On Sun, 03 Mar 2013 16:35:04 -0500, you wrote:
>No, but it isn't that expensive, either. Another (free) alternative,
>as has been mentioned before is Linux and linuxcnc. You would have to
>become familiar with Linux if you haven't used it (I haven't), but it
>should be pretty easy if you are only using it for one program.
>
>I just ordered a 3 axis controller on ebay ($42 USD including
>shipping), though I think I could use the parts from the on-board
>controller to make my own controller, especially if I use the parallel port.
As a note culled from the CAD/CAM group. Please be aware that the
chip/chipset used in the boards can be a limiting factor. Some chips
cannot be pushed all that close to their limits. Almost all (if not
every one) of the chips/chipsets do NOT tolerate removing the load
(stepper) from the chip with power on. It generally blows the output
in the chip (requiring either replacement of the entire chip or the
driver for that channel, if separate).
You need to be aware that connectors used to run the board to the
stepper need to be as robust (read: lock in place) as possible, and
the wires need to be reasonably flexible (wiring from the connection
to the stepper will flex, so you don't want that to break).
That said, I think, (without any kind of data to the contrary), that
these boards will work well within their limits (NO overvoltage, NO
overcurrent, NO disconnects). The only boards/drivers that I know
about that are as rugged as possible are made by gecko, and are the V
series (vampire.... Mariss has a sense of humor). Even then, the
warranty is outstanding.
What I use (because I'm thoroughly cheap and had the chips) is an
L297/L298 chipset with jumper settings for parameters and optical
isolators for step and direction. I haven't blown any of them, ∗but∗
I have been very careful with what I've been doing. Roughly 35 volts
and about 2.5 amps maximum, no behavior compensation, so not the best
in the design.
I'm doing a parallel port if I mess with a controlling PC, a custom
program (to be written) using an XMEGA otherwise, and for a full blown
solution, an FPGA hardware pulse generator (it works, but I need to
see how good it is, implemented in the simulator).
Harvey
>
>Mark
>