Hacker News new | past | comments | ask | show | jobs | submit login
Acer Travelmate 4150 retrofitting to Raspberry Pi Laptop, Part 1 (codeandunicorns.com)
94 points by mitola on March 27, 2017 | hide | past | favorite | 48 comments



I'm interested in the results of this, but I'd suggest holding off and posting this to HN (and elsewhere) when the task is done. At the moment what we've got here is a goal and a disassembled laptop with absolutely nothing hooked up to the Pi, which is not really that interesting.

I'd also suggest that sufficiently interesting stories about why you couldn't manage to do this would also be interesting. ("I couldn't ever get the adaptor"... probably not interesting. "So I had to 3D print it, which was harder than I expected."... interesting. "I tried for weeks to get the RPi working and it almost sorta worked but it was only in monochrome and I couldn't figure out why."... interesting, and probably enough to trigger other people to want to help!)


Indeed not a very interesting read at that point, the idea itself is nice but pretty obvious. There is also nothing concrete about tackling the hard tasks like power supply or interfacing with the keyboard yet.

The author only briefly talked about the LCD driver board but didn't even mention which one he bought let alone share any experiences how it worked out. A "look the screen shows the output of the pi now, thats how I did it" or "I tried … and the screen is still black, not that easy as thought" would have been a better start for submitting such an article series to a wider audience.


Thank you for your kind reply. I will take that into account as you say, since when it will be hopefully more complete I'll publish an update.


Somewhat humorous in light of this comment on Schneier's post about the laptop ban:

"Get the 15" laptop, throw all the electronics away, leave just screen and keyboard. Put Raspberry Pi and three AA batteries inside. Voila - you are able to show TSA staff that this is really laptop, which executes some code, and have a lot of space inside to put in explosives.

Moreover, you have powerful processor and lots of program-controled pins to implement timer initiator."

https://www.schneier.com/blog/archives/2017/03/the_tsas_sele...


The problem with hollywood movie plots is people have been able to remove the optical drive or floppy drive from a laptop for decades, insert a baggie of weed or whatever criminal activity.

To some extent avoiding TSA scrutiny might be a goal behind the apparently meaningless quest for thin laptops at apple. No removable parts means you can't rip parts out and insert a baggie of whatever in the case. Also the laptop being lightweight means whatever you're sneaking in can't be too heavy, if a laptop is lighter than a derringer even if it physically fits its going to be weird explaining why your laptop weighs twice as much as a normal laptop.

Also I'm sure this looks fascinating on the xray machine.


> The problem with hollywood movie plots is people have been able to remove the optical drive or floppy drive from a laptop for decades, insert a baggie of weed or whatever criminal activity.

Indeed. And a decade or more before the movie's purported date of 2025, we realized that Johnny Mnemonic would be better served smuggling data by swallowing a toy balloon with micro SD cards in it. Today he could get away with just three or four cards.


Has the TSA ever weighed of even lifted up your laptop to nice how much it weighs?


For making a smuggler-laptop, it's probably simpler to 3D-print a fat chassis for a little-known brand of thin ultrabook so the agent wouldn't know that it's not supposed to be a half-inch-thick.


Except they all get scanned, so the person who runs the scanner would, arguably, be able to detect this. I imagine the dugout laptop is exactly why we have mandatory scanning of laptops.

I also imagine the software can automatically detect things that looks like contraband, explosives, etc. If not then we'd be seeing a lot more terrorist success using these methods.

There must be a lot of pressure on the devs who make the software that run these things. Whatever computer vision processing it does needs to be pretty good or else.


Much of the software at that level of detection is still research grade. What is usable for industry is either used in automation (eg Finding curled noodles at a straight noodle factory) or used in tech centric companies. Combine that with the government being a decade behind when its trying and I suspect that humans look at every scan.


I'd be very surprised if your average airport xray would be able to distinguish between lithium battery cell and well disguised explosive pack. Doubly so if the cell/explosive is inside an aluminum body of a laptop.


Well if that was true we'd be seeing nothing but explosions on planes. These systems are designed to notice explosives. We know the densities of common explosives vs li-ion batteries and are able to distinguish between them.


Or maybe, just maybe, the explosion threat is massively overblown and airport security is mostly just theater.


Or instead of explosives ten galaxy note 7's with theverge.com or something like that open to make them overheat.


Note7 were quite quickly banned from planes.


...So are explosives.


similarly, sprite_tm hdd controller hack shows that today anything embeds a computer..


This might be of help to you. The OEM version of this laptop seems to be the Compal LA-2601. Which is a very good thing, because the schematics are readily available [1]. So, getting the keyboard working with a PI3 should be a matter of rigging up a micro with enough pins to drive the matrix controller. Looks like an interesting project though, and I wish you good luck.

[1] http://www.s-manuals.com/pdf/motherboard/compal/compal_la-26...


How do you know this and how did you find that?


Google diving. I got curious as to what the keyboard controller was, so I looked up motherboard pictures and specs and eventually found that it uses an ENE KB910 for the keyboard controller. Searching there for how things were set up was mostly fruitless, until I stumbled upon a russian forum which had a pinout for the KB910 [1]. A few more stumbles and I found a result that stated that Compal was the OEM for the laptop -- it's not uncommon for laptops to be made by a white box maker. From there, a search for the model number brought up multiple sites with the schematics.

[1] http://bit.ly/2ob7YaG Not the actual url, because chrome's history ate that.


Impressive :) Thanks a lot for your dive into the subject. It's darn interesting indeed


Fun project. Will be interesting to see how this evolves.

There are also a couple of build kits like [1] for people who don't want to mess with old laptops and weird connectors. I don't have one, but I saw it in Ben Heck's youtube channel and thought it's pretty cool.

[1]: https://pi-top.com/product/pi-top


I think this is a good idea but as mentioned in the article the connectors to keyboard and monitor are hard.

I guess there are XF86 standards and stuff for the keyboard but I still find it hard to think that it will behave exactly like the original.

Further, I would like to see more info on if the LCD gets working with the RPi because it seems like a great project.

Still, good luck with how the conversion works in the end!


I was actually quite suprised how hard it is. The amount of different LCD and subsequent separated driver panels that are possible in configuration is absolutely huge. To be honest I am a lot less worried about LCD working then keyboard, since in the worst case I would need to rewire the whole one manually. But for LCD I will hopefully have everything needed after getting my specific LVDS driver board


A mapping for the keyboard matrix can be "discovered" by pressing all keys in sequence. There are lots of AVR-based projects that emulate an USB HID for a custom keyboard matrix or other input; for an example, see [1][2].

[1] https://deskthority.net/wiki/Easy_AVR_USB_Keyboard_Firmware

[2] https://geekhack.org/index.php?topic=51252.0


I understand the joy of tinkering, but why do you want to downgrade a working Intel Centrino 1.73GHz to an RPi3?

[EDIT] See below, seems that the laptop is already broken.


Because, once it's working, you have an upgradable laptop: every time a new rPI comes out, you can swap it out.

Eventually you'll have a better/faster computer than the Intel Centrino, with potentially much better battery life, modern hardware (USB3+, better radio connectivity), better software support (for instance, mainstream Linux distros are now deprecating support for non-PAE CPUs).

I've been giving this some thought: I have a gutted MacBook Pro I plan to perform the same surgery on. Someone spilled soda on it, so the mainboard was already fried... therefore it's not even a downgrade. It's an upgrade from a brick.


Reason I'm asking is: I have an old Centrino laptop as well as an RPi2 and RPi3. The Centrino is so much faster than the RPi's, it's not even in the same league. And as long as the Raspberrys are married to Broadcom, I think it will take quite some time until we have an RPi that has similar speed like my old Centrino. Don't get me wrong, I love my RPis, and if my old Centrino breaks, maybe I'd do something similar, but I'd never trash it while it's still working. Just seems wasteful to me.


It could also be not only about pure speed - look at what even the RPi 1 can do in terms of media playback. My (working) Centrino laptop can just barely play youtube html 5 videos. If you want a small mediaplayer (remember those portable DVD players? :P) this sounds totally viable.


Ditto. I have two dead macbook pros with good screens / keyboards. The mobos are dead and nobody will buy them from me. This seems like a possible future project. I make writing software and I've always wanted to make a cut down Linux distro for writers (there used to be one but its no longer available). Rpi would be a perfect target, and a MacBook screen and keyboard would be a nice addition

I'd also love to stick a pi in an old keyboard computer like an Atari ST or Amiga.


My email is in my profile. I'd love to exchange tips and encouragement as we progress on this project.


This is the concept behind the (much-better executed) EOMA68 project.[0] You can buy a single laptop chassis, and then upgrade or swap the processing unit as you please.

