https://www.crowdsupply.com/eoma68/micro-desktop/updates/taiwan-micro-deskto...
ok so i'm almost ready to send the laptop pcb1 and pcb4 off for prototyping, this gets about 20% the way towards getting the laptop done, but also is planned to be the basis of a new housing (an all-in-one PC) where sharing the exact same main PCB as in the laptop will cut costs of production of both.
mark (van de borre): about the casework i've sent the DXF files off to a china prototyping company just so i can get something quickly. i don't expect them to be perfect first time, and to have do do another iteration. i'll want to do one last full systems test then the microdsktop v1.7 including casework can go into production. i haven't worked out if i have to do 2,000 of the corner pieces as 3D-printed or to get them injection-molded, at 2,000 pieces it *might* be worth doing as injection-molding.
btw if anyone would like to see the gerber files for the three latest pcbs, they're here:
http://hands.com/~lkcl/eoma/laptop_15in/pcb1/laptop_15in_pcb1_cam/ http://hands.com/~lkcl/eoma/laptop_15in/pcb4/laptop_15in_powerboard/ http://hands.com/~lkcl/eoma/microdesktop/eoma68_microdesktop_cam/
l.
--- crowd-funded eco-conscious hardware: https://www.crowdsupply.com/eoma68
Just gonna ask here coz I'm too lazy. How hard would it be to repurpose one of these cheap $200 macbook clone things with intel atoms to take an eoma68 card instead? I can already imagine the rockchip one in it :)
On 16 April 2017 10:14:42 GMT+03:00, Luke Kenneth Casson Leighton lkcl@lkcl.net wrote:
https://www.crowdsupply.com/eoma68/micro-desktop/updates/taiwan-micro-deskto...
ok so i'm almost ready to send the laptop pcb1 and pcb4 off for prototyping, this gets about 20% the way towards getting the laptop done, but also is planned to be the basis of a new housing (an all-in-one PC) where sharing the exact same main PCB as in the laptop will cut costs of production of both.
mark (van de borre): about the casework i've sent the DXF files off to a china prototyping company just so i can get something quickly. i don't expect them to be perfect first time, and to have do do another iteration. i'll want to do one last full systems test then the microdsktop v1.7 including casework can go into production. i haven't worked out if i have to do 2,000 of the corner pieces as 3D-printed or to get them injection-molded, at 2,000 pieces it *might* be worth doing as injection-molding.
btw if anyone would like to see the gerber files for the three latest pcbs, they're here:
http://hands.com/~lkcl/eoma/laptop_15in/pcb1/laptop_15in_pcb1_cam/ http://hands.com/~lkcl/eoma/laptop_15in/pcb4/laptop_15in_powerboard/ http://hands.com/~lkcl/eoma/microdesktop/eoma68_microdesktop_cam/
l.
crowd-funded eco-conscious hardware: https://www.crowdsupply.com/eoma68
arm-netbook mailing list arm-netbook@lists.phcomp.co.uk http://lists.phcomp.co.uk/mailman/listinfo/arm-netbook Send large attachments to arm-netbook@files.phcomp.co.uk
On Sun, Apr 16, 2017 at 12:42 PM, Allan Mwenda allanitomwesh@gmail.com wrote:
Just gonna ask here coz I'm too lazy.
:)
How hard would it be to repurpose one of these cheap $200 macbook clone things with intel atoms to take an eoma68 card instead? I can already imagine the rockchip one in it :)
yeah me too. ok, repurposing of existing casework comes up as a recurring theme, quite a lot: i was one of the people who believed, back when this project started, that it would be practical and perfectly reasonable. so i wrote it up as one of the updates, "laptop comparison". ha, cool, i just encountered this:
https://www.crowdsupply.com/eoma68/micro-desktop/updates/the-opposite-of-the...
i'm redoing that PCB you can see at the end of that one, except it'll be coloured green.. :)
this was the one: https://www.crowdsupply.com/eoma68/micro-desktop/updates/laptop-comparisons
and... ah. that's strange... i didn't add the bits about the impracticalities of sourcing the components.(which are flat-out impossible in the anticipated quantities). that _was_ the whole purpose of mentioning the update. duuUuh :)
ok so _somewhere_ i have a critique of the strategy which utilises pre-existing casework: it's a comprehensive fail, pure and simple.
why?
well, if you get some existing casework, it's likely to be at least 1 to 10 years old. the company that made the connectors - SPECIFICALLY for that SPECIFIC laptop case as SPECIAL ORDER ITEMS will have a unique relationship with the designer of the laptop.
conversations between you and that supplier would go something like this:
you: "hello! we want to make a PCB based around a proprietary laptop case! please give us 100 of your connectors!"
supplier (very puzzled supplier): "hello! glad to hear from you. are you a representative of the company whom we signed an NDA with whom we have multi-million-dollar supply contracts?"
you: "errr.... no? i just want 100 of your $0.10 connectors that you made 10 years ago"
supplier (who is probably trying to be veery diplomatic by now): "10 years ago? you want to give us $10 for some parts where the tooling's been destroyed over 9 years ago and it would cost us $100k to remake it, and it's a proprietary (copyrighted) design as part of one of our unique client contracts??"
