Building a cheaper handheld (moved from the pyra cost thread)


Chip

[Insert Custom Title Here]
Joined
Jun 25, 2003
Messages
3,503
Age
45
Location
NJ, USA
Website
chipandre.com
Moving this discussion from here.

 

It's like saying the analogy is that you want a wooden chair with a cupholder in it and you have to pay thousands to get it hand made. You don't really care about getting a hand made chair, you just want this one modification. And you wish the other companies would make a chair like that because they're already doing the other 90% of the work to make it and they know how to/have access to methods for making cheap chairs.

What if there was a company that bought off-the-shelf chairs and glued off-the-shelf cupholders onto them? Anybody who is paying attention can see how much work is involved in making the pyra. They're engineering this thing from scratch, and that's no small feat. A lot of the cost of the project-as-a-whole comes from designing, testing, prototyping, re-prototyping, and setting up manufacturing for what will end up being a relatively a small run of finished products. What if instead of completely reinventing the wheel, you find something that's already 90% circular and just round off the edges?

However the hell they're doing it, there are any number of no-name chinese factories pumping out decent-spec smartphones for practically nothing. Here's a phone with a 5" 720p display, MT6735 SoC (64bit A53 w/ mali t720), 2GB RAM, and 16GB flash for all of $122.  They've already designed, tested, and tooled up to build the system board.  They've already sourced the components and are manufacturing these in (presumably) decent quantities.  Why can't you just call up Mr. DOOGEE and say "I'd like to place an order for 2000 of your finest Valencia2 Y100 Pro phones please - do you think you can cut me a break on the price if I don't need the frame, glass, battery, back cover, packaging, or accessories?"  I'm not an expert, but I suspect the answer is "Absolutely!"

Now obviously have no idea how conducive this particular SoC would be to a pyra-like project, but out of the hundreds of cheap chinese phones out there, somebody has to be already making something suitable.  Asus and lenovo are both making inexpensive intel atom-based smartphones - if you could rope one of them into this kind of deal, imagine what you could do with an x86 SoC that is well documented and already has linux support.  That's a good chunk of your development and manufacturing work already sorted.  

What about the rest of it though?  I admit you're going to have to make the case and lower PCB from scratch, but there's still room to be clever.  With the brains of the system up in the lid behind the display, your lower PCB will likely only need 2 layers - much easier to design and cheaper to make.  Some parts can still be outsourced.  Instead of making your own QWERTY keymat/keycaps, buy a bunch from whoever makes this thing - if they can sell the whole deal for $27, the keypad can't cost more than a dollar or two.  Maybe you can find a cheap gamepad with decent buttons/dpad and buy those as well.  Anything you can buy pre-made will cost less than designing and making it yourself.

I have to think that a franken-pyra would be cheaper and faster to make than the ground-up approach.  I'm almost afraid to guess how much cheaper, because there are a lot of variables that I can't make educated guesses about.  But certainly it would be cheaper.  If only I had a few hundred grand lying around and some reliable Chinese manufacturing partners...

I know this is getting borderline offtopic, but this concept has been stuck in my head for the last couple weeks.

Now obviously have no idea how conducive this particular SoC would be to a pyra-like project, but out of the hundreds of cheap chinese phones out there, somebody has to be already making something suitable.

You can't 'frankenphone' anything on the market to something that meets Pyra design specs. Just one example: None of them have Dual-SD slots.
Now obviously have no idea how conducive this particular SoC would be to a pyra-like project, but out of the hundreds of cheap chinese phones out there, somebody has to be already making something suitable.

You can't 'frankenphone' anything on the market to something that meets Pyra design specs. Just one example: None of them have Dual-SD slots.
Actually - to meet the Pyra spec, it would need to be tripple SD slots.  But, you have the right idea.

A Raspberry Pi is a pretty small motherboard - not far from cell phone sized.  In order to 'make it portable' using off the shelf components, the resulting franken-portable-pi winds up being the size of a 1980s laptop.  Use of off-the-shelf components that were not designed to fit together both snug & perfect will by default result in a device that is, "larger than it absolutely must be".

Remember that a primary constraint on the Pyra's design is that it MUST fit in a pocket.  The current Pandora and Pyra designs are already at the stretched limit of that primary constraint.  THAT is why everything MUST be custom created.  There are no 'off the shelf' components that perfectly fit the design constraints.

Combine that size constraint with needing two analog controllers, a VERY nice game pad, two shoulder buttons, a full-on PC capable multi-lingual keyboard, high quality speakers, ports out the wazoo, wireless, high resolution screen, clam shell design so it is self-protecting, etc all crammed inside that limited space constraint and you have a Pyra.  A unique design that should not be compared to cheap Chinese cell phones.

