[Arm-netbook] screwed up the Riki200 plotter design

[Luke Kenneth Casson Leighton]

On Sat, Jul 29, 2017 at 3:22 PM, Neil Jansen <njansen1 at gmail.com> wrote:

> OK, forgive me for going slightly off-topic here, but this whole time, I've
> been sitting here wondering if you're batshit crazy, or if your 3D printer
> design will work as well as you say it will work.  That's mainly because I
> do think it's crazy to be designing a machine in 2017 when there are so
> many existing designs out there that work.

 *sigh* i know it seems that way but i am not kidding when i say that
literally every single 3D printer i've seen has some form of design
flaw or is made from materials and parts that are just... too far-out
expensive, given that the requirements are to maximise mm / sec / $.

 maximising mm / sec / $ has some very _very_ specific implications,
that you can either minimise $ or you can maximise mm/sec (or both).
unfortunately many of the "minimal $ 3D printers" are so shit that to
try to increase mm / sec results in f*****g worse-than-useless
quality.  aaand unfortunately, many of the high-end 3D printers where
$ is high, the amount of $ that went into their construction is
completely insufficient to compensate for their $, such that the "mm /
sec / $" metric is *WORSE* than that of a cheaper crappy 3D printer!

thus i am therefore looking for something *in between* those two
extremes... and i absolutely kid you not, there is *literally* not a
single one.

 the ultimaker-2: absolute stellar mechanical design, i give it a 100%
score for mechanical design rigidity... except it's $2,000 and uses a
bowden tube.  if you know about bowden tubes (latency/play requiring
4-5mm of retract to compensate!) or have ever had nightmare problems
with bowden tubes, you'll know what i'm referring to.

 prusa i3 and prusa i3 clones: vertical carriages (almost 100% of
them), which are just too shit to even remotely consider.  vertical
carriages result in the entire head mechanism twisting and bending the
rods and the X-ends so badly that the head can actually bounce around,
causing both Z and Y variation.

 the list just goes on and on, eliminating every single pre-existing
design out there.


>  I remember your arguments from
> before as to why, so my purpose of writing this isn't to bring the
> motivation back up.

 oh, ok :)

> Originally I didn't really have a leg in the fight,
> but as I've been looking at the CAD screenshots and build pictures from
> your recent work, it is actually starting to take the shape of something
> that I had started to design about a year or so back, for the purposes of
> SMT pick and place.  Looking at the pics at
> http://forums.reprap.org/file.php?177,file=96492 and
> http://forums.reprap.org/file.php?177,file=96491 it looks alarmingly
> similar to what I had originally intended to design.

 very funny :)  well, it's GPLv3+ licensed parametric (python) source
code, so you're more than welcome.

> The only differences
> are what gets mounted on the head, and I probably would have went with
> Hiwin rails instead of pulleys.

 there's MGN9C Hiwin rails underneath the X and Y ends.  two MGN9C
blocks.  actually MGN9 was a mistake: i should have made it MGN12 -
the MGN9s are actually only about 6.5 to 7mm wide where the 2020 slot
is 5.5mm with a 0.5mm curve on each edge.... damn things keep
*dropping into the slot* and end up at an angle, pretty much no matter
what i do.

 this is a pain in the ass, to say the least.  but... *sigh*.... it's
what i got now.


>  There's enough room on that head for
> cameras, a Juki nozzle or two, and a height touch-off probe.

 ... just.  if you don't need fans.  which you shouldn't.  there's
about... 50x50mm of clear space without hitting any bearings or rods,
and there's also space in between the bearings @ 25mm x 25mm to give a
maltese-cross style amount of space to put "stuff".

> Differences in 3D printing philosophy aside, I think that this design could
> possibly be useful for an SMT pick and place frame.

 yehyeh it could.

> At this point I'm just going to sit here and patiently wait with popcorn in
> hand and wait until the whole thing plays out.  I'm not in any real rush so
> it wouldn't even be in 2017 before I would start building one.

 well give me a shout as i do want to make one at some point

> Anyway, good luck with the whole thing.  I still think you're batshit
> crazy, but in some evil genius way, if you can pull it off.

 :)