Hello again,
It was nice to see this latest update on the project:
"Just a brief update: Mike’s factory has assembled the 500 Micro Desktop PCBs, though the through-hole VGA connectors still need to be hand-soldered on. These are the simpler of the two boards, so all of the Micro Desktops are being done, whereas explained in a previous update, the more complex board, the EOMA68-A20, will only be done in a smaller test run of 100 at first."
https://www.crowdsupply.com/eoma68/micro-desktop/updates/500-micro-desktop-p...
It must be satisfying to see all of this start to come together.
On the topic of casings for the computer cards, I get the sense that there are no ready-made solutions for the outward-facing side of the cards, this despite most traditional PCMCIA/CardBus-profile cards exposing ports of different kinds (such as modems, network ports, like two cards I have from many years ago) and thus having similar requirements.
I suppose 3D printing or some other creative casing solution isn't an option. Injection moulding seems to be quite demanding, as the following from the Tomu campaign indicates:
"Right now a shop in China is etching the Tomu case shape into one ton of steel. Once that’s done they’ll manufacture a few units to verify the etching works. After this “T0” shot, they will harden the steel tool and start production."
https://www.crowdsupply.com/sutajio-kosagi/tomu/updates/design-done-producti...
To provide context, the Tomu case is a tiny piece of plastic of around one square centimetre and probably no more than 3mm in depth!
Paul
On Sat, Mar 23, 2019 at 3:56 PM Paul Boddie paul@boddie.org.uk wrote:
Hello again,
hi paul good to hear from you
It was nice to see this latest update on the project:
"Just a brief update: Mike’s factory has assembled the 500 Micro Desktop PCBs, though the through-hole VGA connectors still need to be hand-soldered on. These are the simpler of the two boards, so all of the Micro Desktops are being done, whereas explained in a previous update, the more complex board, the EOMA68-A20, will only be done in a smaller test run of 100 at first."
https://www.crowdsupply.com/eoma68/micro-desktop/updates/500-micro-desktop-p...
It must be satisfying to see all of this start to come together.
yeah - kinda weird as well, "um is that all it takes, you just have to um wait for a factory to um make them??"
On the topic of casings for the computer cards, I get the sense that there are no ready-made solutions for the outward-facing side of the cards, this despite most traditional PCMCIA/CardBus-profile cards exposing ports of different kinds (such as modems, network ports, like two cards I have from many years ago) and thus having similar requirements.
correct.
I suppose 3D printing or some other creative casing solution isn't an option.
not a snowball in hell's chance. resin is too brittle, 3D printing is far too inaccurate, and the plastic is extremely thin. 5 years ago the ones we had stamped out for prototypes (laser-cutting i believe) broke almost immediately.
Injection moulding seems to be quite demanding,
it is. USD $10k is a not unreasonable budget... for soft tooling that will produce decent pieces for the first 1,000 items, start to wear out for the second 1,000 and produce shit thereafter.
that's $10k to $20k to *fail* to get it right, as in $10k to $20k *per iteration*.
i do know someone in the UK who knows of a company that supplies Formula 1 teams. their software is designed, thanks to the extreme amount of work done for Formula 1, to *simulate* the injection mold process.
as a result, they *guarantee* that the molds that are CNC milled from what their software outputs will work first time, and that includes guarantees that bridges that the software recommends be put in place to stop warping (against pull AND push pressures as the plastic cools in the mold) will do the job.
my friend worked for a company that didn't use their software, to manufacture a housing for an articulated lorry mirror. it was curved (for aerodynamic reasons) so there was plenty of opportunity for warping and stress / pressure fractures.
GBP 80,000 to ***FAIL*** to get it right.
the company had to then spend another GBP 45,000 on a second iteration.
injection-molding is a bitch.
so... i've instructed mike to tell the factory staff to get out the cutters, and to snap off the front-facing part of the plastic.
at some point i may get made an extremely thin gluable / foldable metal faceplate. an extremely thin metal sticker, basically, with the holes for the MicroHDMI, MicroSD and USB-OTG.
l.
-----BEGIN PGP SIGNED MESSAGE----- Hash: SHA256
On Sat, 23 Mar 2019 16:56:25 +0000 Luke Kenneth Casson Leighton lkcl@lkcl.net wrote: <snip>
not a snowball in hell's chance. resin is too brittle, 3D printing is far too inaccurate, and the plastic is extremely thin. 5 years ago the ones we had stamped out for prototypes (laser-cutting i believe) broke almost immediately.
