Welcome! Log In Create A New Profile

Advanced

printing a computer

Posted by random 
printing a computer
January 08, 2010 07:27AM
yesterday I was at the fablab protospace and the subject of printing a computer came up.
since the reprap can`t print silicon(yet).
an electronic computer will be impossible with the current reprap
but maybe we can already print a mechanical computer(see link).
so I have a request if there is anybody that can design such a computer
do you want to design it?



[en.wikipedia.org]
VDX
Re: printing a computer
January 08, 2010 07:39AM
... until we can fab nanomechanical structures, this wouldn't be usefull - the mechanical setup for a simple actual table-calculator wouldn't fit in a garage eye rolling smiley

Better look for printing/fabbing adapters from IC's (DIP and/or SMD) to standard (reprapable) connectors ...

Viktor
Re: printing a computer
January 08, 2010 08:09AM
its not really about usability but more about bragging right
that we can say that the reprap can print a computer
Re: printing a computer
January 08, 2010 12:14PM
How about making a "Babbage Machine". This I believe is defined as one of the first computers and is made from gears and such.

Bob Teeter


Bob Teeter
"What Box?"
Re: printing a computer
January 09, 2010 12:10AM
Or figure out and blog/wiki-up pick and place of SMD components. We can't even assemble a board yet, so I think it's a hot thing to work on.
Re: printing a computer
January 10, 2010 01:29AM
Babbage difference engines and other mechanical analog computers require parts made to fairly high tolerances(if it's out of tolerances it will make errors!), so any mechanical computer we consider making should be digital. The Zuse mechanical computer might be worth reproducing:
[www.epemag.com]

We could try printing fluidic logic elements(no moving parts!) and use those to make a computer, in fact Forrest Higgs managed to make a basic fluidic element a while back. I'll post fluidic logic element silhouettes if anyone's interested in turning them into something printable.

[en.wikipedia.org]
[www.tippettsfountains.com]

The [email protected] guys managed to print a working relay so we could make a relay computer if we wanted too...

Keep in mind that mechanical computers tend to be HUGE!
Re: printing a computer
January 10, 2010 09:17PM
What's a good example of a gearchain with small gears that people have RepRapped? I'm fuzzy on how good we've gotten.
Re: printing a computer
January 27, 2010 02:28AM
I wonder if a Curta would be printable. I'd love to have one winking smiley

[www.hpmuseum.org]
Re: printing a computer
January 28, 2010 01:11AM
I don't think so, the curta requires parts made to fairly high tolerances, if we're going to build a curta we're either going to have to increase print resolution or print a larger curta.
[1.bp.blogspot.com]

We might be able to build a Long Now Clock computer, as it's digital so it won't need to be made to high tolerances.
[www.longnow.org]
Re: printing a computer
January 28, 2010 10:45AM
We could print something like a Digicomp (plastic toy programmable computer from the 60's):

[www.retrothing.com]

I had one as a kid and enjoyed programming it to be a three-bit binary counter.

It would probably be more practical to laser-cut one from sheet plastic.
Re: printing a computer
March 10, 2010 01:45PM
I just remner this link [home.earthlink.net]

So first we need reprap resistor and other electronic pasive stuff,
the big jump will be active componentes like transistor. maybeusing negative resistance, or some sort of organic stuff?

Other posibility i think is reprap able of print lens, so we can cofus light easilly maybe make paths using light, or may be photography , oh i heard that before winking smiley
Re: printing a computer
March 21, 2010 10:39PM
How about a reprappable antikythera mechanism?
Re: printing a computer
July 17, 2010 03:03PM
Abacus! That should hopefully be possible.
Re: printing a computer
August 22, 2011 12:40AM
I think the starting point should be a mechanical flip-flop (which is at the hear of all digital devices). This "binary marble adding machine" would be able to be built on a RepRap - even the marbles could be printed:

[www.mekanimo.net]

This device just counts up to 7, but it shouldn't be too hard to expand the concept to a more general purpose digital computing machine like this:

[www.youtube.com]
Re: printing a computer
January 13, 2012 12:43AM
HOLD THE WAGON.

