Hacker News new | past | comments | ask | show | jobs | submit login
Printed Electronics (2016) (thinfilm.no)
134 points by barelyusable on May 9, 2017 | hide | past | favorite | 51 comments



Imagine if this technology was available as inkjet cartridges. The potential for decentralization of electronics manufacturing, is staggering. So is the potential of electronics at the price of fancy photo paper. We would finally be in a position to begin replicating the open source software revolution, in electronics. The power of the hardware manufacturing giants would melt away.

High performance, nanometer-scale silicon would obviously still have to be manufactured the classic way, because, nanometers. But, instead of buying a precious tablet, you order a ribbon of of peel & stick wafers, of commodity CPU and RAM chips. You then stick them onto a printed, flexible nv-mem + screen to enable it to compute at modern speeds.

As people get used to disposable, custom electronics, they learn that for most of what they do with a single sheet of "paper", they actually need very little computing power. Rugged, cheap, micrometer-scale CPUs make a comeback. Kids study them in grade school, using an ordinary optical microscope. IC designers start to optimize for readability. The general public begins to take responsibility for their hardware and software stack, the way they take responsibility for their vehicle and their home.

/dream


This isn't really that much different from etching your own PCB. You just print the circuit on a laser printer, stick it on a copper board, pour some etchant and that's about it.

Of course, you then have to solder the components, which is a time-consuming and difficult process, what with the size of electronics these days. Most components' pads are a fraction of a square millimeter.

EDIT: I may be misunderstanding your point, as the article talks about printed displays etc, but I don't see how you could just get an inkjet cartridge that would allow a home printer to create a display...


> [...] pour some etchant [...] solder the components

I think that's where the differences would start for most people. In the parent's vision, you could assemble complex PCBs using ordinary office supplies and with little knowledge except about the circuit you want to build.

Today, you'd at least need a dedicated hobbyists' level of space, tools, and knowledge how to use the tools safely.


Unless your printer was just spitting out downloaded circuits, I am skeptical that you'd reach the masses. Already the Arduino et al. has helped to let people make electronics via more software and less hardware, and even that doesn't mean that there's one in every home.


This Thinfilm company claims to be able print basic displays already, and hints that the process involves 5 layers of material. I'm guessing that there are two outer conductive layers, of longitudinal and lateral electrode arrays respectively. Two of the inner layers would form the P-N junction of an OLED.

Not sure what that fifth layer would be, though. It could be something as mundane as a non-stretchable structural support. Or, it could be that there is more to OLEDs than the Wikipedia article lets on.

https://en.wikipedia.org/wiki/OLED#Polymer_light-emitting_di...

As for how this would be printed, that is probably just a matter of coaxing the required layer-chemicals into a suspension that has the right consistency for their own printers. I recall an article a while ago, about some pharmaceutical researchers teaming up with HP to develop chemical print heads that spit out protein solutions, and then spit drugs onto them, in order to rapidly test millions of different concentrations. These people might even be using that same printer.


As long as it isn't HP selling the cartridges.


"Aw, man, this thing barely printed two displays and a keyboard and now I need to buy a new cartridge for $10,000."


Hey! Whatever your job is, I like the way you think(tm) and would like to have people like you in the team! If you could use a free pool of computing resources and want to start building something that would actually make the world a better place(not-for-profit, no cheap, outsourced r&d labor === VC* targeted sell-your-soul project) - send me an email!


I can't tell if this is a joke or not (ambiguous X,Y,Z === A), but you don't have an email address if you're really looking for them to send you one.


OK, I may have used a comma instead of a semicolon(read "X; Y,Z).. also, I'm not an active yc contributor in terms of posts/comments - thought my email address would be visible, thanks for pointing that out. I'm also not really good at getting my point across :) and I'm only able to "sell" things I believe in and even then it takes some time - see my previous point - anyway, no joke, just clumsiness.


Good morning, stranger. I am curious about what you all are up to; the idea of a non-profit that has access to computing resources and bandwidth, is kind of interesting. Do you have something like a blog, a twitter feed, or a github org page?


Wow, this is pretty much 90% of the value of the old Plastic Logic business (are these guys related?)

Sad that they are spinning it as a physical products DRM although I certainly understand how that it is unique there. Cheap challenge/response authentication that even if you have all the materials and the printer you can't duplicate without the crypto secrets. Its a much better gizmo that a hologram sticker.

That said, I'd love to see them add OLEDs so that you can print a 'picture' and drop it on a table and have it light up. The signage options there would be pretty awesome too.


So how far we from being able to use this technology on a prototype basis and quickly (and cheaply) print electronics at home or at the office?


Huh. The site isn't completely clear; could this sort of technology cheaply print custom silicon layers, like ICs and some types of components? If so, that would be amazing. If I could print a to-spec FET, a 555 die, some BJTs, a handful of film resistors to +/- 1%...

Well baby, you got yourself a stew!


Most "printed" electronics mean either organic semiconductors, which generally have poor performance but are very cheap to print and flexible. This company also offers polysilicon, which is generally more expensive but is higher performance - it's used for TFT displays, for example. Both of these are usually patterned, so it's more for volume production that one-off like 3d printing.