The Pyra isn't a phone.  It is a multi-function general purpose clam shell Linux Debian computer with high quality gaming controls, audiophile sound, wireless networking, more connection ports than many current laptops which, by the way, still fits in a front pants pocket.  Anyone who tries to compare the Pyra to a phone simply doesn't get it.
Klapse: Assuming the donor phone has OTG, additional SD slots (and whatever else you want) are easily added by way of a hub controller on the lower PCB.

Grench: I think you're overestimating the size of a smartphone mainboard.  See ifixit's teardown of the HTC one - a phone similar in size and shape to the one I linked above.  It's quite tiny, and unlike a RPi, it is designed from the beginning to be in a mobile device.  The mainboard and display of any modern cellphone would fit easily into the lid of the pandora.  Hell, my sony Z3 - with its 2-day battery and all - is thinner than the pandora's lid.  You would need to have the lid shaped specifically to support the display/board, but volume-wise, there's more than enough room for it.

Yes, the pyra is a general-purpose computer.  So is any modern smartphone.  A phone may have had all its peripherals removed, but it's still a computer.  Everything the pyra has that a phone doesn't have are merely accessories, and those can all be added back.  If you want entirely bespoke game controls, you can have them.  You need to make a custom lower PCB anyway, so do it up however you wish.  Personally I don't think the QWERTY keypad needs to be custom, but it still can be if that's a sticking point.  Simply not having to design, test, and manufacture a 10-layer PCB mainboard will save a heck of a lot of time and money all by itself.
 
Last edited by a moderator:
Chip - please get out a set of calipers and measure the -actual- thickness of the Pandora's lid component from the surface of the screen to the back of the lid.  Then measure your smart phone.  Keep in mind that the smart phone's thickness is NOT dictated by the battery, but by the camera(s).

The 'base' part of your conceptual device could not actually be any thinner than the existing Pandora.  Slots, battery, keyboard, etc - all need to live there still.

Even making your very basic assumptions, the device you're describing would be -larger- than the Pandora or Pyra in the most critical 'thickness' dimension.
 
Klapse: Assuming the donor phone has OTG, additional SD slots (and whatever else you want) are easily added by way of a hub controller on the lower PCB.
I don't want to have to trade in my OTG port for basic functionality though.

-Neelix
 
I've looked at a few SoM - including the Pi Compute module. I'm not sure if you could get it quite as small as a Pandora/Pyra, but I reckon you could get it close.

That said, I don't consider the Pandora "pocketable" - unless I'm talking about my coat pocket - It's too big to fit in my trouser pocket comfortably whilst letting me stand and sit. So I carry it around in my rucksack, and as such, it doesn't matter if it's actually a bit bigger then.

I still might try and consider building something, even if it's not "quite" a Pyra. I was thinking of using this SoM. They sell displays too, though they don't particularly shout out
 
Klapse: Assuming the donor phone has OTG, additional SD slots (and whatever else you want) are easily added by way of a hub controller on the lower PCB.
I don't want to have to trade in my OTG port for basic functionality though.

-Neelix
You wouldn't have to.  Just wire one of the ports on the hub controller to a physical USB port (full size or micro) and you're good to go.  4-port hub chips are small and cheap, and that would give you one port to connect to the mainboard, two for SD slots, and one for a physical USB port.  If that's not enough, you can step up to a 7-port hub.  

It's possible there are additional SDIO lanes in the SoC that may or may not be exposed on the mainboard, but honestly, soldering stuff onto the board is probably more effort than it's worth.
 
Chip - please get out a set of calipers and measure the -actual- thickness of the Pandora's lid component from the surface of the screen to the back of the lid.  Then measure your smart phone.  Keep in mind that the smart phone's thickness is NOT dictated by the battery, but by the camera(s).

The 'base' part of your conceptual device could not actually be any thinner than the existing Pandora.  Slots, battery, keyboard, etc - all need to live there still.

Even making your very basic assumptions, the device you're describing would be -larger- than the Pandora or Pyra in the most critical 'thickness' dimension.

After half an hour of looking, I cannot find my caliper.  I attempted to disassemble the pandora's lid to measure the display thickness, but whatever tape they used on it has solidified over the years and will not come off without a fight.  I got one corner off, but I can't get any more off without likely breaking the bezel - something I'm not willing to do to win an internet argument.

Using the totally inaccurate method of simply pinching the pandora and phone between my fingers, they are pretty damn close.  I have no doubt whatsoever that if I felt like destroying a grand worth of gadgets, I could fit the display and mainboard of my phone in the pandora's lid.

And if I want to use it in device mode while also having access to the SD cards and USB ports? 

-Neelix

Then you'd be out of luck.  Though it's a stretch to envision a situation where you'd actually need to do that.  Does the pandora even have a device mode?
 
Then you'd be out of luck.  Though it's a stretch to envision a situation where you'd actually need to do that.  Does the pandora even have a device mode?

It has several actually, as you can at least theoretically load any of the linux gadget drivers.  (though I've yet to get Gadget Audio working properly)

I normally use the USB networking feature myself.   It's also possible to use it as SD card reader.  I had to do that recently at my sister's house when I wanted to use an SDXC card on her machine and discovered and she didn't have a reader for it.  I also seem to recall someone setting it up as a virtual CD drive,  I think they even made a PND to do just that.

Yeah I use a mouse often with the OTG port using this adapter: http://www.amazon.com/gp/product/B00JK1OFOM

That adapter is for using it in host mode, not device mode.

-Neelix
 
That adapter is for using it in host mode, not device mode.

-Neelix
Maybe someday I will get my USB modes straight... so Device mode would essentially be when we use the Pandora as an SD card reader right?

EDIT: NVM you said just that in your post...
 
Last edited by a moderator:
Man I would love a cheaper pyra alternative, I love the idea behind this handheld and the community and everything, I actually just started my first year in University for Computer Science, so I'd think it'd be a fun device to mess with, I can't burn $600+ on a Pyra though..
 
I always wonder how there can be so much  cheap and masses of electronics "made  in china", even with comparable decent quality. They  can do much much better than they did with  the first Pandora for example (had china-releated issues). I recently strumbled over the  german "Pearl" catalogue, tons of  cheap china gadgets, gazillion ones, some even look good. So if they can throw out smartphones, computers, cameras in lightning speed, why is it not possible to make an device "like the Pyra" that is cheaper and still  has good or even better quality? 

Today I was in the german "Saturn" Electronics Market, damn, I must admit these sleak full-metal-unibody "Ultrabooks" are very interesting, I may get one of these (no Apple of course :)   ) I  imagine a Pyra with such a sleak case (just in small), still 10h battery time and still cheaper.  I'm sure it  can be done, I'm not sure how.  :ph34r:

Or the other way round, how about an rugged outdoor-like handheld for the Adventurers of  us?  Basicly a Pyra in an tough case or so. I always liked these military designs.
 
Last edited by a moderator:
Two things occur to me in this discussion.

We have several people who do not understand why the Pandora and future Pyra are and must be built the way they are.  So, they believe that there is an other, cheaper, better way to get to the same end.

Much like the, "Pyra needs a camera!!!" crowd, The belief is not one based in rational facts, but more of an emotional basis.

The "knock off" electronics industry in China is just that - they are unlikely to produce a novel/new device category.  If the Pyra is a resounding success, then they might make a similar product 6-18 months later.
 
We have several people who do not understand why the Pandora and future Pyra are and must be built the way they are.  So, they believe that there is an other, cheaper, better way to get to the same end.

Possibly. I also think there's also probably a few people that think there's another, cheaper way to get to a similar end.
 
I would not say that there is a "better" way, but maybe a more efficient way to get the same results at less costs. Of course, this is just plain theory and nobody can prove it. It may be possible if there is a much wider audience for the Pyra, who knows. But at the moment, yes the way it is is the only one right now.  The Pyra development is already much more smooth than the Pandora one ever was. ED & Co. learned alot from the Pandora, which is already a huge benefit.
 
Much like the, "Pyra needs a camera!!!" crowd, The belief is not one based in rational facts, but more of an emotional basis.

The "knock off" electronics industry in China is just that - they are unlikely to produce a novel/new device category.  If the Pyra is a resounding success, then they might make a similar product 6-18 months later.

There is no "emotional bias" here.  I have no skin in this game.  The pandora/pyra are being designed and built the way they are not because it's the only possible way to build this type of device, but because it's the only way to build that exact device.  Dragonbox et al. have a very specific vision of what they want, and the only way to get exactly that is to do the whole thing from scratch. It's a slow and expensive process, but for them and for those that share their vision, it's worth the time and money.

What I'm trying to say is that if you weren't quite as particular about the exact specifications, you could build a very similar device faster and cheaper by leveraging existing cheap Chinese products.

I like car analogies, so here's one:  The pyra is a tesla and the frankenpalmtop is a rally fighter.  Both are niche automobiles from small startups.  One is a totally bespoke vehicle built from the ground up based on a lot of custom engineering, and the other is cobbled together from various bits and pieces that happened to be convenient.  Rather than engineer and manufacture a powertrain from scratch, local motors just uses a 6.4l crate motor and the transmission out of a yukon.  The taillights are off of a honda civic and the door handles are off a miata.  The frame and most of the bodywork is custom, but anywhere that they could use an off-the-shelf component, they did.

It's not a perfect analogy - a model S is actually cheaper to buy than a rally fighter, and the two cars couldn't possibly be more different in form or function - but I hope you get the core concept.  I can't be bothered to find the R&D and manufacturing startup costs for a tesla vs a rally fighter (if those numbers are even public), but I'm absolutely certain there's at least an order of magnitude between them.  Probably more than one.
 
If the objective is a high quality device tailored to the exact needs of a niche community market, ED's path to the Pyra is spot on.

If the objective is a race to the bottom in quality and costs with no realistic hope of community or support, Chip's path might be a quicker means to the end.

Most people use cheapo cell phones for 1-2 years max.  I plan to have my Pyra for @4-5 years or so - or until ED can make a replacement that is enough of a 'step up' to justify another purchase.

I bought a CC Pandora on pre-order and 'only' waited 2.5 years to get it.  It was fantastic.

The rebirth Pandora didn't strike me as enough of a step up to justify another purchase so soon.

However, the 1Ghz Pandora DID meet that level - so I was one of the early purchasers on that one.

My expectations bar is very high for the Pyra.  It is NOT just a casual game console to me.  The Pandora is a general use pocketable computer that happens to have nice gamer-quality controls and audiophile sound.  I'm not interested in a franken-computer a few inches bigger with slapped together mismatched components, poor battery life and Android OS.
 
My expectations bar is very high for the Pyra.  It is NOT just a casual game console to me.  

Now who has an emotional bias?  ;)

Please don't misinterpret this thread as an attack on the pyra or how it is being developed.  I'm not saying that my way is "right" and everybody else is "wrong", I'm just saying there might be different ways to go about making different devices.  I'm not in any position to actually make the sort of device I'm suggesting.  I don't have the money or the expertise to even get started.  I'm just throwing the idea out there to start a conversation.  It absolutely is possible, to some extent, and I think it bears discussion.

Obviously you're one of the people for who the pyra is ideal and nothing less is acceptable.  That's good - it's a noble project and needs supporters.  But you have to admit that maybe not everybody who wants a pyra-like device wants (or can afford ) to pay a pyra-like price for it.  Certainly there are people who would be willing to trade 10% of functionality for a 50% reduction in cost.  Clearly you wouldn't, but I would.
 
Issue with a franken device is you dont know the exact hardware. For repairs and service the pandora is top notch. A franken device GL.. After awhile they know the common failures with a device. The pandora was just like that.. they knew what was wrong right away and could fix it. 

Making your hardware is expensive but the results are beyond worth it IMO. The quality surpases most chinese crap, and the tailor fit os is a huge step up over andriod, which I do not like.

Ive used the pandora enough to justify the price of buying 3 of them...  2 1ghz and a CC the quality and my specific use are hard to replicate on phone hardware. I type up stories and the pandora is the only system I consider using.  Its far superior to everything esle. ive seen community backed and runs on a proper OS... theres no libre office for android,  and the apps are weak at best if freeware.
 
If the objective is a high quality device tailored to the exact needs of a niche community market, ED's path to the Pyra is spot on.
All of whom have slightly different needs - hence the bazillion keyboard threads...

If the objective is a race to the bottom in quality and costs with no realistic hope of community or support, Chip's path might be a quicker means to the end.
Why does it have to be a "Race to the bottom"? You don't have to go for all the cheapest kit, just use some that make things cheaper.

I'm not interested in a franken-computer a few inches bigger with slapped together mismatched components, poor battery life and Android OS.

That's some baseless assumptions though.

Issue with a franken device is you dont know the exact hardware. For repairs and service the pandora is top notch. A franken device GL.. After awhile they know the common failures with a device. The pandora was just like that.. they knew what was wrong right away and could fix it.
Why don't you know the exact hardware? You'll be picking your source from somewhere, you'll get tech sheets.

Just take the example of switching ED's custom CPU board with a SODIMM slot SoM. You could still keep all your peripherals, just wired up onto a different connector. You'd only have to design the main board, rather than the main board and the CPU board - this would save you some cost. ED said he could only get OMAP SoCs, but with SODIMMs, he could have picked a Pi Compute, or several other ARM SoMs - including a Tegra 3. So he would have had more choice.
 
Back
Top