Help Needed Replicating Success Of Openpandora For A Netbook/laptop

[lkcl]

I have also been looking for a cheap SATA solution for an open hardware project (portable music player). Not really for disk bandwidth, just to get access to higher capacities. I was considering a PATA-SATA bridge such as JM20330, which is used in cheap adapters found on ebay. I've no idea if this chip can be easily sourced, I was thinking of buying the adapter and hacking it.

someone else on another list recommended this, from genesys logic: http://www.genesyslogic.com/_en/product_01_1.php?id=18 - you might find it's easier (and cheaper, for sure) to use USB-to-SATA or maybe (look on their site there are other products) USB-to-ATA/ATAPI

l.

 

[Polossatik]

It seems to me that most small harddrives - the kind people would be most interested in for a netbook - are still PATA so they don't need additional support chips when used with these SoCs.

If I look at local supplier then PATA 2.5" disks are on the way out, the max I find is a 320GB @ 0.25 euro / GB.

A 640GB SATA 2.5" disk is 0.12 euro / GB, a 320 GB SATA 2.5" disk is 0.15 euro / GB.

comparing to SSD, the cheapest SATA SSD is a 128GB @ 1.56 euro / GB, cheapest PCI-e SSD is a 512GB monster @ 3.28 euro / GB

 

[bobc]

I have also been looking for a cheap SATA solution for an open hardware project (portable music player). Not really for disk bandwidth, just to get access to higher capacities. I was considering a PATA-SATA bridge such as JM20330, which is used in cheap adapters found on ebay. I've no idea if this chip can be easily sourced, I was thinking of buying the adapter and hacking it.

someone else on another list recommended this, from genesys logic: http://www.genesyslogic.com/_en/product_01_1.php?id=18 - you might find it's easier (and cheaper, for sure) to use USB-to-SATA or maybe (look on their site there are other products) USB-to-ATA/ATAPI

Wow, you sure like to complicate things LOL. Even if I did have OTG support, PATA-SATA is certainly easier than going via USB.

 

[lkcl]

Do you have any price and availability information on S5PC110? It'd be very attractive compared to any of the other Cortex-A8 offerings out.

this is all i've been able to find so far:
http://www.ebv.com/en/products/categories/details/product/s5pc100.html

go to the "pricing" tab. it keeps changing slightly: they're now at $EUR 14.20 which comes out at around $USD 19 (hmmm, and i thought their prices were in $USD whoops, glad i checked it again). interesting. appx $USD 19 puts it closer to the OMAP3503 600mhz price ($USD 21, 1k), except the S5PC100 is 833mhz.

i'm pressing for more info from people i know.

 

[lkcl]

I don't think the idea to implement an SATA controller in a cheap microcontroller is workable. From end to end it just doesn't seem to work at several points along the way.

[....]

Of course I could be totally wrong, so I guess I'd like to know what Michelle is thinking exactly.

*click*. my mistake. michelle actually said "PATA" and i thought that it was an advancement on SATA, whereas, duhh, of course, PATA is the old, familiar 40-pin ATA interface. that's why she said that the marvell 8051-compatible CPU, the DS80C411, with oodles of GPIO pins could be used (not ATA-6 but at least ... something). i did think she was off her head for a minute, thinking that a 75mhz CPU would be able to do 1.5gbit/sec or 3.0gbit/sec serial but no, it was just me

i took a look at the datasheet, and it looks to me like the DS80C411 instructions take multiple clock cycles (i didn't investigate thoroughly to see if they're pipelined). what you think, exophase? i quite like the idea of saying "sod the off-the-shelf components"

l.

 

[Exophase]

this is all i've been able to find so far:
http://www.ebv.com/e...ct/s5pc100.html

go to the "pricing" tab. it keeps changing slightly: they're now at $EUR 14.20 which comes out at around $USD 19 (hmmm, and i thought their prices were in $USD whoops, glad i checked it again). interesting. appx $USD 19 puts it closer to the OMAP3503 600mhz price ($USD 21, 1k), except the S5PC100 is 833mhz.

i'm pressing for more info from people i know.

Thanks, but I believed you already on the S5PC100 price (okay, even though you were slightly off anyway), I was really curious about the S5PC110 since you mentioned it. I wouldn't call it a full generation ahead of the rest, but the difference between an SGX540 and the 3D in the S5PC100 or i.MX51 would be staggering. A Cortex-A8 at 1GHz with 512KB of L2 cache certainly doesn't hurt either.

I'd definitely pick an S5PC100 over an OMAP3503, it seems better in every way.

*click*. my mistake. michelle actually said "PATA" and i thought that it was an advancement on SATA, whereas, duhh, of course, PATA is the old, familiar 40-pin ATA interface. that's why she said that the marvell 8051-compatible CPU, the DS80C411, with oodles of GPIO pins could be used (not ATA-6 but at least ... something). i did think she was off her head for a minute, thinking that a 75mhz CPU would be able to do 1.5gbit/sec or 3.0gbit/sec serial but no, it was just me

i took a look at the datasheet, and it looks to me like the DS80C411 instructions take multiple clock cycles (i didn't investigate thoroughly to see if they're pipelined). what you think, exophase? i quite like the idea of saying "sod the off-the-shelf components"

Assuming the SoC's with PATA aren't what's wanted for some reason, then the only way any solution will be more desirable than a USB2 bridge is if it's faster. And the only way a microcontroller can possibly let you pick off data fast enough would be to go through an externally mastered parallel bus that the host processor can pick data straight off of as if it were memory. The only cheap/small microcontroller I know of that has this capability is PIC24. It takes data from a FIFO, not DMA, so it still has to be fed constantly, which can be too much for it if it has to do anything on that data (including just getting it from the harddrive). PIO probably has to be used for that, and I just don't see it working out to being faster than USB2.

