If you have access to a lab you can make your own. This video[0] has basic ingredients and instructions. This paper[1] explains some variations on the ink formulation.
This is handy for quick prototyping, but it seems that you can't solder on the parts and have to use "conductive double-sided tape or silver epoxy adhesive" instead.
To etch a PCB, the typical process is to (1) print a photo mask, (2) transfer it using photolithography or toner transfer, (3) dissolve away the exposed copper, and optionally drill holes. The coolest hack I've seen skips step (1) and (2) by directly "printing" onto the PCB using a UV laser, thus reducing the number of steps down to 2.
There was some related method mentioned recently on hackaday too, reportedly somewhat easier, where a guy in steps 1-2 prints with inkjet, and then dusts some toner over the ink and bakes the result in an oven. Never really got into hardware yet, but might help someone:
How conductive? Relatively high-resistance can be worked around for CMOS logic- and MCU-based circuits (mainly high RC constants), but analog that has to be built with high-decade resistances is likely to be noise-prone and flaky, and forget power sections, they'll have to be built on conventional glass-epoxy and hooked up with flex-edge headers.
The lack of a good gas-tight bond between component and conductor is also problematic: silver tarnishes. Silver-clad switch contacts can be made self-wiping, but not static contacts. Absent some kind of hot-weld method, I can just see SMD parts popping off of flex-circuits.
This is a promising start to a new base-level technology, but it's got a ways to go before it's ready for the R&D lab as a tool rather than a patient. I look forward to seeing these issues overcome.
Interesting read, but then again I agree that this is a prototype only protocol with little use because of the fragile nature of the circuits.
And as for prototyping, aren't there software suites around that bring sufficient capability in this regard?
This sounds very promising. If we add different inks that are semi conducting and resistive, we could actually print out whole circuits with components.
Probably still a few years before this becomes reality.
this is awesome - we'll soon be to a point where a person can prototype 100 units of a thing-of-the-internet in a weekend > circuit, mechanical and software.
23 comments
[ 0.21 ms ] story [ 64.6 ms ] threadSpecifically, "Mitsubishi NBSIJ-MU01"
[0]: http://www.youtube.com/watch?v=dfNByi-rrO4 [1]: http://staff.ustc.edu.cn/~yjxiong/paper/Science-2.pdf
To etch a PCB, the typical process is to (1) print a photo mask, (2) transfer it using photolithography or toner transfer, (3) dissolve away the exposed copper, and optionally drill holes. The coolest hack I've seen skips step (1) and (2) by directly "printing" onto the PCB using a UV laser, thus reducing the number of steps down to 2.
http://hackaday.com/2012/08/09/exposing-pcbs-with-a-home-mad...
http://hackaday.com/2013/11/02/perfect-pcbs-with-an-inkjet-p...
(now if you could combine the two...)
The lack of a good gas-tight bond between component and conductor is also problematic: silver tarnishes. Silver-clad switch contacts can be made self-wiping, but not static contacts. Absent some kind of hot-weld method, I can just see SMD parts popping off of flex-circuits.
This is a promising start to a new base-level technology, but it's got a ways to go before it's ready for the R&D lab as a tool rather than a patient. I look forward to seeing these issues overcome.
Lit:
Interesting read, but then again I agree that this is a prototype only protocol with little use because of the fragile nature of the circuits. And as for prototyping, aren't there software suites around that bring sufficient capability in this regard?