Re: Stereolithograpy revisited

While I'd love to be able to have my own STL printer, I have resigned
to the fact that I have to exist with my mill and our ZCorp (3D
starch-type) printer. They're both fun, and we do outsource our STL
requirements when needed. Building my own starch-based printer is on
my project list....

I've only watched the STL process for a total of about 10 hours and
while I think it could be done at the hobby level, it would be tough.
The laser engine is not a trivial component - even though the diode
versions are more stable today, when I was in the industry they were
horribly cursed with drift. We used a number of argon and mixed-gas
units for the entertainment display industry, some with argon
(green-blue) line optics, some broadband ("whitelight") and some UV
specific. The larger units (20+ watts of visible or near-UV) were
quite hefty "Bazookas" that required high-flow water cooling and 480V
3 phase to live. And they were heavy. Did I mention they were heavy?
These things weren't going onto a gantry unless it was the size of an
assembly line crane. As mentioned in a prior post, almost all gas
lasers insisted on being mounted horizontally. Diode units had a
similar requirement, mostly due to the cooling channels. The only
units I saw that would have been viable, would have been the
SpectrPhysics Millenia series, unfortunately there wasn't a UV
verision (their new Explorer series is tiny and can put out a whopping
120uJ at 350nm....) but again, a tough mount for direct beam pointing.
As a matter of practice, legality and design, we placed our lasers
inside specially designed projection tables, hosting a multitude of
switchable mirrors on both fast and slow galvos, that would provide
all of our direct beam, fiber and remote effects control. This was a
niche industry - a lot of these were "production-prototypes" - only 10
or so units made a year, if that. The benefit is that the state of the
art was highly advanced by a couple of companies, especially in
graphics projection.
While I would be amiss to not mention the pioneers like LaserMedia and
NML, one major player was (and still is) Pangolin, that created both a
hardware and software platform (of course proprietary) for high speed
graphics projection - their ISA or PCI cards control the galvo
amplifiers and if required, AO crystal (for beam switching and color
tuning) and you're good to go. We could have 20-30 watts of full color
visible pumping onto 3x5mm mirrors and a scan angle of 40 degrees -
without breaking a sweat. The biggest problem with this type of setup
was divergence. Yes, the laser was mounted conveniently. Yes, the
output was conveniently mounted and controlled. Unfortunately, each
time you hit even a "perfect surface" mirror, the beam will diverge,
and not always consistently. Collimating optics are a double-edged
sword - and that's when you can see the beam!
The software and control platform can be simple and straightforward -
there are even a couple of older "open-source" XY only (no beam
control - we called it "blanking") that are essentially little DA
boards that you drop off of your printer port. For $500 or less a pair
of galvos are yours, cheaper if used.
One issue is with repeatability. Drawing quickly over the same path
was preferred by the galvos we used (as well as it was to draw
grpahics, anyway) but if you slowed the scan rate down the beam could
wobble with the additional heat produced by the galvos having to
maintain higher currents in holding "still" longer.
I no longer work in that field, but still in optics, and a lot with
blue and UV LEDs - I've not yet seen any stock unit that would work
well in this application - even the 3W Lux "star" knockoffs would
require a LOT of help to remain stable.
The number one reason we don't have a STL printer in house, aside from
the initial investment, is the recurring materials costs - our
building is full of merc and fluorescent lights - slow emitters of UV.
A horror story I heard was that a new operator had "left the lights
on, and the tubs open over the weekend" - at a couple hundred $$$ a
gallon, that's a lot of material to waste......
But if you need some control system help, I'll be happy to assist....

Cheers,

Ted.


--- In CAD_CAM_EDM_DRO@yahoogroups.com, "turyga1963" <TIntruder@...>
wrote:

