CAD CAM EDM DRO - Yahoo Group Archive

Re: [CAD_CAM_EDM_DRO] Re: EZ-USB

on 2001-05-24 16:08:06 UTC
> The USB bus, like anything on a PC, is subject to have
> some problems. The early compatibility and lockup problems
> seem to be easing as the USB interface matures and more
> OEM's get their act together. Many problems that have occurred
> with device interfaces (i.e. serial, parallel, or network) are
> often traced to the system software (Windoze especially) trying
> to run them. The USB interface provides a very structured device
> interface which should result in fewer system bugs in the long
> run.
>
> It would probably be a good idea to have any CNC device on
> a USB bus by itself. This would reduce the possibility of
> any problems related to hubs and multiple devices. Most
> computers now have two or more USB busses so this should
> be reasonable.

Seperation in USB busses is a relative thing. The computer may have
multiple jacks, but they all ultimately hang off the same controller
chip. Mess with that chip and you loose all the USB devices. I think
my preference would be to have the CNC controls be the only USB
devices in the system - that way under something like rtlinux, you
could leave the USB driver out of the main kernel, so that only the
realtime portion of the system could talk to the controller. Of
course things may get harder when all keyboards and mice are USB...

> CNC control run away is an issue with any system. The USB board
> that we are working on will have a "dead man" timer so that any
> loss of USB control would cause a controlled shutdown. Also, USB
> communications have built in CRC checking of the data content
> being transferred so corrupt data is not likely to cause problems.

Stepper systems have a unique advantage in that they don't move
without constant input. A servo or hardware pulse generator can lead
to a runaway. At a minimum, I think you need to be protected against
computer crashes and microcontroller crashes. A typical embedded
systems trick is a timer that will reset the processor if it is not
periodically set back to 0. Having it also trigger someting like an
Estop that cuts power to the drives would be a good idea, just
in case the micro can't recover.

Chris

--
Christopher C. Stratton, stratton@...
Instrument Maker, Horn Player & Engineer
22 Adrian Street, Somerville, MA 02143
http://www.mdc.net/~stratton
NEW PHONE NUMBER: (617) 628-1062 home, 253-2606 MIT

Discussion Thread

Chris Stratton 2001-05-24 09:52:22 UTC EZ-USB Dennis Cranston 2001-05-24 10:44:12 UTC RE: [CAD_CAM_EDM_DRO] EZ-USB stevesng@n... 2001-05-24 12:47:40 UTC Re: EZ-USB Ian Wright 2001-05-24 13:35:58 UTC Re: [CAD_CAM_EDM_DRO] Re: EZ-USB Jon Elson 2001-05-24 13:52:26 UTC Re: [CAD_CAM_EDM_DRO] Re: EZ-USB stevesng@n... 2001-05-24 14:43:42 UTC Re: EZ-USB Alan Marconett KM6VV 2001-05-24 15:55:28 UTC Re: [CAD_CAM_EDM_DRO] EZ-USB Alan Marconett KM6VV 2001-05-24 16:06:30 UTC Re: [CAD_CAM_EDM_DRO] Re: EZ-USB Alan Marconett KM6VV 2001-05-24 16:07:26 UTC Re: [CAD_CAM_EDM_DRO] Re: EZ-USB Chris Stratton 2001-05-24 16:08:06 UTC Re: [CAD_CAM_EDM_DRO] Re: EZ-USB stevesng@n... 2001-05-24 17:11:24 UTC Re: EZ-USB stevesng@n... 2001-05-24 17:21:06 UTC Re: EZ-USB Alan Marconett KM6VV 2001-05-24 18:24:38 UTC Re: EZ-USB stevesng@n... 2001-05-24 18:51:52 UTC Re: EZ-USB Chris Stratton 2001-05-24 19:09:04 UTC Re: [CAD_CAM_EDM_DRO] Re: EZ-USB Jon Elson 2001-05-24 20:32:58 UTC Re: [CAD_CAM_EDM_DRO] Re: EZ-USB Smoke 2001-05-28 10:02:00 UTC Re: [CAD_CAM_EDM_DRO] Re: EZ-USB Jon Elson 2001-05-28 15:00:08 UTC Re: [CAD_CAM_EDM_DRO] Re: EZ-USB Hugh Prescott 2001-05-28 16:34:00 UTC Re: [CAD_CAM_EDM_DRO] Re: EZ-USB Carol & Jerry Jankura 2001-05-28 17:45:19 UTC RE: [CAD_CAM_EDM_DRO] Re: EZ-USB Hugh Prescott 2001-05-28 19:59:42 UTC Re: [CAD_CAM_EDM_DRO] Re: EZ-USB