Hacker Newsnew | past | comments | ask | show | jobs | submitlogin

I've often dreamed of jury rigging four iPad 3 panels together. Think about it: a 19.4" diagonal 4096 x 3072 display, for the not-unreasonable price of $1,596. Drop the flash memory and A5X etc inside the iPad, and maybe you could get it down to $1000 or so (!).

No idea how technically feasible that is, but "tiling" could be one way to build a really high-res 20"+ display without requiring absurdly large, continuous error-free panels.



Great idea, I might do this myself. It's actually much cheaper than that - I'd say around $400.

1) Get just the panels off ebay for $60-$80 /each.

2) You can get the spec sheets for the panels with enough Googling. It's a 6-bit (yeah, not 8-bit) LVDS signal, same as most modern laptops.

3) The toughest part. Create a DVI-> high-speed LVDS converter. There might be something already available, though. A quick google only shows a Toshiba DSI -> LVDS converter capped at 1080p.

Ballparking the LVDS clock rate: 2048 * 1536 * 6 * 60 Hz = 1.1 GHz

Interesting idea, though.


I would have hoped that displayport would replace LVDS, which would allow theese kind of hacks to be much easier and a lot cheaper.

Have laptops made the switch yet? (since they use displayport externally seems like it would be in their interest to use it internally as well) In that case the new MPB panel would perhaps be more suitable (and probably cheaper in the end - as well as better quality (I'm assuming/hoping it has 8 bits, but I don't know)).

And with displayport you could just hook it up to a decent gamer graphic card and be done. I'm not sure how feasible 2048x1536 is for a single-link DVI (at 60 Hz you would probably be in trouble), and graphic cards with 4 dual link DVI connectors isn't exactly mainstream so you would have to get (at least two) active displayport to DL DVI adapters, and they are not cheap either.


It is a common misconception that DVI is limited to 1080p. The standard provides a budget of 165 million pixels per second, shared between active image, blanking and frame rate. It's easy math to figure out how far it can go based on a given frame rate and percentages of non-image time. I remember doing the math years ago and the results were well beyond 1080p. I don't know the Toshiba DVI->LVDS converter you are referring to. I have personally designed a couple of them and they have non such limitations.


Hey, can you email me at garbage1234@live.ca ? I'm interested in a DVI->LVDS converter, would like to talk with someone who has worked with them in the past.


I would love there to be some kind of hacker's hardware gadget that allowed people to take various panels and connect them to computers.

While you're talking about big high res panels, I'd be happy with an extra little low res panel; any of the Nintendo DS screens, for example. But being able to use these old little panels could prevent some of them ending up in landfill.

I guess the other problem is the weirdly sub-optimal nature of multiple monitors - despite years of people wanting to use multiple monitors it's still a bit kludgy.



it's not hardware and doesn't exactly what you'd like, but have a look at MIT's Junkyard Jumbotron: http://jumbotron.media.mit.edu

Essentially a bit of software to add other devices (ipads, phones etc.) as virtual screens




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

Search: