f382f8 No.13
How about an open source phone/umpc/phablet?
I have had this project in mind for some time now.
My current idea is a system based around the freescale imx6quad, that fucker has everything and it is very open. 2d/3d graphics acceleration, sata, gigabit LAN, chip level encryption, tamper security, LVDS, MIPI DSI, CSI camera interface and much more. The gsm module power and mic connections can be controlled by the processor, allowing full control and fredom over your device. It would be the size of an UMPC while still allowing you full freedom and even stuff like sms and call encryption. If enough people are into the idea we could might even get it kickstartet. I hope the EE's of cyber are interested since this would grant us full control over our personal data.
The project github is at
https://github.com/Anongineer/OSPhoneIRC channel is at #OSPhone on rizon
Post last edited at
f382f8 No.14
>>13OSH GSM is currently non existing. The only way to make it secure would be to isolate the microphone and everything else from the gsm module and have the processor "wake it up" when its needed.
03f779 No.15
>>14Breaking the circuit to any component isn't exactly hard.
ea4189 No.17
>>15Exactly, the GSM module could be anything off the shelf.
40fdff No.19
sounds good
03f779 No.20
This is outside my area of expertise but if a subject matter expert feeds me schematics I can do production. I have enough spare cc laminate laying around I can order components and make as many prototypes as it takes to get it right.
Thinking of setting up a vacuum table so I can mass mould plastics.
03b04a No.22
>>20Now this is a fucking board.
22 get
3fa720 No.26
>I hope the EE's of cyber are interestedOh, we are. Or at least, I am. If you want a dev board for starting software work while the actual hardware gets ironed out,
http://www.solid-run.com/ sells the hummingboard, which is not too bad.
03b04a No.27
>>26ditto, show us some more, maybe we can play around
03f779 No.30
>>26I'll order one on the first
cfcc02 No.32
OP here. I can start designing some parts of the circuit needed and find the matching components. If there are anons that can do some case designing, or maybe find some nice prebuild case, then it would help alot.
cfcc02 No.34
>>20As in you can make the cases? We''ll have to order the circuit boards from a prototyping service that does parts assembly. Unless someone here has an infrared heating station for bga's or a masochist with a heat gun :^)
03f779 No.36
>>34>masochist>Autism is a neurodevelopmental disorder characterized by impaired social interaction, verbal and non-verbal communication, and restricted and repetitive behavior.http://www.tools-super.com/scotle-ir6000-infrared-bga-rework-station-new-version-p-9.htmlSoon…
cfcc02 No.37
>>36500 bucks? I just came.
03f779 No.38
>>37Best you are going to find in that price range.
0b5249 No.39
this sounds awesome, please do this.
f382f8 No.41
Here is the fact sheet for the SOC. I'm working on the power distribution and sequencing right now.
f382f8 No.42
f382f8 No.43
f382f8 No.44
>>43It seems the NSA has stolen the encryption keys for most of the worlds sim cards
https://firstlook.org/theintercept/2015/02/19/great-sim-heist/ I hope this project will be able to add a second layer of encryption to communications.
80d59c No.51
PLEASE do this. I will help fund it in any way I can, aside from being rather poor.
3fa720 No.52
As far as assembly of the PCB, I recommend only getting the BGA CPU assembled by a shop, unless you can find one with reasonable prices. I can probably be better/cheaper assembly down to at least 0603 parts. QFN can be a bitch unless you have stencils (maybe impossible with a BGA package presoldered.), but TSSOP or QFP (and similar) are not too bad.
03f779 No.53
>>51Educated people who are willing to put the time and effort to help are invaluable. Anongineer is posting resources so you can get smart. The more people we have working on this the better at this point. That said the posts that started this board were basically about how he is broke as fuck so maybe he should make a thread offering his services.
>>52Working on saving up for equipment to do it all at home.
3fa720 No.54
>>53If shit works out I may have a pick and place machine in a month or three. Going to depend on what money I make off of what projects to afford it, but if things work out how I hope I will both be able to afford it and need it.
For designing something like a high-end ARM board a pick and place is invaluable because the frequency the board operates at demands a whole lot of terminating resistors and caps for signal integrity.
03f779 No.56
>>54>in a month or threeOn my budget that is a long way away. What model are you looking at?
3fa720 No.57
>>56Probably the cheapest one I can pick up. We're debating how cheap we can get away with but as we're capable of improving the hardware we really just need to make sure we can get something reliable.
As I said this is more or less dependent on a certain very SMD-heavy project going as well as planned but it certainly looks good so far. Currently we have a couple cheap airstations and do everything by hand but for 100+ SMD parts per board that's a real stretch.
f382f8 No.60
You guys really surprised me with your enthusiasm towards this project. I think I'm gonna make a github with all the project files, data sheets and other relevant info. I just hope we can collaborate to make this a reality, and i even think a lot of normal people could get interested if they find out that privacy is a choice.
03f779 No.62
>>60I mentioned it in an email to Wheels and he said he is interested.
f382f8 No.63
>>60Found some nice memory too
http://www.alliancememory.com/pdf/ddr3/2GB-AS4C128M16D3L.pdf>>62Seriously? If hotwheels gets behind it, it could become a sitewide project. I also think we should get some programmers from /tech/, since its going to run linux.
Post last edited at
f382f8 No.64
>>63Cpu power getting worked out.
3fa720 No.66
Ok, after looking at some of what solid-run.com offers, I think it might be a good idea to look at the micro-SOM modules it offers.
http://www.solid-run.com/products/micro-som/micro-som-specifications/Probably the Dual is a good option, with a GC2000 GPU, 1 GB RAM, 802.11 and bluetooth all on a 47x30 mm module. It's going to add a bit of thickness but I find it very tempting to use something which can already be purchased in bulk and would push us forward to the stage where we develop phone-specific hardware. At least as a rev 1.0.
Thoughts? Should we limit our options using someone else's module? Or do we re-invent it and then have to figure out how to assemble it?
f382f8 No.67
>>66I think this is a great idea, though we should develop our own module along the side. This would enable software development and we could make some hardware to give the cpu full control over GSM-mic connection and so forth.
3fa720 No.68
>>67Yeah, we'll have to build another board to socket onto the module with power, GPS, GSM, MIPI DSI, and camera breakout, among other things.
Something we're going to have to deal with at some point is MIPI. MIPI is a bitch and most of the info concerning timing is usually not open so we're going to need to figure that out somehow or other. If anyone has more info on that I would certainly be interested.
f382f8 No.69
>>68I know mike from mikes electric stuff reverse engineered the mipi lane on an ipod nano screen. And it looks like the mipi driver is all ready built into the chip, and that should make everything a lot easier.
3fa720 No.70
>>69I'm hoping that helps but I assume timing is screen specific so we'd probably have to figure that out ourselves.
f382f8 No.71
>>70https://community.freescale.com/thread/305613 Yeah. But it looks like the mipi configuration is integrated into the linux distribution provided by freescale.
f382f8 No.72
>>71We just have to find a screen with semi-open specs.
f382f8 No.73
>>72I think we are golden if we just pick a mainstream phone display. Looks like most of them are reverse engineered.
3fa720 No.74
>>73Not too much of a problem, then. That's good.
f382f8 No.75
>>74We also need an anon with a 3D printer or some plastic skills, since most design flows in the industry is
case design(with dimensions)->pcb/hardware design
3fa720 No.76
>>75We probably need a pretty good 3D printer to make something decent. However I would say you could probably use the screen size as the primary case size, and then the only variable would be thickness- which would be about as thin as you could make it.
I suppose screen size is up for debate but 5.5 inches is pretty standard for a 1920x1080.
f382f8 No.77
>>76Agreed, 5.5 is large enough for general purpose computing and it fits in your pocket.
f382f8 No.78
3fa720 No.79
>>78Ok, looking forward to working on this. I have a good basic understanding of pretty much every part of this but my strong suit is software and linux, so I'll probably focus on that.
f382f8 No.80
>>79We need all the help we can get anon, its much appreciated.
03b04a No.84
03f779 No.85
>>84Admit you lied or provide a source.
03b04a No.86
OP Check out this other imx6quad machine being made by the brilliant hacker Andrew "bunnie" Huang (of "Hacking the XBOX" fame
https://www.crowdsupply.com/kosagi/novena-open-laptop 03b04a No.87
Let's make an IRC on rizon dudes
03f779 No.88
f382f8 No.90
>>86Looks awesome, though 2k+ for a laptop is a bit much.
f382f8 No.91
>>87I made one at #OSPhone.
f382f8 No.92
>>91why does it keep removing my name, god damnit.
03b04a No.93
03b04a No.94
f382f8 No.95
>>94Okay I'm trying to compile a list of parts. So far i have the ram, processor and power supplies down. We i think we need most normal phone sensors, some flash memory for the kernel/firmware and all the other chips i forgot about.
f382f8 No.96
>>95File with parts is now on git
f382f8 No.97
>>96Block diagram added to git.
f382f8 No.98
Edit any mistakes i make if you find some.
49c928 No.99
>>95I dunno about anybody else but I wouldn't mind losing the fancy gimmicks (accelerators and such) so long as it ran linux proper (not android or shitty knockoff android, but not necessarily desktop linux) and the baseband processor did not control the main processor, which I can't imagine would be hard if you just had a smaller "slave" SoC that would itself be controlled by the main processor. Also it would be really cool if more ports were broken out like ethernet or serial so you could hook it up and do networking stuff you'd normally need a laptop for.
I'm not much good at hardware design but I can help with software if it gets that far.
f382f8 No.100
>>99Sensors are luckily very easy to implement since they work over i2c/SPI. Full duplex gigabit ethernet chip has already been tracked down.
49c928 No.101
>>100They add cost though and for me that's an issue. The IC may only cost a few dollars but you have to count the extra design work, the added board complexity and the extra manufacturing steps. They add up when you try to do everything. There are open source(ish) phones out there but they're all like $2000.
I know everybody has their own priorities but what I'd be interested in would be something cheap and functional and highly connectable. I wanna be able to whip out an ethernet cable and configure some VLANs with telnet, or plug a flash drive into its proper USB port and make an emergency boot flash drive. I don't want to waggle my fingers in front of it to see the pretty effects and never carry it with me because it costs more than my car. That's what iphones are for.
f382f8 No.102
>>101We could of course implement different versions of the hardware. But sensors like accelerometers are very cheap and extremely easy ti implement and use. They will only add like 0.50-1 dollar of the overall cost.
ab84aa No.125
>>32
On it. I'll just use a 'tag' at the bottom so when I reply back you'll know it's me.
-rv
ab84aa No.126
>>13
I'm going to do some concept images.
ab84aa No.127
>>126
Question, do we want a front camera? I don't think many of us here are into taking selfies really so I'm not sure if it's needed.
14d56c No.130
Does /tech/ know about this? They'd totally be in.
2a0c5c No.150
Has anybody gotten anywhere with this in the last seven months or has everybody given up?
It sounds like a pretty awesome project.
6dd6f9 No.151
So, I was involved in the initial discussion of this, but was more interested in software than hardware and this project never even got that far. (yet, to be fair)
For reasons unrelated, I was reminded of this sort of thing a couple of days ago. I've been working with the LPC1788, a cortex M3 that runs at 120MHz- slower even than a feature phone from 2008 or whenever.
However, the processor has a lot of peripheral features and could run a phone, though it could never run an OS like android. I was contemplating making a dumb phone out of it, based on UClinux with an ncurses interface, where you could perform basic operations like looking up numbers in a phone book, texting, and making calls. It occurred to me that really the only thing necessary to make that into a "smart" phone is a browser. So if netsurf was running on it you'd have an (admittedly terrible) smartphone. The only difference between the smart and dumb phones (based on this processor) is that one has to have enough screen/keyboard to use a browser. This gives a somewhat interesting path of being able to build a dumb phone and upgrade it into something that could be used as a smarphone.
I wonder if this is actually a good idea, it's possible we could base it on something like the i.MX233 which is a little faster (though less RAM I believe). The benefits would be as follows:
-> the processor is slow enough that anyone can build a dev board without worrying about impedance matching or trace length. Easy to do with free software.
-> the parts could be all 0603 / QFP and bigger, which are easy to assemble- anyone with a $50 soldering iron can do it, though maybe not quite as well as people with a PnP and oven.
-> you get quick gratification from building a dumb phone, which you can go ahead and use, and develop programs for it. You don't need to spend years building something good enough to compete with the latest crap from apple/samsung/whoever doesn't suck.
So now instead of suggesting that we build something that would be challenging and time consuming even for a large, experienced team, I'm suggesting that we build something simple, get a simple result, and improve it as we go.
An obvious downside is that it could never compare to a quad core zillion GB of RAM new phone. But if you're trying to compete with a new phone as someone just working with a couple other dudes in your free time it's pretty obvious you'll get nowhere anyways. I suppose if the phone was a big enough success you could always make a much better, more expensive version. Thoughts?