On Friday 22 February 2008, Ian W. Wright wrote:
>Hi,
>
>I've been looking at the 'nist-lathe' config files and trawling the web
>for more info on this but pretty much drawn a blank as far as the info I
>want is concerned. So, can anyone please tell me:
>
>1. What electronics were involved in hooking up the spindle motor to the
>PWM output of the parallel port.
I would, for the scale of the work you are doing, make the headstock motor a
big stepper, nema 23 at least, and timing belt driven with enough gear down
to get the torque required, which I suspect isn't much. 3 to 1 maybe? And I
wouldn't do pwm, I'd do step/dir into a xylotex driver.
>2. What kind of encoder is needed - does it need to be a full quadrature
>encoder or a simple 'disk-with-one-slit' and a photocell?
With the above setup, none, as the math within emc can handle the synch
required, even for a many pass single tooth cutter if all the gear ratios are
integer values. Probably even if they aren't since the fp format emc uses
internally is a quite long number representation.
>3.How accurately could this system cut a thread?
As accurately as you can measure it. The rest is just math in the gcode
>Let me explain more - I am looking for a way to cut threads which match
>those used in the antique timepieces I restore. Most of these from the
>17th and 18th centuries were cut by hand and were of no standardised
>pitch. The craftsman simply had a 'screw lathe' in which the 'headstock
>spindle' was screwcut (often filed by hand originally) for half its
>length and this ran in a 'nut' in the headstock. The work was held in
>the other end of this spindle and the thread was cut with a hand tool as
>the spindle rotated and wound itself through the headstock screw. So,
>whatever the diameter of the screw, the pitch was the same and the depth
>of thread was governed by the workman's eye'. As I said, most of these
>lathe threads were originally hand filed and a system of repetition was
>used to generate progressively better (more consistent) screws from the
>original until one was achieved that was considered good enough to fit
>to the 'screw lathe'. Consequently, the pitch, even if aimed for a
>particular number, was always approximate and 40 TPI might well come out
>at 41 2/3 TPI. For proper restoration work, it is essential that any new
>screws match the originals as closely as possible and so I'm looking for
>a way to do this using EMC2 and my little milling machine, with a lathe
>bed attachment.
One other effect I'd like to point out Ian, is cold flow, where the screw is
stretched a small amount in that area between the head and the threaded hole
its been installed in for 75 years. Because of that, I think I'd try to
measure not the screw, but the nut or threaded hole, and this I'd imagine
would be a tricky operation to do accurately also. Because of the size, I'd
expect to have to make the device doing the valley to valley measurement,
over more than one valley if the internal threads are deep enough.
>I can probably do it by using a stepper driven spindle
>in the lathe head but I'm wondering if it would be better to try to get
>a smoother drive using some kind of servo drive - if possible avoiding
>the expense of buying a commercial servo driver. I do have a servo motor
>with encoder attached that might be useful but it doesn't have an index
>pulse - maybe I could provide that separately on the lathe spindle?
With a 262 oz stepper, microstepped at 8x, that's 4k steps/rev, and if timing
belt ratio driven at 3x spindle, that's 12,000 steps per spindle rev. Smooth
would not be an issue, only speed, it might not be fast enough if you had to
make 100 screws like the first one. Spindle speeds would be in the hundred
range, as I've never seen my motors doing much over 250 rpm at the motor. I
can do an x axis move at 30 ipm with such a motor turning a 20 tpi screw, so
that's 600 rpm but its all tapped out & may stop at any time, <=200 is a much
more reasonable figure if its to stay locked to position while doing actual
work.
>If, as I suspect, the nist-lathe setup uses only a single pulse per rev
>for spindle location, would this be accurate enough to cut odd threads?
No, an index pulse only won't cut it except in runs long enough to set an
average. Screw making is often no more than 20 turns of the spindle &
usually less.
>Any advice will be gratefully received...thanks.
My $0.02. Not much, but that is how I would do it. And may yet. But the
lathe is going to be bigger than my present 7x10 too when I do. I'm only
half a days drive one way from a Grizzly showroom. :-)
--
Cheers, Gene
"There are four boxes to be used in defense of liberty:
soap, ballot, jury, and ammo. Please use in that order."
-Ed Howdershelt (Author)
I put instant coffee in a microwave and almost went back in time.
-- Steven Wright
-------------------------------------------------------------------------
This SF.net email is sponsored by: Microsoft
Defy all challenges. Microsoft(R) Visual Studio 2008.
http://clk.atdmt.com/MRT/go/vse0120000070mrt/direct/01/
_______________________________________________
Emc-users mailing list
[email protected]
https://lists.sourceforge.net/lists/listinfo/emc-users