What material did you laser cut? I would have thought that laser cut aluminum would be perfect.
Thanks, David
On Tue, Mar 26, 2019 at 11:52 PM David Niklas doark@mail.com wrote:
-----BEGIN PGP SIGNED MESSAGE----- Hash: SHA256
On Sat, 23 Mar 2019 16:56:25 +0000 Luke Kenneth Casson Leighton lkcl@lkcl.net wrote:
<snip> > > not a snowball in hell's chance. resin is too brittle, 3D printing > is far too inaccurate, and the plastic is extremely thin. 5 years ago > the ones we had stamped out for prototypes (laser-cutting i believe) > broke almost immediately. >
What material did you laser cut? I would have thought that laser cut aluminum would be perfect.
ten prototypes of the mass-produced PCMCIA plastic surround from Litkconn had holes laser-cut to make space for the micro-sd, micro-hdmi and USB-OTG ports.
the result was plastic under 1mm deep that had only around 0.5mm height below e.g. the micro-sd card and it of course snapped immediately.
as a result we considered very thin metal sheet (thick foil in effect) that could be stamped (or laser cut) and was effectively a "metal sticker" that could go over the end and be bent round. probably even by hand.
l.
The metal sticker sounds good to me. Better than nothing or anything else i can do my self. i guess it will last a good while too :). More fancy options could be saved for future cards with bigger budget i guess.
Thanks for the update/documentation posts again btw. They’ve been my lessons in design, engineering and bringing a product to market. :D
Helped me to have ideas like what if old signs made from aluminium plastic composite sheets where used to make a lightweight portable subwoofer or a light weight, cheap shipping box.... shaving the grams off :) and cheap :)
On Mar 27, 2019, at 06:21, Luke Kenneth Casson Leighton lkcl@lkcl.net wrote:
On Tue, Mar 26, 2019 at 11:52 PM David Niklas doark@mail.com wrote:
On Sat, 23 Mar 2019 16:56:25 +0000 Luke Kenneth Casson Leighton lkcl@lkcl.net wrote:
<snip> > > not a snowball in hell's chance. resin is too brittle, 3D printing > is far too inaccurate, and the plastic is extremely thin. 5 years ago > the ones we had stamped out for prototypes (laser-cutting i believe) > broke almost immediately. >
What material did you laser cut? I would have thought that laser cut aluminum would be perfect.
ten prototypes of the mass-produced PCMCIA plastic surround from Litkconn had holes laser-cut to make space for the micro-sd, micro-hdmi and USB-OTG ports.
the result was plastic under 1mm deep that had only around 0.5mm height below e.g. the micro-sd card and it of course snapped immediately.
Sounds like a more flexible plastic is needed but that clearly impacts the manufacturing processes available and thus the cost.
as a result we considered very thin metal sheet (thick foil in effect) that could be stamped (or laser cut) and was effectively a "metal sticker" that could go over the end and be bent round. probably even by hand.
Did you try the foil? How well did it work?
On Wed, Mar 27, 2019 at 11:16 AM Richard Wilbur richard.wilbur@gmail.com wrote:
Sounds like a more flexible plastic is needed but that clearly impacts the manufacturing processes available and thus the cost.
the existing plastic comes as-is. making *any* changes would require a minimum of around USD $20k in design and injection-molding costs.
as a result we considered very thin metal sheet (thick foil in effect) that could be stamped (or laser cut) and was effectively a "metal sticker" that could go over the end and be bent round. probably even by hand.
Did you try the foil? How well did it work?
no - just a design idea. shelved for the first Cards as there's insufficient funds. can always send them out later.
l.
On Mar 27, 2019, at 11:59, Luke Kenneth Casson Leighton lkcl@lkcl.net wrote:
the existing plastic comes as-is. making *any* changes would require a minimum of around USD $20k in design and injection-molding costs.
Yes, that's prohibitive when you are on the wrong side of out of money!
as a result we considered very thin metal sheet (thick foil in effect) that could be stamped (or laser cut) and was effectively a "metal sticker" that could go over the end and be bent round. probably even by hand.
Did you try the foil? How well did it work?
no - just a design idea. shelved for the first Cards as there's insufficient funds. can always send them out later.
Any idea how much it would cost to test the idea?
arm-netbook@lists.phcomp.co.uk
-
Alexander Ross -
David Niklas -
Luke Kenneth Casson Leighton -
Paul Boddie -
Richard Wilbur