Skipping steps, avoiding resonance
Posted by
Randy Gordon-Gilmore
on 2002-12-01 22:35:26 UTC
I have a perplexing problem. Recently I've put together a gantry milling
machine from mostly surplus components--a Techno-Isel stepper-driven gantry
table (500x500x100mm theoretical travel, with zero-backlash leadscrews),
three General Controls "Dragon Driver" 2.5A bipolar chopping drives (fed by
a 28V transformer) and a Sherline headstock, for which I made an adapter
plate to bolt it onto the Z-axis carriage.
I finally got the thing running this weekend, and it is not running
reliably. I have the drivers in half-step mode, which with the 4mm-pitch
leadscrews gives 2540 steps per inch. I can jog the X and Y axes at
insanely fast speeds (200 ipm,) but I have fast jog set at 60 ipm. I'm
using CNCPro running under DR-DOS 7.03.
Running 3-axis Gcode generated by Visual Mill Free, all 3 axes "drift"
during machining. I'm cutting acrylic with a .50 roughing endmill, with
cutting speed set at 10ipm. The motion is actually more like 2.5D--it goes
to a depth and then pockets that level before going to another depth and
repeating. It spirals down into the material.
Cuts at each depth should line up, but I end up with a "terraced" effect
because the X and Y will shift by .005-.050 inches between depths. I have
also measured the depths (each "level" is set at .050") and they are not
consistent either.
I'm baffled, but suspecting that the spiraling between levels is what is
throwing things off, along with maybe arcs in XY. When I got the drives,
they were set for full stepping, and the table lost steps very easily. I
think that the coarse leadscrews contribute to that, because the motors are
stepping fairly slowly. Now that I have the drives set for half stepping,
and have a decent acceleration programmed into CNCPro to get to running
speed quickly, and keep all my motions to 5 ipm or more, I seem to be
avoiding the low-speed resonances OK for straight moves.
But what can I do about arcs? By their nature they will be moving the axes
slowly through the suspected resonance range, and spiral leadins will do
that for all 3 axes. How do I avoid that?
I truthfully don't know for sure that the arcs are the problem. But
straight-line motion with handwritten Gcode works fine. I have the stepper
motor cables away from AC cords as much as possible. CNCPro works pretty
reliably on my Sherline mill, so I don't think that is the problem. I have
run the Gcode through editNC to backplot it, and it looks fine so I don't
suspect that. So that pretty much leaves arcs and spirals, doesn't it?
Newbie again,
Randy
Randy Gordon-Gilmore ,----.___________ ______________ _________________
ProtoTrains // = = === == || == == == = || == == == = == =|
Rio Vista, CA, USA /-O==O------------o==o------------o==o-----------o==o-'
zephyrus@... http://www.prototrains.com
machine from mostly surplus components--a Techno-Isel stepper-driven gantry
table (500x500x100mm theoretical travel, with zero-backlash leadscrews),
three General Controls "Dragon Driver" 2.5A bipolar chopping drives (fed by
a 28V transformer) and a Sherline headstock, for which I made an adapter
plate to bolt it onto the Z-axis carriage.
I finally got the thing running this weekend, and it is not running
reliably. I have the drivers in half-step mode, which with the 4mm-pitch
leadscrews gives 2540 steps per inch. I can jog the X and Y axes at
insanely fast speeds (200 ipm,) but I have fast jog set at 60 ipm. I'm
using CNCPro running under DR-DOS 7.03.
Running 3-axis Gcode generated by Visual Mill Free, all 3 axes "drift"
during machining. I'm cutting acrylic with a .50 roughing endmill, with
cutting speed set at 10ipm. The motion is actually more like 2.5D--it goes
to a depth and then pockets that level before going to another depth and
repeating. It spirals down into the material.
Cuts at each depth should line up, but I end up with a "terraced" effect
because the X and Y will shift by .005-.050 inches between depths. I have
also measured the depths (each "level" is set at .050") and they are not
consistent either.
I'm baffled, but suspecting that the spiraling between levels is what is
throwing things off, along with maybe arcs in XY. When I got the drives,
they were set for full stepping, and the table lost steps very easily. I
think that the coarse leadscrews contribute to that, because the motors are
stepping fairly slowly. Now that I have the drives set for half stepping,
and have a decent acceleration programmed into CNCPro to get to running
speed quickly, and keep all my motions to 5 ipm or more, I seem to be
avoiding the low-speed resonances OK for straight moves.
But what can I do about arcs? By their nature they will be moving the axes
slowly through the suspected resonance range, and spiral leadins will do
that for all 3 axes. How do I avoid that?
I truthfully don't know for sure that the arcs are the problem. But
straight-line motion with handwritten Gcode works fine. I have the stepper
motor cables away from AC cords as much as possible. CNCPro works pretty
reliably on my Sherline mill, so I don't think that is the problem. I have
run the Gcode through editNC to backplot it, and it looks fine so I don't
suspect that. So that pretty much leaves arcs and spirals, doesn't it?
Newbie again,
Randy
Randy Gordon-Gilmore ,----.___________ ______________ _________________
ProtoTrains // = = === == || == == == = || == == == = == =|
Rio Vista, CA, USA /-O==O------------o==o------------o==o-----------o==o-'
zephyrus@... http://www.prototrains.com
Discussion Thread
Randy Gordon-Gilmore
2002-12-01 22:35:26 UTC
Skipping steps, avoiding resonance
Tim Goldstein
2002-12-01 23:05:59 UTC
RE: [CAD_CAM_EDM_DRO] Skipping steps, avoiding resonance
Randy Gordon-Gilmore
2002-12-01 23:53:15 UTC
RE: Skipping steps, avoiding resonance
CL
2002-12-02 06:59:56 UTC
Re: [CAD_CAM_EDM_DRO] Skipping steps, avoiding resonance
Carol & Jerry Jankura
2002-12-02 08:06:49 UTC
RE: [CAD_CAM_EDM_DRO] Skipping steps, avoiding resonance
Randy Gordon-Gilmore
2002-12-07 23:34:24 UTC
Re: [CAD_CAM_EDM_DRO] Skipping steps, avoiding resonance
Randy Gordon-Gilmore
2002-12-07 23:54:02 UTC
Skipping steps, avoiding resonance and the importance of being grounded
Chris L
2002-12-08 07:52:11 UTC
Re: [CAD_CAM_EDM_DRO] Skipping steps, avoiding resonance and the importance of being grounded
Peter
2002-12-08 08:43:19 UTC
Re: [CAD_CAM_EDM_DRO] Skipping steps, avoiding resonance
Peter
2002-12-08 08:48:00 UTC
Re: [CAD_CAM_EDM_DRO] Skipping steps, avoiding resonance and the importance of being grounded
Chris L
2002-12-08 09:34:44 UTC
Re: [CAD_CAM_EDM_DRO] Skipping steps, avoiding resonance
Peter
2002-12-08 12:00:36 UTC
Re: [CAD_CAM_EDM_DRO] Skipping steps, avoiding resonance
gadget_c <billsand@c...
2002-12-08 12:29:52 UTC
Re: Skipping steps, avoiding resonance
mariss92705 <mariss92705@y...
2002-12-08 12:33:18 UTC
Re: Skipping steps, avoiding resonance
Chris L
2002-12-08 16:22:36 UTC
Re: [CAD_CAM_EDM_DRO] Re: Skipping steps, avoiding resonance
Chris L
2002-12-08 16:37:20 UTC
Re: [CAD_CAM_EDM_DRO] Skipping steps, avoiding resonance
Peter
2002-12-08 18:57:35 UTC
Re: [CAD_CAM_EDM_DRO] Skipping steps, avoiding resonance
Peter
2002-12-08 19:02:38 UTC
Re: [CAD_CAM_EDM_DRO] Skipping steps, avoiding resonance