangerous; Mach 3 passed one of its soft limits during machining and the axis motion was stopped by the hardware limit switch in place. These two scary events were started in a regular manner, no warning of Mach 3 telling me that the job goes beyond a soft limit, for example.   

On the second such event I could verify that the machine coordinate system of two axis had shifted by some amount, 170 mm for one axis, some 23 mm for the second.

I extracted these two numbers by jogging the "crashed" axis off of the limit and gently jogging to each of the home switches until they trip. The values indicated by the DROs in this situation should be identical than the values after a homing procedure, zero in my case.  A clear shift in machine coordinate was present. Scary and disappointing.

The configuration of this basic 3 axis open loop system:
X, stepper, no encoder, home switch on its own pin, limits in series on pin 15
A is slave of Y, steppers, no encoders,  Y home switch on its own pin, Y limits in series on pin 15
Z, stepper, no encoder, home switch on its own pin, Z limits in series on pin 15
in short: 3 pins for homing, one for limits, and one for e-stop.
The working volume defined by the soft limits is smaller and lies inside the volume defined by the hardware limits.

Any comment you may provide will be appreciated.

enjoy your WE, Philippe

Offline TPS

*
  •  1,608 1,608
    • View Profile
Re: Mach 3 - driving beyond soft limits ?
« Reply #1 on: April 07, 2018, 04:11:07 AM »
are you sure that softlimits had been turned on?
anything is possible, just try it.
if you find some mistakes, in my bad bavarian english,they are yours.

Offline Filou

*
  •  44 44
    • View Profile
Re: Mach 3 - driving beyond soft limits ?
« Reply #2 on: April 07, 2018, 05:54:04 AM »
TPS, thank you.

Yes, "softlimits" is always on for regular machining.

(Ansonsten waere ich ein bischen der Depp gewesen, alles erklaeren und genau das zu vergessen, oder nicht ?) ;D

Enjoy the nice spring day, Philippe

Offline Davek0974

*
  •  2,597 2,597
    • View Profile
    • DD Metal Products Ltd
Re: Mach 3 - driving beyond soft limits ?
« Reply #3 on: April 07, 2018, 08:44:31 AM »
As you explain that the machine coordinates had shifted it sounds very much to me like something had jammed the router and caused the machine to lose position - this would explain the soft limit failure.

I would check/clean/lubricate all motion and bearings etc then run some tests to see if it does it again.
Bridgeport Mill, Mach3 V062, CSMIO-IP/A controller, AC Servo Drives.
Plasma table, Mach3 V062, Step motors, C&CNC THC.

Offline Filou

*
  •  44 44
    • View Profile
Re: Mach 3 - driving beyond soft limits ?
« Reply #4 on: April 07, 2018, 09:05:27 AM »
Howdy Davek0974, thank you for this thought.

We are trying to reproduce the issue, one axis at the time, by disabling two out of three axis, no work piece.
Jamming does not seem a good candidate; when we disconnect the spindle (design on Y and X made on purpose for easy disconnection) they slide freely.   What we hope for now is something like "slipping".  If the stepper misses steps or swallows a few,  i.e. the axis does not move as much as expected and Mach3 will not know about it. This could also result in a shift. No smoking gun yet ! and the sun shines outside ... But these jobs must be ready by Monday ...

Philippe

Offline Davek0974

*
  •  2,597 2,597
    • View Profile
    • DD Metal Products Ltd
Re: Mach 3 - driving beyond soft limits ?
« Reply #5 on: April 07, 2018, 11:18:45 AM »
Maybe check your PSU voltage too, maybe it dropped while running and the motors just lost steps.?
Bridgeport Mill, Mach3 V062, CSMIO-IP/A controller, AC Servo Drives.
Plasma table, Mach3 V062, Step motors, C&CNC THC.
Re: Mach 3 - driving beyond soft limits ?
« Reply #6 on: April 07, 2018, 10:10:23 PM »
Hi,
a troubling problem to be sure.

If I follow the evidence you've posted it is as if the machine coordinates have shifted 170mm an 23mm from the physical location. I would not expect softlimits to be triggered in this
circumstance.

Imagine you have referenced your machine at the beginning of the session and at a given time through the job, while at machine coordinate  location 400,400 say, a fault ocurrs and the X
axis machine coordinate changes to 500. Mach believes the controlled point is at 500,400. If a move is called that returns the controlled point to machine zero, ie 0,0 the machine will
drive 500mm on the X axis when in fact its only 400mm from the X boundary by virtue of the previous fault and it will crash. Mach however believes its a legitimate move and within the boundaries
and will therefore not trigger a softlimit event. You are saved only by physical limit switches.

It is an unusual fault that Mach's machine coordinates should get out of whack with respect to its physical location unless it missing steps and missing 170mm should stick out like dogs balls!

Were you observing the machine while it was working?
Could it have suddenly stopped and only start again after the code had 'moved on' by 170mm?
Could the machine be losing steps incrementally but only show up when a move was called that moved out-of-bounds?

Craig
My wife left with my best friend...
     and I miss him!

Offline Filou

*
  •  44 44
    • View Profile
Re: Mach 3 - driving beyond soft limits ?
« Reply #7 on: April 08, 2018, 01:56:52 AM »
Good morning / evening Joeaverage alias Craig, thank you for your input.

"Were you observing the machine while it was working ?"

no, it was a long job, and I was doing something else. But we are now having a team (+ video recording) and hope to be able to observe this type of event.

"Could it have suddenly stopped and only start again after the code had 'moved on' by 170mm ?"

on this job, surfacing, there are X stops at the end of each travel, a dY to move the tool and a reverse command to the starting X point. The legs on X are always 700.0 mm long, there is no "170 long move" in the code anywhere. I feel 170 could have been anything, same for the 23 mm.  The only "known" stop is when the screen of the PC dims, there is a short interruption, but it is a clean one; machine and Mach 3 remain sync. This event is known and way too short to produce such gaps.

"Could the machine be losing steps incrementally but only show up when a move was called that moved out-of-bounds ? "
 
Loosing step is an obvious candidate, but no smoking gun yet.

Enjoy this nice day, Philippe

Offline Davek0974

*
  •  2,597 2,597
    • View Profile
    • DD Metal Products Ltd
Re: Mach 3 - driving beyond soft limits ?
« Reply #8 on: April 08, 2018, 02:46:29 AM »
Losing 170mm really sounds like a loose motor coupling - any possibility??
Bridgeport Mill, Mach3 V062, CSMIO-IP/A controller, AC Servo Drives.
Plasma table, Mach3 V062, Step motors, C&CNC THC.
Re: Mach 3 - driving beyond soft limits ?
« Reply #9 on: April 08, 2018, 03:12:21 AM »
Hi,
establishing how this error accumulates is going to be the key to solving the problem.

There is three ways I can see it happening.
1) The machine is missing steps on an occasional basis and an error is building up over a period of minutes before a move is attempted which results in a crash
2) There is some fault which causes an axis stall but shortly, several seconds, after the stall the axis resumes but now is well behind the Gcode contolled point.
3) There is some fault in Mach which is causing some instantaneous jump in the machine coordinates.

This last possibility could be detected by a suitably written macro running repeatedly with the macro pump.
The macro pump runs every 40ms or so. If the machine coordinates between successive passes through the macro pump differ by more than the maximum allowed by virtue of max motor tuning speed you
would have your smoking gun.

Craig
My wife left with my best friend...
     and I miss him!