I am not actually sure what the benefits of sheet-based logic are - RFID is obvious, but existing vendors prefer tiny, inflexible monocrystalline silicon dies instead. Are the Thinfilm RFID tags cheaper?


ARM demoed some printed SoCs (based on Cortex M0) a while back: https://semiaccurate.com/2015/11/18/arm-charts-path-printed-... The process is described here: http://www.molecularimprints.com/technology-overview/j-fil-o...


I'm thinking so, given they say that they can print displays and memory. That suggests that they can do a pretty decent number of semiconductors, which indeed could be rather useful.


Here are details (from 2015) of the collaboration with Xerox mentioned on the Thinfilm page: https://venturebeat.com/2015/01/21/thinfilm-teams-with-xerox...

They describe being able to print memory, with logic planned in the future. They also talk about commercializing and scaling the process.

Rights management for physical goods seems to be the most commonly touted usecase for this. Another example usecase I have heard mentioned is a thermometer and an indicator that changes colour to indicate spoilage that can be printed into the label of temperature sensitive goods.



For those in the know, how does the cost of 12-15¢ for an NFC chip compare with what a packaging company would pay now?


If this is cheaper than RFID tags, I'm seriously considering starting a company around selling a fridge device that could keep track of food that had a tag printed on them.

The Jetsons fridge that has always been promised, but never made. A fridge that could buy your food for you!


I think it was never made because it's useless. I know perfectly what's in my fridge. The problem is in deciding what's still edible and then maybe go for some grocery shopping.


Then explain why you keep opening it, wasting energy, to look for some nonexistent goodies.


For the same reason I keep opening Hacker News :)


There are other companies in this field too. Right next town here is PolyIC who does printed circuits for years for use as touch buttons, printed RfID, and alike. http://www.polyic.com/


TPM?

Standard critique of TPM includes pointing out that manufacture is a black box of trust. The ability to completely control both software and hardware seems like it would make this scheme more desirable.


now if the top layer could be a display and the other 6 layers are the rest of the computer.

I'd create a cube of all those sandwiched together.


To mitigate against the concern of parasitic capacitance raised by @pmoriarty, I would decouple the display layer from the computational layer using a shielding substrate.

You could get a very similar effect to what you're proposing if you used a future-version of a flexible OLED screen like the ones that LG[1] (and others) are beginning to produce. It's not exactly the case (yet) where you could fabricate it at home with your super cool electronics inject printer. It is, however, definitely a step in the right direction for giving enterprising people the ability to create open-source and bespoke tablets, phones, and other media devices that previously were only within the reach of very well-funded companies.

---

[1] http://english.yonhapnews.co.kr/business/2015/05/19/80/05010...


I'm not sure how practical this would be, due to parasitic capacitance.[1]

[1] - https://en.wikipedia.org/wiki/Parasitic_capacitance


What is the pitch of this technology?


Reading the sample use cases on the site makes it seem like smarter DRM is the pitch :(


I'm guessing about 300 dpi.


It must be too early, but I am failing to see the major benefits here.

Is it cheaper and more durable?


If you can "print" circuits instead of having to image them one wafer at a time that could be a big throughput boost.

My guess is they can't do 10nm or anything competitive with it yet, so it's only suitable for displays and such.


The application seems to be cheaper RFID tags.


> Ideal for simple indicator displays

Looks to me like just about any low frequency, low power circuit is a possibility.


Low density, too. Not for rolling out flash memory by the mile.


Fucking terrible that what they considered the number one item to mention on that page is how many patents they hold, rather than what the thing is, does, or is good for.


It's a form of signaling[0]. In this case, it means: "Like you what you see here? Well unfortunately we can't sell it to you since we don't have the manufacturing capacity. So you'll have to license our IP so you can build it yourself."

Case in point, Xerox licensed their IP and is already putting units in the field[1].

[0] https://en.wikipedia.org/wiki/Signalling_(economics)

[1] http://www.packagingdigest.com/brand-protection/smart-label-...


To be honest: they have been around for 20 years and haven't really done anything yet. I don't think they will. I think they want to sell their patents to a troll.


Someone once said "Patents have become a kind of currency now".

It's sad to see that people are scaling innovation using only patents.


50+ ways to ensure this won't be adopted universally until patents expire. 50+ ways to hold back innovation.


> Xerox opening 1.3Billion unit plant in early 2016

Sounds like the site needs an update.


"printed memory needs a refresh"


"If you want a vision of the future, imagine a human face booting on a stamp forever."

(https://slatestarcodex.com/2016/10/17/the-moral-of-the-story..., CTRL+f for "By 2050")


"CTRL+f" - you sound like a Chrome loser. (o;


I almost wrote C-s, but decided against it for the sake of people here who are not fluent in Emacs-speak.

Also, don't Firefox, Opera and IE use CTRL+f as well?


Firefox has an option to search text when you start typing.


why settle on GUI idioms when you can /

;)


heh good one.


test




Guidelines | FAQ | Lists | API | Security | Legal | Apply to YC | Contact

Search: