Embedded Systems Engineer's Dream - "PhoneBloks" The 100% Customizable Phone That LASTS - Polidicks

It's a very nice idea, but it's impossible to have anything worth mentioning in this setup.

I'd imagine having one of these will be a nightmare. Every time you drop it you'll have to pick up 20 lego-sized pieces from underneath all of your furniture.

I don't think people throw phones away for breaking so much as they do for just being an old model.

Wait, wouldn't all of those blocks have to be individually packaged

I don't know of anyone who actually throws away a phone, they just sell it on.

I haven't changed phone in a good 6 years. Hell I know this because my phone resets back to 2009 when it was activated occasionally when it shorts. I have no reason to upgrade so I never have. I used it for texting and calling people and that's all I've ever needed it for.

From electronics point of view, this is complete bullshit. The way parts are swapable in a PC is because specific components are made and optimized for different sockets/connectors. It's down right impossible to have a universal connection that handles all sorts of components anywhere close to acceptable speeds at this size. This concept comes down to "We have this cool connector thingy and it makes the things work! Magic!" The reason why phones can be so small is because there are no connectors between all the indidual bits and pieces; its all soldered together tightly. I embrace the concept, and it could do miracles for the phone market, transforming it into the model we see with PCs; inter-exchangeable components and software. But we are not there yet technology wise. Perhaps a middle ground where certain components conforming to socket-types can be upgraded is possible, but I find it highly unlikely; unless you want to go back to 1995-style box phones in terms of size. Hell, whats the motivation to give this choice to consumers, when the current business model is making them money by the truckload? This is a design concept, pure and simple, with no possibility of becoming an actual product. No actual hardware design or even fact-checking was involved here.

That's all cool but where do you actually insert the SIM card?

the people who say "This is impossible" "This is stupid business wise" no it isn't, you just need cooperation to make the CPU's easily similar to each other but better than the last, and business wise why is it? Many people beg for customization.

[QUOTE=Splash Attack;42161490]I can understand swapping out the camera for another camera or some other identical component in its place, but once you go rearranging shit like that, the idea doesn't hold up. [B]That's just not how electronics work.[/B][/QUOTE] Don't post if you don't know what you're talking about It could easily work, you just need a standardized interface. [editline]12th September 2013[/editline] [QUOTE=Krahn;42163939] It's down right impossible to have a universal connection that handles all sorts of components anywhere close to acceptable speeds at this size.[/QUOTE] [quote] No actual hardware design or even fact-checking was involved here.[/quote] So please provide a source for your little statement then It's all about using a standardized communication protocol. This would work exactly like it does on a motherboard, except you'd need a layer on top which identifies where each component is connected on every startup, or something of that nature.

Impossible? No, nothing is impossible. Impractical? Pfft yeah.

This is definitely innovative And god would I love to have a phone with a feature like that [editline]12th September 2013[/editline] [QUOTE=S31-Syntax;42164247]Impossible? No, nothing is impossible. Impractical? Pfft yeah.[/QUOTE] If everything is connected with screws, it seems pretty practical. I have features in my phone I would love to be able to take a way to improve others, at times.

[QUOTE=Jad Hinto;42161461]Don't exactly care for how plain it looks, but it's a more than worthy sacrifice for that much on the fly customization.[/QUOTE] you're implying it's any more than an (impossible) concept

Okay, say you succeed in making the common, modular interface required for this. How do you secure the parts to the baseplate? They mentioned a couple of screws in the bottom. Well that doesn't tell me anything, and there is no way that those teeny screws in the bottom can secure the blocks at the top [editline]12th September 2013[/editline] The only way I can think of to combat the constantly changing connection and pin standards is to overdesign the interface from the start. Make the interface far more capable than the things we're attaching to it. I'm doodling something right now that could maybe work. Don't expect much though, I'm unfamiliar with the back end requirements of something like this. Regardless, in order for blocks to scale, the interface needs to be tiled. I'll post pics when I'm done.

[QUOTE=aydin690;42161896]EE here. The amount of bandwidth flowing between different components and frequencies involved on a pcb and the sheer number of connections just screams integration. Not to mention the amount of engineering that goes into things like thermal management, antennas, signal integrity, emi compliance, and power management. It may be doable but I don't think its worth doing. [editline]11th September 2013[/editline] LOL, more like any engineer's nightmare.[/QUOTE] I'll admit crosstalk and thermal management will be a pain.. Also, I apologize for posting this in I guess the wrong subforum, excitement got the best of me.

[QUOTE=demoguy08;42164213]Don't post if you don't know what you're talking about It could easily work, you just need a standardized interface. [editline]12th September 2013[/editline] So please provide a source for your little statement then It's all about using a standardized communication protocol. This would work exactly like it does on a motherboard, except you'd need a layer on top which identifies where each component is connected on every startup, or something of that nature.[/QUOTE] Easy there dude... I said it was impossible in its current concept. If it used specialized slots for each type of component (Slot for the ARM CPU, isolated circuit for the modem to prevent noise on the antenna, that sort of thing) then yeah, it might be possible. But that requires connectors that take up space, which makes your phone much bulkier. Look at the electronic circuit they show. Those connectors make no sense, nor does the shown circuitry. If an actual hardware designer was involved, then they either chose some artistic license, or simply don't have a solution to the problem, or don't want to show it (yet). A big copper nub like that is gonna pick up a hell of a lot of interference. And with just a few of those, you are never gonna have enough bandwidth to do anything meaningful, let alone connect a CPU. Want a source? how about you take a look at an ARM chip or your PC's CPU socket. See how many connectors it has? That's the basics; each connector can send a high or low signal (0 or 1), more connections means more throughput. A higher frequency means you can send signals faster, but too high a frequency and you get interference, missed signals and a whole bunch of issues. This is a way simplified explanation, but I think you get the point If each of those connectors was more than what they show; like an actual socket, it might work. But they don't show that simply because no connection exists that does everything you need to support all different hardware components at the same time. If you can find a connector or component that can do this, please link it and I'll gladly eat my words.

From a logical standpoint I understand just how ridiculous and improbable this is without some sort of new fangled technological advancement that would allow this to work. But I've got to say, the design and idea of it is fucking awesome. If we could make this work some time in the future I'd totally be down to try it out.

Seems like a lot of people are taking this guy's use of 'CPU' a bit too literally, my impression was that he meant CPU, GPU and RAM would be on a single PCB, which is doable. Inefficient as hell, but doable.

Nice idea, but it gets completely cockblocked by the problem of connecting the modules to the main board. [editline]12th September 2013[/editline] Yeah yeah someone will say "BUT ITS NOT IMPOSSIBLE!" yeah well invisible cars aren't impossible either, just extremely expensive and/or impractical for what they offer. Sure someone could probably throw a few hundred millions into developing this under 5 years with an entire team of engineers, programmers, designers and whatnot, but you could just chuck the same effort and money into a normal phone design and get a much better result.

irl ideas guy

Okay, here's what I came up with. [t]http://i.imgur.com/xYNDpZI.png[/t] This first one is the smallest size. I scaled it to 1"x1", and I know the actual thing would be smaller, but my autocad was in inches and I couldn't be arsed to change it. This being the smallest size, there are still a shitton of pins to choose from. This is good though because there are plenty of pins that developers can use and still have room for growth. There is a peg to limit the orientation, but thats for consistency. [t]http://i.imgur.com/4rekEwC.png[/t] This one is a size up, a 2x2. There are now 4 times as many pins to choose from, yet it will still fit in the same shape setup because it is literally tiled. Because there are more pins, bigger and more complex things can go in it. I'm not expecting all the pins to get used, in fact there are this many pins so that they have room to play with. The hard part past this though is getting the backboard to be adaptive. It'd have to be intelligent, with its own firmware. And the phone's power switch will need to be here. Because of the abundance of pins, you can afford to reserve a couple of identical pins across all designs for power. With dedicated pins for power, you can easily have power routed to the backboard without anything having to figure out where it needs to go. Once the backboard is powered, it can boot up with its own firmware and start discovering the other blocks installed and routing them as needed. With such a massive abundance of pins and I would assume pathways, there [I]should[/I] be plenty of room to route stuff without conflicts. After staring at it for 2+ hours now, it honestly looks like this is indeed quite possible. I can't even begin to know how the backend/firmware would work, but on a conceptual level and a hardware level, this should be doable.

[QUOTE=acds;42165547]Nice idea, but it gets completely cockblocked by the problem of connecting the modules to the main board. [editline]12th September 2013[/editline] Yeah yeah someone will say "BUT ITS NOT IMPOSSIBLE!" yeah well invisible cars aren't impossible either, just extremely expensive and/or impractical for what they offer. Sure someone could probably throw a few hundred millions into developing this under 5 years with an entire team of engineers, programmers, designers and whatnot, but you could just chuck the same effort and money into a normal phone design and get a much better result.[/QUOTE] I think I would have good use for it. Sometimes I need more batterypower and not so many features.

[QUOTE=Coffee;42163852]I don't know of anyone who actually throws away a phone, they just sell it on.[/QUOTE] Exactly This video assumes that everybody literally throws away their phones after 1-2 years of use because they get outdated or somehow unusable. That's incredibly stupid and naive. Who seriously does that? A even if you have to sell it for $25-$50 to get someone to buy it, that's still better than throwing it out. Almost everyone sells it or re-introduces the phone into the system at a lower cost to someone who doesn't mind having the last generation model, or they just keep it for pretty much forever. Especially since it allows them to front the cost for their next device they buy. If you trade in your phone, the company will almost always refurbish it or take it down for parts. This is like complaining that when people are done with cars they buy new they just throw them out. Uh, no they don't. Besides, how does making the phone modular somehow "solve" this problem even if it actually existed? Lets say you upgrade to a new screen. What are you going to do with the old one? Its much less sellable on the market than an entire phone so in all likelyhood, it'll just get thrown out. Especially for parts like batteries. And the #1 reason why this is silly, is because it assumes people go through phones like junk food because they have to. The reality is, they go through phones because they want to. People buy the new iPhone every year or every other year because its the newest phone, with the newest style even though its only marginally improved over the previous one most of the time. Nobody's gonna get a modular phone and have it last 10 years unless they would have gotten any old phone that lasts 10 years anyways, because by the time a few years pass they'll want the newest phone anyways. Most people don't care about specs, they care how the user experience and stuff is.

I like the idea of this, especially the incentive to help recycle and cut down on e-waste.

Such a bullshit idea. A cool idea. But its bullshit. Never going to happen.

Having different slots for different items (cpu, gpu...) would defeat the purpose of this concept. But the concept seems to have different sized tiles which means you'd need to play a more basic version of tetris to stack the blocks together without holes.

[QUOTE=Supacasey;42162688]Straight up impossible, sorry.[/QUOTE] Do you guys have any electrical engineering background at all? Because there are very good solutions for modular pin configurations. Google Cypress PSoC for one of many examples.

[QUOTE=Krahn;42163939]From electronics point of view, this is complete bullshit. The way parts are swapable in a PC is because specific components are made and optimized for different sockets/connectors. It's down right impossible to have a universal connection that handles all sorts of components anywhere close to acceptable speeds at this size. This concept comes down to "We have this cool connector thingy and it makes the things work! Magic!" The reason why phones can be so small is because there are no connectors between all the indidual bits and pieces; its all soldered together tightly. I embrace the concept, and it could do miracles for the phone market, transforming it into the model we see with PCs; inter-exchangeable components and software. But we are not there yet technology wise. Perhaps a middle ground where certain components conforming to socket-types can be upgraded is possible, but I find it highly unlikely; unless you want to go back to 1995-style box phones in terms of size. Hell, whats the motivation to give this choice to consumers, when the current business model is making them money by the truckload? This is a design concept, pure and simple, with no possibility of becoming an actual product. No actual hardware design or even fact-checking was involved here.[/QUOTE] And I'm sure back in the 1980s, if you told someone you'd be able to carry a computer with wireless internet/media device/phone/camera all in one device the size of a credit card, they would have told you it was impossible too.

-snip-

This is another KONY 2012 except less public masturbation.