... you get the general idea, allan? :)
even if it's a common design, as i've found out already, you need a *personal* connection - someone who *actually* has worked with that casework and knows *all* of the components *and* suppliers, has a good relationship with them, and is prepared to risk that because you're *guaranteed* to order at least 1k and preferably 10k units...
.all of this should give you the general impression that it is a f*** of a lot of work and risk for almost zero return. it's similar to the hilarious "how i made a $3 toaster for $1800" ted talk, which is well worth watching.
https://www.ted.com/talks/thomas_thwaites_how_i_built_a_toaster_from_scratch...
l.
God damn NDAs. And it really is a common design too, chinese companies are basically slapping logos on the same chassis. That is unfortunate, the situation you describe.
On 16 April 2017 16:01:43 GMT+03:00, Luke Kenneth Casson Leighton lkcl@lkcl.net wrote:
On Sun, Apr 16, 2017 at 12:42 PM, Allan Mwenda allanitomwesh@gmail.com wrote:
Just gonna ask here coz I'm too lazy.
:)
How hard would it be to repurpose one of these cheap $200 macbook
clone
things with intel atoms to take an eoma68 card instead? I can already imagine the rockchip one in it :)
yeah me too. ok, repurposing of existing casework comes up as a recurring theme, quite a lot: i was one of the people who believed, back when this project started, that it would be practical and perfectly reasonable. so i wrote it up as one of the updates, "laptop comparison". ha, cool, i just encountered this:
https://www.crowdsupply.com/eoma68/micro-desktop/updates/the-opposite-of-the...
i'm redoing that PCB you can see at the end of that one, except it'll be coloured green.. :)
this was the one: https://www.crowdsupply.com/eoma68/micro-desktop/updates/laptop-comparisons
and... ah. that's strange... i didn't add the bits about the impracticalities of sourcing the components.(which are flat-out impossible in the anticipated quantities). that _was_ the whole purpose of mentioning the update. duuUuh :)
ok so _somewhere_ i have a critique of the strategy which utilises pre-existing casework: it's a comprehensive fail, pure and simple.
why?
well, if you get some existing casework, it's likely to be at least 1 to 10 years old. the company that made the connectors - SPECIFICALLY for that SPECIFIC laptop case as SPECIAL ORDER ITEMS will have a unique relationship with the designer of the laptop.
conversations between you and that supplier would go something like this:
you: "hello! we want to make a PCB based around a proprietary laptop case! please give us 100 of your connectors!"
supplier (very puzzled supplier): "hello! glad to hear from you. are you a representative of the company whom we signed an NDA with whom we have multi-million-dollar supply contracts?"
you: "errr.... no? i just want 100 of your $0.10 connectors that you made 10 years ago"
supplier (who is probably trying to be veery diplomatic by now): "10 years ago? you want to give us $10 for some parts where the tooling's been destroyed over 9 years ago and it would cost us $100k to remake it, and it's a proprietary (copyrighted) design as part of one of our unique client contracts??"
... you get the general idea, allan? :)
even if it's a common design, as i've found out already, you need a *personal* connection - someone who *actually* has worked with that casework and knows *all* of the components *and* suppliers, has a good relationship with them, and is prepared to risk that because you're *guaranteed* to order at least 1k and preferably 10k units...
.all of this should give you the general impression that it is a f*** of a lot of work and risk for almost zero return. it's similar to the hilarious "how i made a $3 toaster for $1800" ted talk, which is well worth watching.
https://www.ted.com/talks/thomas_thwaites_how_i_built_a_toaster_from_scratch...
l.
arm-netbook mailing list arm-netbook@lists.phcomp.co.uk http://lists.phcomp.co.uk/mailman/listinfo/arm-netbook Send large attachments to arm-netbook@files.phcomp.co.uk
On Sun, Apr 16, 2017 at 2:31 PM, Allan Mwenda allanitomwesh@gmail.com wrote:
God damn NDAs. And it really is a common design too, chinese companies are basically slapping logos on the same chassis.
yyep... because the casework molds and PCB design files have made their way into the "system", where it's easy for that to happen.
the problem is: the moment you want to do something different - such as put in a 5mm x 54mm slot - that's a complete redesign of the molds. fees for developing molds are anything from $20k to $250k. there's a good article on bunniestudios (actually several):
https://www.bunniestudios.com/blog/?p=189 https://www.bunniestudios.com/blog/?p=4146
l.
arm-netbook@lists.phcomp.co.uk
-
Allan Mwenda -
Luke Kenneth Casson Leighton