>
> Grahm
>
> Drilling holes requires hitting the intended spot as does cutting
> out a shape or hitting the point of the uv-hardening-adhesive you
> intend to hit. If you set up a machine and press the green button,
> you sort of expect the thing to start cutting/drilling/marking or
> whatever at the correct point, huh?
>
> Most commercial systems from Lumonics, Baasel(now RofinSinar I
> think), and Hitachi use granite instead of steel and those that
> offer the option of a steel frame often require a pier isolated from
> the building foundation. This is what it takes to get repeatable,
> absolute indexed positioning of .0005" or better in 3 axes.
> Not just microvias, but controlled depth skiving, resistor trimming,
> micro-marking, or any of dozens of processes done daily in the high
> tech industries.
> Sure this differs from the processes run where lasers are machine
> shop evolutions of mills, routers, oxygen torches and other classic
> machine shop tools.
> And also obviously if you wish to use sheer mass in place of
> isolated/stabalized platforms, a big CNC laser for cutting steel
> plate can just use adequately large beams/castings or whatever for
> the frame.
> In any case, I highly doubt many homebrew CNC Laser builders will
> have a good experience with galvos, if they get them working at all.
>
> Obviously there are systems made for shop floor use, but just like
> normal "machines", temperature affects accuracy. Even new
> commercial CNC machines with temperature compensation features are
> only certified for a fairly narrow environmental range. I wonder
> what a 10 degree change will do to the aimpoint of your unprotected
> galvos? OOPS...there goes .001", maybe even .005" or more!!! Guess
> we're back to conventional X-Y-Z axis motion for most of us.
>
> Another issue with galvos vs. moving table or gantry is that when a
> shape is cut out, the edges have a bevel related to the beam angle
> of incidence galvo-to-table. If you have a cut line that
> transitions from the right edge of one window to the left edge of
> another, this will cause a visible reversal of the bevel direction.
> And any tiny error in Z Axis focus will cause an X-Y misalignment at
> the transition between windows on the table. Laser marking has less
> problem with this as the beam/mode is typically slightly de-focused
> anyway (mode shaping or intermediate optics), and there is no z Axis
> dimension...just a surface effect.
>
> Curious though, how much wattage at which wavelengths have you been
> able to pump through galvos? I know this thread was mainly about
> stereolithography, but it seems that most of the interest in hobby
> CNC lasers is more along the lines of the mill/router design, and I
> simply can't find available components to build a galvo system that
> will survive being used in a laser system where the power level is
> such that you can cut the materials people seem most interested in,
> such as common metals or thicker organics. All I seem to find are
> moving gantry or moving table.
>
> Also, what software are people using for motion control with their
> homebrew galvo systems? Sure, the X-Y-Z of the table is easy, but
> what about the galvo control? And what are you using to post your G-
> code? Not a trivial task to write the converters needed...
>
>
>
>
>
> --- In CAD_CAM_EDM_DRO@yahoogroups.com, "Graham Stabler" <eexgs@>
> wrote:
> >
> > --- In CAD_CAM_EDM_DRO@yahoogroups.com, "turyga1963" <TIntruder@>
> > wrote:
> > >
> > > In order to drill a hole in the right place, it has to sync
> > > perfectly with the correct laser pulse and be absolutely steady
> when
> > > it does.
> >
> > no one mentioned drilling holes.
> >
> > > If you plan on using galvos effectively, you also better plan on
> > > buying a really big and heavy granite optical bench to mount
> > > everything on and provide rigidity for the entire system, galvo,
> > > motion, and beam path.
> >
> > Most commercial laser cutting systems (see synrad.com etc) using
> > galvos have nothing like that.
> >
> > > Hmmmm...better budget for environmental control system that
> keeps
> > > the temperature +/- 2 degrees too.
> >
> > Laser marking systems are commonly used on the shop floor.
> >
> > I agree with your overall point but as commercial STL machines are
> not
> > operated in refrigerated rooms and do not have granite bases I am
> not
> > sure where you are getting this from. Perhaps micro via drilling
> needs
> > this, STL I doubt.
> >
> > Graham
> >
> >
> >
> >
> >
> >
> > >
> > >
> > > --- In CAD_CAM_EDM_DRO@yahoogroups.com, Jon Elson <elson@> wrote:
> > > >
> > > > Graham Stabler wrote:
> > > >
> > > > >--- In CAD_CAM_EDM_DRO@yahoogroups.com, juan gelt <juangelt@>
> > > wrote:
> > > > >
> > > > >
> > > > >>has everybody had a chance to expound on his opinions?
> > > > >>very well, now here are some facts:
> > > >
> > >
> >>http://www.technology.niagarac.on.ca/people/mcsele/lasers/index.htm
> > > l
> > > > >>
> > > > >>
> > > > >>
> > > > >>
> > > > >
> > > > >Firstly your assumption is that just because you can make a
> UV
> > > laser
> > > > >it must be suitable for stereo lithography but I doubt you
> have
> > > any
> > > > >idea if that is the case or not. I don't.
> > > > >
> > > > >
> > > > >
> > > > Yeah, the Nitrogen transmission line laser is usually run at
> 1 -
> > > 10 Hz
> > > > rep rate.
> > > > It would be pretty hard to do any real assembly with only 10
> > > pulses per
> > > > second.
> > > >
> > > > >What I should of said was we can't afford galvos anyway, my
> real
> > > point
> > > > >being there is not much point worrying about issues
> pertaining to
> > > > >technologies that are out of our reach (practically). I was
> > > refering
> > > > >to the post directly before mine only.
> > > > >
> > > > >
> > > > >
> > > > Actually, galvanometer-mirror drivers can be found on the
> surplus
> > > market
> > > > (they
> > > > used to use them in direct-writing oscillographs 50 years ago)
> and
> > > they
> > > > wouldn't
> > > > be that hard to make, either. I have some little ones,
> basically
> > > two
> > > > wires suspended
> > > > in a magnetic field, with a tiny mirror glued in the middle of
> the
> > > wires.
> > > >
> > > > Jon
> > > >
> > >
> >
>