GP2X Gp2x As A $100 Laptop?


recycledelectrons

Still Fresh
Joined
Jan 4, 2006
Messages
4
I'm convinced that a machine with a laptop-sized touch screen, the guts of a PDA, and a stack of SD slots would be a winner.

I'd love to have a wi-fi tablet that allows web surfing at any hotspot. Touching the screen would handle 95% of my input, but an on-screen keyboard or a wireless keyobard would be nice.

I'd also love to have an ebook reader. The big screen is critical, as is power savings. I don't need a big, bag processor to read a HTML file.

It should be ruged enough to drive nails, and still weigh less than 2 lbs. Clearly, something's got to go. I say dump the multi-gigahertz procesors in favor of the 200MHz systems that I drooled over in graduate school. If a 200 MHz system could run the university's web server, it should be able to work for my laptop. After we pack a PDA's guts and a laptop's screen in, we can encase the whole thing in a very tough case.

I like the $100 laptop's idea of an easily changable power supply, so only a socket for an easily-build power adapter would be standard. Maybe a large place on the back with a nearly indestructible, rectified, protected connector and several 1/4-20 threaded inserts to allow physical attachment of the power supply would be acceptable.

Mine would quickly be customized so that it could run off of AA batteries, an internal rechargable battery, a hand-crank charger, or a solar panel built in on the back.

Extra SD card slots only cost 1 I/O line each, and allow much needed expandability. Most SD card slots would be accessed rarely, with atleast one acting as an internal HDD, a sd ROM card with textbooks on it, an SDIO card acting as a permanent wi-fi port, and another one allowing bluetooth keyboards and mice. These would not need to be external (where water could get into them) so I'd seal them inside under a screwed-in and gasketed panel. I know that this has not been tried, but the SDIO slot is today's version of the ISA slot, and I plan to treat it as such. That way, expansion is trivial over the next 5+ years.

My hope is to add a huge touch screen, and several extra SD card slots to a GP2X or GP32X.

Only one SD slot would be external. Keyboards would attach via bluetooth with the bluetooth SDIO card inside a gasketed, screwed-on panel. The power adapter would be sealed in when the (replacable) power adapter is correctly screwed on. That means that 4 pins have to be protected against shorting out when the whole thing is soaking wet. That should help reliability in adverse conditions considerably.

I've heard a rumor that the GP2X and GP32X's schematics are open-source. Is this true? If so, where do I get them?

Thanks!

Andy
 
I appologize if I seemed to imply that a GP2X or GP32X could be hacked for under $100. I meant that the basic concept of the $100 laptop (a cheaper-than-the-current-standard laptop-like system) could be built around the same core as these systems.

As for my ideas of hacking hardware, it seems that you are software-only guys. Could you point me to the hardware section?

Andy
 
Perhaps he has been confused with the 'Real' $100 Laptop:

http://laptop.media.mit.edu/

The GP2X has NOTHING to do with any of that.

It is a gaming system, nothing more and nothing less.
(which it is starting to do quite well)

The hardware spec is owned and only available to GPH. Heck we don't even have the updated kernel sources. So technically its not even fully open sourced yet.
 
Get a Sharp Zaurus if you want a cheap handheld laptop replacement to do those things, but then the cheapest i've seen them is $400 or so.
 
If you are actually serious, maybe some of the guys in #gp2xdev (efnet) could help but it seriously looked like random blabber to me - as you explained no details on how you planned to actually do this.

And the GP2X docs are not open source!
 
If you are actually serious, maybe some of the guys in #gp2xdev (efnet) could help but it seriously looked like random blabber to me - as you explained no details on how you planned to actually do this.

And the GP2X docs are not open source!


Thanks for that answer. I had heard a rumor that the hardware design was open-source. That must have been in error.

As for the details...those depends on the schematics, don't they? It's impossible to rework a hardware platform if you don't have the schematics.

I've interfaced SD card to Atmel AVR micros, so interfacing them to ARMs should not be all that hard, especially with the first one being done already. My only concern was: would there be an extra half-dozen unused pins on the main procesor?

The larger screen would be interesting. Given an ARM processor as the video processor, and the ability to give 720 x 480 output, an 8" to 11" screen would be usable at this resolution. Of course, LCD interfacing has never been easy. It would be ideal to get a screen that uses the existing driver chip, or a driver chip that's compatible with the existing one. (Again, without schematics, I can not design a modification.)

