[Arm-netbook] Existential 3D Printing Moments

[Neil Jansen]

On Thu, May 18, 2017 at 11:29 PM, Luke Kenneth Casson Leighton <
lkcl at lkcl.net> wrote:

>  size, power budget.  in about 5-8 years it won't be an issue.
> EOMA200 is better suited to clustering.  bigger PCB size and a much
> higher power budget.

I'm not really concerned about size.  What do you mean by power budget?
What's to stop me or anyone else from buying, say, 8x EOMA68's and
networking them?  (gigabit or otherwise)?  Assuming that a backplane is the
only thing in the way, that could happen pretty easily.  The only other
thing missing at that point would be availability of the 1st gen cards, but
that should get better in time, I'd think.  Cost wise it's still less than
an Intel NUC, if you're not counting the cost of the backplane.


>  and now you can use a 24v heater you can spend another extra $5 on an
> E3Dv6 volcano clone, now you can get *another* 20% increase in speed
> for only a 2.5% increase in budget.

As you can see from the pics, we ran on the cheapest 12V power supplies
that we could find.  Before that I tested 24V, it wasn't worth the cost.
Again, brickwall economics here.  We went cheap.  The 12V power supplies
were purchased in bulk and were maybe $14 USD each?  The 3D printed mounts
and the little PCB's were practically free and it would turn the supply on
and off between jobs whereas our 24V bricks were on all the time.  The ONLY
thing that we splurged on at the time was the E3D nozzles and that was more
of a crapshoot.  I would have done better to cheap out on those as well, I
could have printed more reliably with the cheaper J-Heads.

>  yehyeh.  i was considering optimising the design for minimising
> extrusion, and rigidising the frame with panels (like the ultimaker),
> but if it's $1.50 for a 350mm length why bother with minimising
> extrusion, just get some strips of lexan 50mm wide and strap them in a
> cross across the diagonals.  i've got a little video showing how
> effective that is: https://www.youtube.com/watch?v=Qb-WjZY5qyI

Don't bother minimizing extrusion if you do end up redesigning (gah!).
It's cheap as dirt nowadays if you're buying the generic stuff.  If you
want rigid, well there you go.


>  it may surprise you to know that i'm a big fan of the melzi 2.0....
> except they're 1.3A driver ICs and that's not going to be enough.  i'm
> already getting regular extruder-skipping on the anycubic, and that's
> even at 50mm/sec.

I have a junk box full of Melzi's, they were horrible, but it was all
manufacturing defects from a crappy Chinese company.  The Chinese version
took some artistic leeway that the original (British IIRC?) designer
probably never intended.


> i can't bring myself to use RAMPS.  i just... can't.  phil hands knows
> the story very well, he had one of the original sanguilinos (huxleys)
> just like you: the polulus are *prototyping* boards *specifically*
> described by the manufacturer as for *evaluation purposes only*.  his
> printer used to stall out for a couple of seconds on one axis as the
> ICs overheated.  he read the datasheet and found that there's a
> *ceramic insulator* on the *top* of the IC (meaning that a heatsink is
> pretty useless), and that the IC is designed for heat to be wicked
> away *through the PCB* via the exposed pad.  and  a frickin 18x18mm
> PCB just ain't gonna cut it.

I've used both as I've said.  Mine never stalled out.  I used cheap-as-dirt
A4998's.  Of course, I was running them < 100mm/sec and they were happy
there. I'm sounding like a broken record again, lol.  But, it was cheap but
it worked extremely well.

>  i just... i can't bring myself to spend backers' money on stuff that
> i know is crud, neil.

You're starting to sound like a German engineer now :)  They're not crud if
you use them within the constraints that I outlined.  No need to turn your
nose at them.  What I'm trying to get at is that you've got this huge point
of diminishing returns, you can place yourself on either side of it.  Your
choice obviously.  But don't act like it isn't viable, because I literally
did it.


>  sso i've been spending some time tracking down board designs and so
> on.  Arduino Due: https://world.taobao.com/item/539393961702.htm RMB
> 75 so that's around $12.

Dang those Due's are getting cheaper, back in my day those were a pretty
penny.

>  and TRAMS uses TMC2100s, where their Reference Design has full PCB
> and schematics available: if i'm doing 10+ i can just send that to
> mike and he can make them.  TRAMS is *real* basic.  4 steppers, 2
> beefy power MOSFETs (extruder, printbed), 2 smaller ones for fans.

<3 TMC2100's.  Our PnP was going to use TMC2130's. Great German drivers.
However #1 they're hard as shit to import into China, which sucked for us
at the time.  You can get damn near anything in China but this was one of
those parts that just isn't really something that they use.  It was, to
this day, the only part that I could not find on Taobao.  We may have
smuggled our samples in from Hong Kong.  #2 they're only really necessary
if you want to squeeze performance out of your stepper motors.  For our
farm we never did that, we didn't need to.


>  MGN9C rails so that the problems associated with rods go away.
> triple lead screws (i might consider quadruple) on the printbed, NO
> CANTILEVERING.

You're a madman.  You sure like to over-engineer things, don't you? :)


>  well, here's the thing: i actually quite like trying out things that
> other people aren't doing.  but also taking calculated risks.

Sounds like you've already got your mind made up.  I'm not here to tell you
what to do.  I'm just sharing my experience and what worked for me.  Like
many technical problems, it's all about the approach.  There are as many
different approaches as there are engineers and business men.  You know
what is ultimately best for your situation.  If it were me in your shoes
though .. well, I'd never put myself in that position again, haha.  Nope,
one and done, thank you very much.  Any of my future products I make will
be CNC machined, laser cut, or injection molded, and then outsourced.  As
long as it's a durable product, it's not really any worse than the energy
expended to setup a printer farm.


...annd from your previous-previous email, I forgot to reply to this little
bit:

>   love it.  well let's get you on the list for a pre-production prototype
ok?

Yea, hook a brother up.  The pre-production is the A20 or is it the older
one?  Are there any basic breakout boards or dev boards for it to plug
into? If you need an address or anything like that just let me know.
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.phcomp.co.uk/pipermail/arm-netbook/attachments/20170519/2ab27eea/attachment-0001.html>