RE: [CAD_CAM_EDM_DRO] EMC Linux vs Windows
Posted by
Carol & Jerry Jankura
on 2002-03-09 22:07:36 UTC
Hi, Terry:
|With out giving a lot of background I have building a CNC mill for
|the past two years. It is step and direction and I have it running
|with CNC PRO. After getting it running I decided it would be more
|accurate with an encoder feed back system. I went on line and found
|out that I could do this with one of Dan Mauch DRO kits using either
|EMC or DeskNCrt to control it, I have the encoders setup the DRO kit
|is waiting for me to setup the computer with the control software.
Have you determined the exact cause of your inaccuracies? A properly
designed stepper motor system, running something like TurboCNC, should
exhibit only errors due to backlash and nonlinearity of any leadscrews. In
many instances, the effect of backlash can be compensated out of the system.
Many folks on this list are running their systems with Stepper motors and
achieve acceptable accuracy (precision, repeatability, and backlash
compensation) so I'd suggest that you determine a baseline (where you are
now) and investigate and eliminate the causes of your inaccuracies rather
than use a DRO to compensate for them.
What level of inaccuracy are you experiencing?
Now, when you ask about Windows, the only program that claims to run
properly under Win 98 is Master5, but I don't know if it has an encoder
feedback system or not. The problem that you have with Windows is that the
internal clock interrupt (19.2 msec rate) allows the OS to take over the
machine, resulting in some pulse rate inconsistencies that can allow the
motors to stall out. Note that, with a Windows system, you'll get all of the
pulses transferred to the motor, but the pulse rate may dither enough for
the motors to stall out much sooner than they otherwise would.
|With out giving a lot of background I have building a CNC mill for
|the past two years. It is step and direction and I have it running
|with CNC PRO. After getting it running I decided it would be more
|accurate with an encoder feed back system. I went on line and found
|out that I could do this with one of Dan Mauch DRO kits using either
|EMC or DeskNCrt to control it, I have the encoders setup the DRO kit
|is waiting for me to setup the computer with the control software.
Have you determined the exact cause of your inaccuracies? A properly
designed stepper motor system, running something like TurboCNC, should
exhibit only errors due to backlash and nonlinearity of any leadscrews. In
many instances, the effect of backlash can be compensated out of the system.
Many folks on this list are running their systems with Stepper motors and
achieve acceptable accuracy (precision, repeatability, and backlash
compensation) so I'd suggest that you determine a baseline (where you are
now) and investigate and eliminate the causes of your inaccuracies rather
than use a DRO to compensate for them.
What level of inaccuracy are you experiencing?
Now, when you ask about Windows, the only program that claims to run
properly under Win 98 is Master5, but I don't know if it has an encoder
feedback system or not. The problem that you have with Windows is that the
internal clock interrupt (19.2 msec rate) allows the OS to take over the
machine, resulting in some pulse rate inconsistencies that can allow the
motors to stall out. Note that, with a Windows system, you'll get all of the
pulses transferred to the motor, but the pulse rate may dither enough for
the motors to stall out much sooner than they otherwise would.
Discussion Thread
terry twombly
2002-03-09 20:37:58 UTC
EMC Linux vs Windows
Carol & Jerry Jankura
2002-03-09 22:07:36 UTC
RE: [CAD_CAM_EDM_DRO] EMC Linux vs Windows
Paul
2002-03-10 05:11:46 UTC
Re: [CAD_CAM_EDM_DRO] EMC Linux vs Windows
Dan Mauch
2002-03-10 06:27:28 UTC
RE: [CAD_CAM_EDM_DRO] EMC Linux vs Windows
Gail & Bryan Harries
2002-03-10 07:26:02 UTC
RE: [CAD_CAM_EDM_DRO] EMC Linux vs Windows