The touch screen would be interesting. They coem in a variety of interfaces, with the exact specification depending on the number of unused I/O pins that I can steal from the main ARM processor.

Once I had schematics for each part, I'd make my own PCBs for each part, and develop drivers for them. After each part worked, I'd try etching (or routing) PCBs for the entire system. I'd have an advantage over guys putting togeather palmtops...size is not as critical in a latop as it is in a palmtop.

Ohh well...this obviously is not the place to ask, because the hardware is not open source, and because of the lack of a development community capable of dealing intelligently with proposed hardware changes.

Andy
 
Last edited by a moderator:
I'd personally try to build a $100 laptop on top of a reference board instead of reverse-engineering an existing commercial product like the GP2X. It's much cheaper that way.

I think the Gumstix machines should be able to be sold to you cheaply enough in bulk to produce a decently priced Linux machine, though building the thing is of course very expensive.

MIT has a lot of sponsors giving out a lot of money to cover Negroponte's $100 laptop scheme, and they are definitely not getting their parts at cost, and absolutely not at retail.
 
I think this is a good idea, all you need to do is go to the marketplace and look for a broken gp2x or even gp32(new or used are cheap). I know of some sites that have good components if you need them.
 
I'd personally try to build a $100 laptop on top of a reference board instead of reverse-engineering an existing commercial product like the GP2X. It's much cheaper that way.

I should have killed this thread once I realized that the GP series of palmtops (hardware) was not opensource. I would never hack a commerical product to the extent of adding a new screen (when the old one is soldered it) and adding new SD cards (when I have no schematics.)

Having dropped the GP line of products from my search (they're great, but obviously not what I'm looking for) I am looking for an opensource (hardware) design for a palmtop.

If I can not find that after burning many hours of time, then I'll look for an SBC based on either an ARM or an Intel chip, that has an expansion card that drives a full-size (8" to 11") touch screen, an SPI connection, and a half-dozen TTL lines that can interrupt the processor. (multiple SD cards require a SPI interface, plus one TTL wire for an interrupt from each card, plus power and ground connections.)

If I can not find that, them I'l stuck. I'm not willing to spend the time to custom interface big LCD screens.

Andy
 
Last edited by a moderator:
worldwriter: I think this is more closely related to the STFU files.

Edit: geh... I'm drunk.
 
I'd personally try to build a $100 laptop on top of a reference board instead of reverse-engineering an existing commercial product like the GP2X. It's much cheaper that way.

I should have killed this thread once I realized that the GP series of palmtops (hardware) was not opensource. I would never hack a commerical product to the extent of adding a new screen (when the old one is soldered it) and adding new SD cards (when I have no schematics.)

Having dropped the GP line of products from my search (they're great, but obviously not what I'm looking for) I am looking for an opensource (hardware) design for a palmtop.

If I can not find that after burning many hours of time, then I'll look for an SBC based on either an ARM or an Intel chip, that has an expansion card that drives a full-size (8" to 11") touch screen, an SPI connection, and a half-dozen TTL lines that can interrupt the processor. (multiple SD cards require a SPI interface, plus one TTL wire for an interrupt from each card, plus power and ground connections.)

If I can not find that, them I'l stuck. I'm not willing to spend the time to custom interface big LCD screens.

Andy
There are a lot of mighty technical words in there. Perhaps it would be best if you lay off the amphetamines and had little time to rest before you hurt yourself.
 
Last edited by a moderator:
I think the big screen would be a battery guzzler, so I doubt AA's would suffice to run the system for an extended period of time.

Designing the system for networking is going to be a winner: ability to network will be the most important aspect of a device in the very near future, if it isn't already. WiFi is obviously necessary to do any real amount of work; Bluetooth is a good addition because it will allow you to interface with peripherals (as you mentioned -- though they will be rather expensive compared to the standard wired versions, and you'll need more batteries), but also to devices such as mobile phones and PDA's.

Maybe the Noka 770 is something along the lines of what you're looking for. Though the hardware isn't open, the software is. It's also way outside your price range, weighing in at $359.

You have a nice idea there, but I doubt you'll be able to manufacture it within a $100 budget. And you definitely won't be able to use the GP2X as a basis.
 
Back
Top