***Building an Android Phone***

[mumera]

Hi Guys,

I am a complete newbie, but very ambitious none the less. I have yet not completed any projects but a want to build an Android Phone.

The first step, I guess as with any project should be to build the first prototype.

/***********/
/*PROTOTYPE */
/***********/

From browsing the various projects on the site, I am thinking (please some one correct me) that we will need to combine the following projects to get this to work:

1. A Linux based SBC (Single Board Computer) -> to Run Android -> Beagle Bone
2. A Touch Screen -> for I/O -> Beagle Touch
3. Battery -> Beagle Juice
4. A GSM Module -> for the Phone Functionality -> Something like the ADH8066

Nice to haves would be:

1. 3 Axis Accelerometer
2. A Camera

What do you guys this..... does this make sense even...... Please HELP!!!!..... or this just plain crazy

[cstratton]

That's a huge project, and would result in a stack of boards much larger and costlier than a phone.

If you want to do it, maybe start with just the sbc and the touchscreen and make some progress on the software stack before adding the expense of the GSM modem and battery system.

Another option would be to find a root-able android device, which would let you hack on the software at about the same level, ie system stuff but not the firmware of the radio processor. Or port android to a smartphone that shipped running something else. On the plus side, these are relatively less expensive paths and the results are physically compact; on the downside the hardware documentation available to you for the chips in consumer phones may be more limited than it is for modules marketed to integrators.

[TheFallen]

Right, using a Beagle Bone presents some difficulty. You will need design <3 boards. A screen board as the Beagle Touch is designed for normal Beagle boards. A battery board as the Beagle Juice is again designed for normal Beagle boards. And finally a GSM module for the Beagle Bone.

For you to DIY this then it will be about $1000 (parts + time, minimum) and be about an inch thick and a battery life of 5 minutes, maybe a couple of hours.

I must admit you could definitely get in on the ground floor with some Beagle Bone shields, sorry, capes, at the moment. Probably best to focus on a screen board first.

[mumera]

Hi Guys,

Thank you so much for your response.

@ cstratton
Well this is a first prototype, I am not sure how to approach this honestly. Why I am asking you guys for help. Regarding the suggestions, you made

a.
"start with just the sbc and the touchscreen and make some progress on the software stack before adding the expense of the GSM modem and battery system"
Well I was thinking the exact opposite of using a GSM Module and connecting it with my PC and using an Android emulator i dont know if or how that would even work but sounded like a good place to start. What do you think.

b. Regarding your second option, I agree it would be very hard to manipulate and program the chips its hard enough doing the ones which have proper documentation.

@ TheFallen

Point taken maybe it would be a better idea to move onto a normal beagle board. What exactly do you mean by beagle bone shields. Do you think though it can be done. What would you suggest as a good starting point.

Regarding the battery life, I thought the typical beagle juice 4500 mAh battery gave around 6.5 hour of battery life, wouldn't it be pretty much same for us to.

[cstratton]

Well I was thinking the exact opposite of using a GSM Module and connecting it with my PC and using an Android emulator i dont know if or how that would even work but sounded like a good place to start. What do you think.

This might get you an awkward way to make phone calls or use a mobile data connection, but it would be a fairly small and un-representative part of the task of making an android phone. Assuming you are buying a module and not putting bare RF chips on your own board, most of the work has little to do with telephony, and a lot to do with getting the software and drivers running on your board and talking to all of your other peripherals, dealing with power management, etc. By using the emulator, you miss diving into that core challenge - of the peripherals the GSM module is actually pretty simple and standardized in that it's likely a serial port plus an analog audio channel.

Something you can do at basically no expense is to download the Android sources (either AOSP or a popular alternative ROM, which may be more complete), examine them to get an idea of what you are up against, and go through the process of building and installing them on an existing phone.

[mumera]

@cstratton
ummmm i'm not too worried about the "getting the software and drivers running on your board and talking to all of your other peripherals, dealing with power management, etc." because its already been done, which is why I chose beagle board for this project. I'm not exactly sure if you've seen this http://www.youtube.com/watch?v=TLYmSSzWjI0, does what I'm saying make sense now or am I missing something here. See the Beagle Board gadget already has android running on it along with power management, touch screen and wifi. ( I got my idea after seeing these videos really, that if android is already running on the beagle board modules why not just connect a GSM module to it n make it a complete phone)

So if this is all done should'nt I be thinking about, how I will get android to interact with the GSM module.

[cstratton]

I'm aware that android runs on some beagleboard versions, yes. But if you are using someone else's port of android already supporting your chosen modules, I'm not sure quite where the project "value" is, compared to starting with something that already has the GSM integrated and is cheaper and more compact, ie, phone known to be flashable? I understand the idea of "did it myself" value, but it seems like the original part you would be doing is the part where doing it yourself implies a large, clunky, overcostly result.