You can minimize the mass that the plotter has to transport by using one of
the DREMEL tools that has a flexible cable, then support the motor
independently of the plotter. I built a mechanism of that sort back in the
early '80's, and it failed miserably to do what I wanted. Those plotter
servos are fast at the expense of having the muscle to accomplish work.
An X-Y table built from some drill rod on pillow blocks with linear bearings
of which there are many relatively inexpensive types these days (check the
robotics manufacturint technology mags) is not terribly difficult to build
if you can get access to a machine shop with the basic equipment, i.e.
lathe, vertical milling machine, surface grinder.
When selecting materials, it's important to remember that steel has a
typical thermals expansion coefficient of 6 ppm while aluminum has one on
the order of 27 ppm/degree F. If your dimensions are referenced to a steel
base, having an aluminum carriage for the "Y" feed that's pillowed on
aluminum may cause some grief.
The way I built mine, intended for a PCB tester of sorts, and by someone
(me) who didn't know nearly enough about the job, I had to haul the motor
for the "Y" feed everywhere I went.with the carriage that provided the "X"
position. BTW ... a 5 TPI thread on a ball lead-screw will produce 1 mil of
translation when propelled by a 1.8-degree stepper. Likewise, the
pitch-diameter of a cable/pulley arrangement for hauling the carriages about
needs be about 0.3183" (not easy to hold within manufacturing tolerance by
an amateur, but terribly critical if you want <1.5 mil error, the industry
standard having then been 3 mils) if you want 1" per revolution. However,
1.8 degrees will give you nominally 1/200". Half-stepping a 0.9-degree
motor is a mite coarse. Ball lead screws are slightly heavier, but require
less torque to propel the carriage. Cables worsen the potential for
resonance problems, lead screws cost a lot more. It's a tough problem.
The worst part is that the bearing in the handpiece on the flexible-shaft
dremel can't take the side-load required for routing a hole in a PCB, though
it will last a little while if all you're routing is the foil. One other
thing to contemplate, by the way, is the difference between drag-cut and a
climb-cut when milling/routing, since the top speed of the DREMEL tool is
still on the low side for milling and drilling PCB's.
There ... thats a preliminary core dump of what I encountered back twenty or
so years ago. I hope I gathered what was wanted, as I couldn't find the
original post.
Dick
----- Original Message -----
From: "Gary M. Wheeler" <g.wheeler_at_worldnet.att.net>
To: <classiccmp_at_classiccmp.org>
Sent: Tuesday, December 19, 2000 1:25 AM
Subject: Re: HP plotters
> re: From: Philip.Belben_at_powertech.co.uk
> Date: 09/03/99-07:23:55 AM Z
>
> >I like the idea of a router in the pen part of a plotter - but it's a lot
> of
> >mass for the pen carriage to move around!
> >
> >Philip.
>
> Am working on a router table based on the HP 7225A single pen 8 1/2"x11"
> flatbed plotter. To make architectural models out of foam etc.
>
> Built an aluminum gantry-crane type arrangement that supports a Dremel
tool.
> Tried to keep mass and friction low as possible. Connected it to the
plotter
> with a pantograph arrangement.
>
> The plotter will drive the dremel tool well at slow speed. HOWEVER, when
> the plotter gets the message "PU" pen up, it RACES to the next x,y
> coordinate. This quick movement throws calibration and accuracy off.
>
> The other problem is printing from a Windows CAD program to an old plotter
> like this.
>
> Gary
>
> Gary M. Wheeler, AIA
> home page: http://g.wheeler.home.att.net
>
>
>
>
Received on Tue Dec 19 2000 - 11:08:46 GMT