[0] https://www.crowdsupply.com/eoma68/micro-desktop


According to the article, it wasn't a working laptop, he bought it off eBay for £15 as a parts machine. His only requirement was a working screen and keyboard since he planned to gut it for the Pi.

I was thinking about doing something like this myself a while back, and my oldest laptop, a Dell Latitude CPx with a PIII 500MHz CPU, would be ideal since I could reuse the power supply and batteries (it has two battery bays) with a simple charge controller/voltage converter. But, the laptop is in nearly mint condition and runs so well I can't bear to destroy it. It's easily as fast as the newest Pi, and it runs OpenBSD very well, which will likely never be ported to the Pi.


Definitely get you on 'not fussing with it' and cost/effort - but for completeness:

There are other small SoC/Maker boards would fit the bill:

https://www.openbsd.org/armv7.html

And people are working on Pi support, after all:

http://marc.info/?l=openbsd-cvs&m=147059203101111&w=2


I was under the impression that Theo had put a hard "fuck no" on ever porting OpenBSD to the Pi until Broadcom fully opens up the hardware (full documentation and blob-free booting). While Broadcom has been releasing documentation, there are no official plans that I'm aware of for ever booting the Pi without the firmware blob.

Don't get me wrong, I will gladly welcome OpenBSD on the Pi if it ever happens; one of my favorite OSes on one of my favorite platforms. But I'm not holding my breath.

As for other SoCs/boards; there simply is no other board with the combination of support, community, power, cost efficiency, and versatility that the Pi enjoys. Pretty much any other SBC is lacking in one or more areas critical to making such an endeavor worthwhile. The only thing that comes close is the Nvidia Jetson TX2, and it's cost prohibitive for this kind of project when you can get a mainstream AMD64 ultrabook for half the price of that board alone.


Broadcom released documentation and driver source code for VideoCore in 2014 so it should be easier to port things to the RPi now: https://en.wikipedia.org/wiki/Free_and_open-source_graphics_...

AFAIK Linux distribution for the RPi still use the closed-source bootloader, which is some commercial RTOS that needs to sit on a FAT partition and runs on the VideoCore chip (!). There is work on a Free Software bootloader: https://github.com/christinaa/rpi-open-firmware


> According to the article, it wasn't a working laptop, he bought it off eBay for £15 as a parts machine. His only requirement was a working screen and keyboard since he planned to gut it for the Pi.

Ah, I missed that part and mistook "everything seems to be in order" to mean that this is a working laptop. My apologies.


Maybe power usage? The Pi3 will probably get way more battery life than the original laptop did. That is, of course, if the battery still has some life left in it. Also, since the rPi is such a popular platform, you can expect everything (WiFi, BT, etc) to work out of the box.


And how will he charge the battery without the charging unit on the mainboard?


He's adding that.. read the article?


I've got an old Asus L3000D from early 2000 in a bag. I took it out and looked at it. The keyboard is great, the screen is almost 1:1, a two buttons touchpad, floppy disk, DVD recorder, mic, audio jack, two 1394 ports (I forgot what they are), parallel, serial, VGA, two USB (1.1?), PCI express?, ethernet and modem :-) TV out, PS2 mouse and keyboard. Maybe without all the internal components it won't weight more that 1.5 kg. I remember I made it crawl with Ubuntu circa 2010. Perhaps I should give it a try as a Raspberry case.


I often think (dream) of a partial solution; using a rpi as a coprocessor, daughter card. My 2006 core 2 duo is fine for 90% of tasks, except for video decoding. Delegating that to a tiny gpu/vpu would release the main cpu for logic and IO.


How much weights the gutted laptop? How much can it weight after it will be finished?


I can update you on that. At the moment the heaviest part of the laptop is the display. But I think it may not be bellow 2kg, but above. I can update when I have a better estimate.


Sounds like you might have a balance problem after you get everything working. Might need to machine in a surface style kick stand


After a bit more investigation while I wait for the parts... I think it should be quite fine, beacuse the battery will probably be the heaviest part, but it is quite possible that either left or right side of the laptop will be quite heavier. Hopefully I'll make it fit :D I found an interesting battery with 12V in and separate 12 out and another 5V USB out, so it should be fairly straightforward hopefully with that to power everything with it.


RP3 laptop cases making appearance on Alibaba in 5....4.....3....


There's already the pi-top [0]. I've been considering getting one for work to use without a Raspberry Pi. It connects to the Pi using HDMI and USB so it might work as a portable, battery-powered version of a rack mount console.

[0] https://pi-top.com/product/pi-top




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

Search: