Re: Home built PCBs

[John Griessen <john_g@xxxxxxxxxx>]

If you imagine using a macro circuit printing method to 
get a flatter system, or to do glue logic, then speed may 
not be the main thing. If speed is the main thing in your 
design, then small is beautiful, but you have to get up to 
visible scale to connect to sensors or motors or displays 
eventually, so that is another place to apply homebrew PCBs. 
I am going to try out the direct toner method of printing conductive 
paths with a okidata LED printer in a while, so I would like to 
hear if anyone has suggestions for that.

Is it really feasible to print plastic transistors?  Are the ratios 
of length and width anything like in silicon?  IF so, then the 
registration of the paper handling path in the printer might 
allow 4X8 mm transistors with good tolerances.   What is it 
like using semiconducting plastic juice in a ink jet cartridge? 
Messy?  Short life of cartridge?

Aside from glue logic and connecting chips, I ran across the 
ball semiconductor method web page, which many MISC readers 
will be interested in.  
http://www.ball.co.jp/ballsemi_e.html

They propose a total abandoning of current semiconductor mfrg 
methods.  Forget flat surfaces of large wafers.  Do photo-
lithographic printing on little 1 mm balls of substrate material 
designed to hold only one, or maybe just part of one system. 
 Circuit interconection is a very limiting barrier right now, 
and I did not see where they address it, but what if each 
little circuit ball had 
a network connection circuit on it--a serial lightwave 
bus--then all the little circuit nodes could collaborate.  
Think of a 1 mm ball as a substrate for its surface area of 
about 2 mm.  2 mm square is 2 million microns, so if the 
transistor gate size is .5 micron you could get a 2 million 
transistor circuit.  If you design it like an F21 istead of 
a Pentium, a powerful processor 
and memory will fit.

The really exciting near term potential for the ball semi 
method is that it can allow quick small batches to run 
through the same fab as the high volume stuff. 
 I did not read if they use low cost masks with 
their spherical photolithography method, but I imagine there 
is something very different there too.  Mask registration 
may be done more locally than with masks for planar circuits.

We all see how high volume allows low prices, so that is 
a huge benefit to creativity and would enable development 
of real circuits by small groups or individuals.

At 11:34 PM 11/21/97 +1000, you wrote:
>:Massive instructions deleted ;
>
>Not like me a number of months ago, testing out an old idea 
>of where ever photocopier toner was conductive or not so I 
>could make Macro Roms.  Of course I got infinitive 
>resistence on my Multi-metre.
>
>Anybody interested in homebrew, there are a number of these schemes 
>using ink jet printer technology and macro-circuits on the peices of 
>plastic, using variouse drawings for the components.  So maybe we 
>could make processor designs with this.
>
>Certainly would be cheaper than fabbing, but who would be interested 
>in buy these macro products, wouldn't they be too slow and not as 
>effective as surfacemounting a real chip? 
>
>Wayne.
>
>

From: jfox@dnai.com (Jeff Fox)
Subject: Re: MISC-d Digest V97 #49
Resent-From: MISC@pisa.rockefeller.edu
X-Mailing-List: <MISC@pisa.rockefeller.edu> archive/latest/684
X-Loop: MISC@pisa.rockefeller.edu
Resent-Sender: MISC-request@pisa.rockefeller.edu

Dear MISC readers:

I think the discussion of homebuilt hardware design is quite
interesting.  I think there are many people in the group who
are interested.