Wouldn't an automaton be so much better?

Automaton Link

Video Version.

Add a spring to this to make it self tapping.

Edited 1 time(s). Last edit at 01/13/2012 12:46AM by Penny Garfunkel.
elf
Re: printing a computer
April 01, 2012 06:00AM
You could try to design a hydraulic computer that uses water passing or not through valves, as its gates.
I think you just need 4 types of gate, AND, OR, NAND, NOR.
Wouldn't it be great if we could print the controller as part of a hydraulically driven (no stepper motors) RepRap printer.
Re: printing a computer
April 02, 2012 07:52AM
Elf, how large would a simple gate be in such a system ? I'm afraid the hydraulic reprap controller would not fit in my room smiling smiley

Actually, how would you make such a "hydraulic logic gate" ? I fail to imagine a design that wouldn't require electricity at some point.


Most of my technical comments should be correct, but is THIS one ?
Anyway, as a rule of thumb, always double check what people write.
elf
Re: printing a computer
April 02, 2012 09:42AM
Since you ask, I have just found a pdf describing a pneumatic 8-bit microprocessor, I have not had a chance to read it yet, but you can pneumatic microprocessor

Some years ago I saw a photograph or drawing of a microscopic steam engine which had a drop of water trapped in a very small hydraulic cylinder, a heating wire turned that drop of water into steam that then expanded, pushing against a piston that transferred the linear motion to the work that had to be done, this was microscopic so a hydraulic computer need not be large.

I suggested a hydraulic computer since water is less compressible than air and therefore more reliable, a pneumatic system would require an air compressor whereas a hydraulic system could run using a gravity fed water supply.

Yes the printer would require electricity to convert serial/USB communication to hydraulic pulses, and probably for the heater in the extruder (unless we use the sun to melt the plastic).

But this is a possibility not a final design, there is a lot of development to be done, like every other part of reprap, the long term aim is a printer that can print itself, so we should try to eliminate those components that must pe purchased, I hope that one day we will be able to print stepper motors, but until then this is one possibility.

But the OP was asking about a printable computer and that is the post I was addressing.

Edited 2 time(s). Last edit at 04/02/2012 09:52AM by elf.
elf
Re: printing a computer
April 02, 2012 10:07AM
Here is an article on fluidics from encyclopedia Britannica Fluidics
Re: printing a computer
April 02, 2012 02:57PM
Thanks for the links. Interesting...


Most of my technical comments should be correct, but is THIS one ?
Anyway, as a rule of thumb, always double check what people write.
Re: printing a computer
July 30, 2012 12:49PM
Hmm I think eventually the technology that would win out is DLP UV printing...theoretically that can go down to the nanoscale, if somehow one could change out the material in the bath and print multi-material style. But even then a lot of circuits are made using other technologies such as doping and electron-beam milling, etc.

Ink Jet 3D printers like the objet seem to only go down to about 16 microns, still too large to print a computer.

Not sure if one could *print* a PCB just yet, apart from the silicon pathways...hmm molecular assemblers...or perhaps another way of making computers altogether, that doesn't involve multi-modal construction...

Or perhaps the 3DP tool head of the future would combine multiple fabrication technologies, just by swapping out tool heads. So you could be printing polymer now, then shining an EBM onto it, then using a UV projection next, etc.
CPS
Re: printing a computer
December 29, 2012 11:44PM
Sure design a printer that can etch copper into silicon wafers accurate to sub-22nm

cpu are already 3d printer (not really), you just need a billion dollar factory
Re: printing a computer
February 17, 2014 10:35PM
It's been a while since the last post, but hey... someone here may also be interested in this one:

[xiaoji-chen.com]

It's a mechanical linkage computer, where "lever up" is one and "lever down" is zero (could even have "analog" setups using aangles maybe?). I tried to replicate it for a "mathematical toy" but didn't get very far before being distracted by other projects like handwinding motors :p.
Re: printing a computer
August 30, 2014 08:06PM
We could always print an Abacus. That's a mechanical computer. tongue sticking out smiley
Sorry, only registered users may post in this forum.

Click here to login