Re: 3D printer progress

Graham,
setting aside gantry movement problems, how does it actually works?
I mean how the object is being built up?
How does the printer prints?
Alex

--- In CAD_CAM_EDM_DRO@yahoogroups.com, "Graham Stabler" <grezmos@...>
wrote:

>
> --- In CAD_CAM_EDM_DRO@yahoogroups.com, "ballendo" <ballendo@> wrote:
>
> > Seems there's got to be a way to shift from dance and pre-print to
> > carriage travel... Cuz that's the only way we're gonna seamlessly
> > use existing printer routines...
>
> The way I intend to do it is to monitor the paper feed encoder, detect
> the high frequency when it actually feeds (much faster than dancing)
> and then take over control. Its not seamless but as the system is
> going to need something to control the supply/build bins anyway and
> will already need two stepper drivers why not have a third. It also
> means you can size the gantry motor appropriately for carrying the
> counter roller or whatever else.
>
> There is a way to mechanically do what we want but its only half the
> story, the printers themselves have a special mechanism that means
> that when the paper roller goes in the reverse with the print head in
> a certain position a clutch/ratchet engages and the paper pick up
> roller turns, you could in theory make some kind of clutch system that
> works on that principle.
>
> But even then you have to mess around to get the gantry to return, its
> actually easier to just take control.
>
> >copiers, so parts and replacements should be easy to get.
>
> I wondered about some form of electromagnetic clutch, it would make
> engagement easy but you still need to mess around to do the reversing
> for gantry return (assuming you don't want to sit there for an hour).
>
> > Trying to cover your back<G>,appreciate your efforts,
>
> Please keep on thinking, I'm bound to miss all kinds of options.
>
> > P.S. Doesn't the propeller seem custom made for multi-axis motion
> > control!?
>
> Yes, in fact I implemented Grex type pulse generation in about 10
> lines of code, each processor has two counters that can be used as
> programmable numerical oscillators, so each cog can generate two pulse
> trains well into the 10's of Mhz without doing anything, so you set a
> fixed length loop up to act like the Grex interrupt and modify the
> frequencies each time just like the Grex does.
>
> What I haven't done is the pre-motion calculation stuff, if you read
> the G-rex documentation and follow the example you will see that it
> doesn't work for just any old maximum velocity and acceleration
> because of the way it works, you need to tweak these so you end up
> traveling the correct distance.
>
> Graham
>