[gta02-core] Openmoko Beagle Hybrid

Dr. H. Nikolaus Schaller hns at computer.org
Fri May 14 17:36:51 CEST 2010


Am 14.05.2010 um 16:36 schrieb Werner Almesberger:

> Dr. H. Nikolaus Schaller wrote:
>> There is now a new Wiki page for the project:
>>
>>      http://wiki.openmoko.org/wiki/Openmoko_Beagle_Hybrid
>
> Kewl. But where's the duct tape ? :-)

Between PCB and LCM :-)

>
>> I have received some questions why we did not put all this into a  
>> nice
>> design. The main reason is that we can't redesign the Beagleboard (it
>> has fixed dimensions) and we can't afford to build plastic injection
>> moulds (if someone has an idea how to reduce cost this is very
>> welcome).
>
> Low-volume injection molding should be quite affordable if you
> provide the cast (aluminium) or at least a machine-ready design.
> Of course, if you have to pay for the entire design work too,
> things will get expensive.

We have done some research and asked for quotation and the result is  
that if you want to get below 50 USD per unit (which is still a lot of  
money) you need a volume of >200 units. If you make any small mistake,  
you have to repeat the process and cost will increase by factor 2...  
So it is not good for learning and protoyping.

> However, you may also want to consider making the parts directly,
> without going via a cast. This is much more expensive for larger
> quantities, but if you only need a handful of cases anyway, it
> should be more efficient.

The alternative would be 3D-Printing. There are now some quite good  
machines that can produce in ABS. Unfortunately these machines are  
rather expensive and operators want to have fast amortization. This  
raises cost of small quantities of cases like the freerunner well  
beyond 100 USD.

So our observation is that there is no method which allows to produce  
1-100 units for experimenting and testing - and still meets the  
expectations in quality and cost.

But we may have missed something. When I walked through the "Golden  
Mall" during CeBIT this year, I wondered how approx. 500 small asian  
booths could display approx. 50 different products each (headsets,  
mice, keyboards, batteries, plugs, covers, fake phones, real phones,  
etc...). Each one must have its own plastics. So I would estimate  
several million USD in moulds have been on display :)

> The issue then becomes access to equipment and experience. I think
> making a simple case should be little more than a weekend project
> for someone who's set up to do such things. The challenge seems to
> be to find such a person, or - if you're looking for an exciting
> new hobby - to become one :-)

A third alternative could be to experiment with Makerbot / Cupcake.  
Anyone tried to "print" a Freerunner case from the open 3D data?

BR,
Nikolaus



More information about the gta02-core mailing list