DoN. Nichols wrote:
Jon Elson wrote :
Just as it is the one advantage of a drum plotter which I have.
Resolution is 0.010", almost direct drive of stepper motors from the
computer, something like a 15 V input signal, even though it was low
current. It required a special wire-wrapped interface from my SWTP
6809.
I still have a Calcomp 1076C plotter, takes up to size E paper,
has servos that give 3 G accel and 50+ IPS motion. Mostly it
just tears paper and wrecks $30 ceramic pen nibs (probably
because I don't know how to set it up right.) I haven't turned
it on in 5+ years, but being high-quality gear it most likely
would fire right up if I did.
I did just get a color laser printer off the loading dock at
work, where people set cast-offs that somebody may want. It
almost worked, and with a little web searching, I found there's
a little damper pad that works out of position and fouls up the
pull-in time of a solenoid. Difficult to get to, but insanely
easy to fix once you have it apart. Prints incredibly good
color output.
The advantage of that roll paper feed was that I could directly
plot the fret positions for a stringed instrument.
A friend of mine did exactly the same on an ancient Calcomp at
the Wash Univ. data center some 30+ years ago.
[ ... ]
Maybe or maybe not. Solaris is supposed to be a real-time OS to
start with, and with a PCI version of the Servo-To-Go card it should be
possible to bring it up on any of the Ultra systems (which all have PCI
slots instead of Sbus.)
There IS no PCI version of the Servo-to-Go card. Strictly ISA,
and thus woefully obsolete.
So I see after visiting their web site. There is simply the
original (which I have) and the second version, which is still ISA.
Yes, can't imagine they sell a whole bunch of them anymore.
My hardware runs on an IEEE-1284
capable parallel port. Is the real time compatible with rtai?
RTAI? What is that?
Rtai is the current real time "extension" for Linux. It
actually runs above the Linux kernel, and makes Linux the lowest
priority task. The RT programs run as loadable kernel modules,
and run with kernel privileges, and also kernel restrictions.
The later Suns (including my SB-1000s)
have a fully programmable parallel port, once you find the right termio
signals to send to the driver.
The advantage of the IEEE-1284 (EPP) mode of the PC parallel
port is it does all the handshaking between the PC CPU and the
device(s) on the parallel port in hardware, so you can read or
write a register on the device with a single CPU instruction,
and do single-byte transfers at the rate of every 800 ns or so.
Do you know that the X-axis ballscrew does not turn?
Yes, and I've actually got a screw made like that from some
other machine, but it is obviously not a Bridgeport BOSS part.
It actually would work pretty well, and I may use it, but I will
have to machine all the attachment hardware to use it. I'm
still deciding whether that screw should go on the mill or the
lathe.
And do you have the Series-I head for the CNC version? The ball
screw for that is hollow -- surrounding the quill, with the ball nut
rotating around that, so the thrust is truly on axis.
No, I do not have this. I have a usable 1J head that I have
added a ballscrew assembly to the front of the housing, as close
to the quill as I can get it. It has a little flex in the
linkage, but the backlash is barely over .001", so I am mostly
satisfied with it. You can see my insane hack job of a retrofit
http://jelinux.pico-systems.com/CNCconv.html
Jon