A New Prop Printed Circuit Board

Humanoido

Can you tell me if anyone has made printed circuit boards in three dimensions, i.e. in mathematical shapes with curves? To minimize challenges with the chip plane, a 44-Pin QFN Chip and a retainer could be used.

Leon

I've had large PCBs made that were bowed so badly in both planes that they didn't fit in the rack. I never checked to see what function the shape was, though.

Flexible PCBs could be constrained to fit many mathematical functions. I can't think what use it would be, though.

Leon

▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔
Amateur radio callsign: G1HSM

Post Edited (Leon) : 12/29/2009 11:46:21 AM GMT

TonyWaite

Flexi or flexi-rigid PCB's are an established technology, but dramatically more expensive than the usual rigid ones.

It's easier mechanically to have areas of stiffening in the design, ie a rigid area to place the components, with the flexible areas handling the interconnect.

I've only used them where there was no alternative spacewise; but you have to be really careful with the mechanical design regarding bend radii etc and understanding the consequences of flexibility or movement or reliability suffers.

Regards,

T o n y

Humanoido

Curving can be constrained to that typical of OPTICS and EMR to comply with 4th order differential curves. I would start with a simple curve at first. Eventually the project would use a multidimension of EMR conducted by the prop chips. Simple PLANAR curving is the most easy to achieve. However, at the moment, I'm more interested in compounding curves. The idea is to dimensionally stratify the Prop chips in higher density with a more rapid hierarchy of EMR communication that follows the physical laws of optics. This could break the barriers set by previous tree dimension tactics - obsoleting a twelve dimensional hypercube for example. Some interesting illustrations follow.

Post Edited (humanoido) : 12/29/2009 3:50:21 PM GMT

Leon

What on earth does that gobbledygook mean?

Leon

▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔
Amateur radio callsign: G1HSM

heater

What?

▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔
For me, the past is not over yet.

Humanoido

Leon: I am attempting to marry the field of optics, curved space, and EMR, with the prop chip for a faster way of "networking" hundreds, even thousands of prop signals simultaneously. If it works, maybe future desktop supercomputers can run faster, cooler, and need no circuit paths. I built a few simple examples in 2000 using the BASIC Stamp which were successful (not posted), but the Propeller chip has features to make the project even more attractive and feature-laden, so I have re-initiated the work. Having curved PCBs would make everything much more simple, though maybe not so easy to explain. To define propagation and fields, there are differential curves already derived for space time. There are already laws of optics. This can simplify the geometry for curving boards. Except I have no process for making curved boards, hence the question about how to make a dimensionally curved board. Reheating the polymer substrate will warp the board but do some awful things to the existing pathways. It is likely these pathways will mostly be eliminated but I would like to retain some pathways for minimalistic circuits to include eeprom, etc. and have some mounting positions intact, the propeller chip included. The diagrams show this is not a simple case of bending a two dimensional board. I hope this explanation helps.

Leon

It's still gobbledygook! I don't see the relevance of optics, unless you are using optical comms. As for curved space, you need a very strong gravitational field for that, and I can't see how that would be relevant, either.

Leon

▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔
Amateur radio callsign: G1HSM

Post Edited (Leon) : 12/29/2009 4:53:30 PM GMT

Humanoido

What?

Leon

Hpw do you propose to curve space, then, if you don't use gravity? The only other way is with an accelerating system, which Einstein showed is equivalent to gravitation.

If you want to build a massively parallel system, there are much better devices than the Propeller.

Leon

▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔
Amateur radio callsign: G1HSM

Humanoido

Leon: The curved space is represented by the printed circuit board. It's an analogy. How to curve it is the question I asked you.

Leon

But why do you need a curved PCB? A flat PCB can accommodate any connection topology one can imagine, from a simple systolic array to a multi-dimensional hypercube or a torus.

Leon

▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔
Amateur radio callsign: G1HSM

Post Edited (Leon) : 12/29/2009 10:41:25 PM GMT

pacman

Whilst not curved (and definitely not optics) I do recall that one of the supercomputers in the mid 80's was using a ring topology with it's circuit boards. IIRC the 'logic' was that the electrons would have less distance to travel and thus the machine could run faster. At the time I remember thinking "how cool is that!" - They probably wern't the first to use that technology (but I was young{ish} and only just getting onto computers..

Though the PCB's were flat (like normal) they were curved (like a piece of slot car track).

Instead of a curved surface, could you get away with a series of flat boards arranged in a sphere? (much like a 20 {or more} sided die?)

▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔
=================
The future is in our hands.
Which way to the future?
=================

Martin Hodge

I curve my PCB's using synnergystic synaptitudes with poly-limitric cragameeses. You you can align the phase induced impedance with a hysteresis matched soldering tip...

A little fairy dust and yetti poop will get things moving nicely.

James Long

LOL.........

I hate to say it...but that is funny. You guys crack me up.

I think Humanoid was being serious........I took it that way.

I see what you are getting at, but I don't think there is a way to implement it. Even Flex PCB are going to have restrictions. So a compound curve is probably never going to happen, without some major bucks spent to do it. A compound curve will never sit flat (without relief cuts), so having one made would be extremely difficult and expensive.

Interesting twist (no pun intended).......but probably not feasible.

James L

▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔
James L
Partner/Designer
Lil Brother SMT Assembly Services

Are you addicted to technology or Micro-controllers..... then checkout the forums at Savage Circuits. Learn to build your own Gizmos!

TonyWaite

To answer the question seriously: if you needed to make such a 3D interconnect then the following would work:
- define the detail of the lattice using a 3D CAD model;
- micro-fabricate the resulting 3D matrix in titanium using laser sintering;
- plate with an insulating layer using plasma-deposition;
- build a conductive layer similarly;
- etch out the unwanted interconnect using 3D laser ablation;
All of the above is doable with existing lab technology.

Regards,

T o n y

Toby Seckshund

I have a sick bunny, could the parameters be retuned. There is a shortage of Yetis over here (but not according to HER ).

Fairy dust and a stick, with a star on it, was my Christmas wish, saves lugging around a toolbox, and thinking.

▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔
Style and grace : Nil point

Post Edited (Toby Seckshund) : 12/30/2009 10:34:37 AM GMT

Phil Pilgrim (PhiPi)

The Propeller is fast, but not so fast that connection topology has to be mimicked in fabrication geometry. Even one of the Cray supercomputers, whose mechanical geometry resembled the folds of the brain's cerebral cortex used flat circuit boards. Anything else would be an escapist flight of fancy — and an expensive one at that.

-Phil

kwinn

The Cray supercomputer I think you are talking about had it's boards mounted radially with the backplane towards the center of the circle. This was done to reduce the distance signals had to travel and allow adequate airflow for cooling.

ElectricAye

Feeding sea urchins lots and lots of silver and copper and tin???

I've heard some fungi will deposit heavy metals into their cell walls, and the growth patterns for some of these are phototaxic and/or galvanotaxic.

Why not grow your next circuit?

Humanoido

In review of some postings, I cannot stop laughing, yes, please send one package of Fairy Dust and Yetti droppings. If it's cracked up to what you claim, it could save light years of project time and increase garden vegetable yields to supply the entire world this winter. After feeding the world and curing hunger, we can stop war, perpetuate peace, obsolete sickness (taking care of Toby's sick bunny) and old age, let money grow on trees, and travel beyond the stars to enlist in the Alien Exchange Program.

Seriously though, there was an article about growing computers and circuits using ideas from DNA and crystals in respective solutions. Was it in Scientific American or Science Digest? It seems it would take time to develop this technology and crystal growing techniques are somewhat limited. It could be a future project though, to grow some simple pathways on dimensional printed circuit substrates, and components. I am sure someone is already working on this. [noparse]:)[/noparse]

Certainly the Cray put new ideas into actual use, and reducing wire length was a critical and important aspect of gaining speed. Nothing says that will be the important and critical aspect of using a Propeller chip. The prop has other features and attributes. I think everyone knows that prop speed by itself is not going to set a supercomputer speed record today. However, consider hybrid technology where the sum of the two techniques is greater than the one.

Although one could go the CAD route in three dimensional etching "routering" techniques for manufacturing and repeatability with precision, and even though it is possible to cheaply go with a store bought substrate and a paint-on copper clad kit so to speak, the idea came up for a less expensive approach. After some prudent shopping, there are a number of products with compounded curves, cubes, spheres, toroidal cores, etc. that are pre-made can be purchased off the shelf cheaply. The circuit pathways can be localized to a smaller point and affixed to the curve. A hole can be drilled from the obverse side to feed through and in that way, reflectance etc. can honor the original pathways and the curved surface. If the islands are reduced toward infinitesimally small sizes, layered or whatever, high densities can be achieved.

I have one project to complete and may begin a prototype of this to show a working model. Building it is one thing, programming it in Assembler is another. How is your prop assembler programming skill?

James Long

humanoido said...
How is your prop assembler programming skill?

Ha I can answer that for myself..........NOT GOOD ENOUGH!!

Wow.......I thought I got deep into projects.........

(background check here......my wife says I always dive way to deep into the small interests I get into. She says I take a small project and turn it into a life long adventure, studying the past, present and future of things remotely relevant to what I'm working on. She states, "that's why you learn a lot but finish little.")

James L

▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔
James L
Partner/Designer
Lil Brother SMT Assembly Services

Are you addicted to technology or Micro-controllers..... then checkout the forums at Savage Circuits. Learn to build your own Gizmos!

Humanoido

James Long said...
(background check here......my wife says I always dive way to deep into the small interests I get into. She says I take a small project and turn it into a life long adventure, studying the past, present and future of things remotely relevant to what I'm working on. She states, "that's why you learn a lot but finish little.")

What an intelligent wife - some people have all the luck! She knows that great scientist minds often turn projects into a lifetime adventure. Take for example, Albert Einstein. Even as a small child, he was envisioning how it would be possible to travel on a light beam. Later in life, as an adult, he gave us the laws of time travel based on the speed of light and he showed how gravity can bend light.

James Long

humanoido said...

James Long said...
(background check here......my wife says I always dive way to deep into the small interests I get into. She says I take a small project and turn it into a life long adventure, studying the past, present and future of things remotely relevant to what I'm working on. She states, "that's why you learn a lot but finish little.")

What an intelligent wife - some people have all the luck! She knows that great scientist minds often turn projects into a lifetime adventure. Take for example, Albert Einstein. Even as a small child, he was envisioning how it would be possible to travel on a light beam. Later in life, as an adult, he gave us the laws of time travel based on the speed of light and he showed how gravity can bend light.

What is funny.......after 16 years (wow....funny she put up with me that long), she now knows how to "steer" me in a direction to get results.

I have to say......she is quite intelligent for someone who is not technology minded. I got her a Stamp education kit for her birthday. She is reluctant to get past the "hello world!" part.

I try to push her (gently) but she isn't moving much on it. I remind her every now and then to see if more progress could be made. It hasn't worked yet.

I should enlist people from here to inspire her. I don't want her to do anything but enjoy herself. She has a great mind.

James L

Now back on topic.

▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔
James L
Partner/Designer
Lil Brother SMT Assembly Services

Are you addicted to technology or Micro-controllers..... then checkout the forums at Savage Circuits. Learn to build your own Gizmos!

kwinn

@humanido, we already grow computers. The problem is it is very difficult and takes many years to program them, and quite often they do not want to accept the programming.

While we can connect computing nodes in a variety of architectures (arrays, cubes, hypercubes, etc) to produce supercomputers we are unfortunately limited to building them in 3 dimensions. That means our interconnect wiring may get rather long as the number of computers increases.

As much as I like the prop (awesome chip for I/O and control tasks) I would not use it if I were to spend the time, money, and effort to design a supercomputer of any kind. Probably one of the graphics GPU's, an Intel/AMD/or similar chip and 1 to 4 gig of ram per node.

Martin Hodge

Apologies. I really thought it was a joke. I've been marveling recently at all the "free energy magnet motor" videos on youtube and jumped to conclusions. Shows what I know. Just an hour ago I erroneously spent an entire 90 seconds trying, with ardor, to solder a joint with a spool of insulated hookup wire... "Damn, this solder sucks"

eiplanner

Humanoido,

I'm hearing you, but I think you are starting in the wrong place. The magnificent ideas you have for warping the circuit boards in such a manner to accommodate

a specific supercomputer array of prop chips is sort of like making the wagon faster and more durable when there are already more highly advanced automobiles

in the world. The designs you are considering will no doubt be incorporated someday at the silicon or crystal level. The will be put into form on a microscopic plane

and not so much the large component level that you are speaking of.

pacman

One other idea I've seen recently is something called "Circuit in Plastic"

Basically it was a plastic substrate (that could be any shape) with the tracks laid on it (sort of like tin foil wrapped around a mould) then you encase the whole thing in plastic again.


So you could do a lens or a ball or some sort of weird plastic impeller with truly embedded electronics.

follow the link to see more...www.abc.net.au/tv/newinventors/txt/s2616421.htm

▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔
=================
The future is in our hands.
Which way to the future?
=================

ErNa

Hello, the year comes to an end, and this task will surely not be solved this year! But, there are interesting thoughts popping up. The first is: it makes no sense. so this is the best starting point to make no mistakes: put something together, see what happens. The prop gives you 8 cogs and means to establish communication. So 2nd: create a software which transparently communicates from cog to cog. Attach a LED to a prop and a phototransistor: slaughter an optoswitch from an old floppy drive. 3rd: mount the props to moving platform. 4th: by sending messages via the led and receiving the messages, the individual props can start to move and arrange themselves to communicate with as much others as they like. Install mirrors to ceiling and walls. Replace the mirrors by moving mirrors from a laser show. And let control the props those mirrors. You will create a living community, watch, what's going and report "the daily soap" to us here in the forum.
Happy New Year, ErNa

▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔
cmapspublic3.ihmc.us:80/servlet/SBReadResourceServlet?rid=1181572927203_421963583_5511&partName=htmltext
Hello Rest Of The World
Hello Debris
Install a propeller and blow them away

heater

I can't for the life of me get the point of all this multi-dimensional curved space-time Propeller layout idea but...

...the idea of an arrangement of PCB tiles into some platonic solid, say an icosahedron, with a Prop sitting in the middle of each facet surrounded by multi coloured LEDs, sensors and shining polished ground plane like some demented Sputnik is starting to appeal to me.

Lets see, 20 Props at 10 euro each plus sensors, LEDS and PCBs....nah not going to happen here.

▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔
For me, the past is not over yet.

ErNa

A hobby just a hobby. The propeller allows hobbyist to solve problem on the go, which some scientist believe to be unsolvable. It is just a question to answer: I have a propeller, so what to do. I have 20 propellers: what to do? It is like having a magnet: a magnet, a sheet of paper, a coin, and children are bewitched. Have 20 magnets and put them onto a fridge, and you can study the behavior of a multi-particle system with bipolar interaction. Put them on a table, keep them from flipping and make stray experiments with magnetic monopols. Discover, that dipols become monopoles, if you have a potential, that keep them from flipping. Yes, it's not going to happen here. It happens already! ErNa
p.s. How could I ever imagine to emulate a z80 on a prop?

▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔
cmapspublic3.ihmc.us:80/servlet/SBReadResourceServlet?rid=1181572927203_421963583_5511&partName=htmltext
Hello Rest Of The World
Hello Debris
Install a propeller and blow them away