At any rate, this 8051 seems like a bad fit, given that it costs $13+ and isn't very well suited for this application (you want fast 16-bit buses, not multiplexed 8-bit buses).. I think you're paying a lot for ethernet and CAN bus controllers you wouldn't end up using. A faster Cortex-M3 or PIC24 can be had for substantially less. An FPGA (or possibly even CPLD) would probably be the best fit, if you had to use something custom for this.

 

[lkcl]

http://focus.ti.com/docs/prod/folders/print/am1808.html

... don't freak out - the idea of having a 375mhz ARM9 computer, i mean.. it'd be great an low-cost 'an'all... but that's _not_ why i'm looking at it: i'm looking at it because it's only approx $USD 9 in 1k volumes and it has a 3gb/sec SATA controller _and_ a 16-bit HPI (host port interface) which can operate asynchronously at the host's clock rate. it also has 64k on-board ROM and 128k on-board RAM. that means that it would be possible to connect the AM1808 and _just_ the AM1808 directly to the main CPU and use it purely for its peripherals. thanks to it being an ARM9 CPU, standard gcc could be used to write the drivers.

i really haven't seen a cheaper way to get an extra high-speed USB2 port, an extra 10/100 ethernet, a 3gbit/sec SATA and two extra SD/MMC slots. writing the linux kernel drivers for the host system is going to freak out the linux developers just a tad, though

oh, for goodness sake, it has 144 GPIO pins as well.

 

[Exophase]

Yeah, HPI is the kind of bus I was describing, and it's very useful. One of the cool things about HPI on TI chips is you can also boot over it, so you don't need to tie flash (or any external memory at all) to the chip, and its firmware can be changed just by changing a file on your host filesystem.

I do think you could do a little cheaper with something else, and this might draw more power (probably pretty negligible), but it's a pretty cool SATA solution.

With HPI you don't even have to have anything running on the ARM9 in order to access the chip's peripherals. You can basically merge its address space with the host SoC's. The only difference is that you have to load addresses indirectly to a fixed location, since the HPI bus is multiplexed. If there are available drivers they can be ported to this architecture without too much effort, although rigging DMA is going to be tough.

 

[bobc]

Did the survey of open netbook/laptop requirements ever get done? And if so, what were the results?

 

[Exophase]

We're seeing S5PC110 in non-Samsung devices now (Odroid-T, Odroid-S.. still Korean though), so if this does get off the ground in the nearish future it'd be nice to see this SoC in it. S5PC110 should be very similar to Apple's A4 chip.

 

[Kangal]

To the OP,
The correct answer is to find what people need, then find the way they want those needs full-filled.
You need to structure your research into facts, then into ideas, then a concept.

I think your going to find great difficulty in the netbook field due to the competition.
There are many great offerings both from China and from Retail, and at affordable prices.
I think if you try to do what everyone else is doing it will fail.
But if you do something too different, it may be counter-intuitive.
I think your best bet is to go with a tablet and design it that an accessory doubles it as a netbook. '10/'11 tis the year of tablet craze.
See what others have: iPad, VAIO-P, Malata SMB-A1011, ASUS EePad, ExoPC.
Go big or go home! :lol:

Well here's the process, your welcome

1) The first big question is what will it be used for:


2) Next question is what are its features:
What does it need to have, what are things that are expected but not vital, what are things consumers look for.


3) Next big question is design:
How will you design something that does those and appeals to the market.


4) Next question is what platform will be used:
Will it be widely adopted, versatile, cross-compatible, intuitive, secure?
What OS would best suit your needs?

5) And now you worry about the processor:
The equation is Performance/(Battery x Price), now go do the maths which has the highest value for your needs?

6) Now that you have a pretty clear cut idea, its off to the drawing table.
Make a sketch, think about how it would be used... (eg position of USB-input, will it hinder the way you hold it once plugged etc etc)

7) Get feedback, more research about the market and how the device would be built (blueprints)... etc etc

8) This stage is when your serious, sell your idea OR get a partner OR fund it yourself and continue to processing

9) Beyond these are the manufactoring and business aspects (marketing, wholesaling, advertising) of your device [the real challenge]

 

[Kangal]

I will help you out in these processes with my input.
Enjoy and do send some feedback. I was thinking if this device was made by standard materials (plastic, glass, aluminium) and was not marketed/advertised there would be significant cost reductions able to make this somewhere between cheap netbooks and the iPad.

#1
It will be your home entertainment unit.
It will spend its idle time on the coffee table infront of the television in the living room.
When not used it will be an electronic picture frame
It will be:
your Family organizer/schedule
the thing you write notes in for your family (Honey food in oven etc)
for web browsing from your couch including flash
your Music/Movie player (smooth 720p flash)
your Networking device (video chat, facebook etc)
your sketchpad, notepad and Document editor
your console to play flash games and emulators
Connects to your HDTV

#2
Need:
16:9 720p screen
Touchscreen
Wifi
Front Camera (3MP)
HDMI
SD card
Loudspeakers
Microphone

Expected:
Bluetooth
Accelerometre/Gyro
infrared
internal storage
Video capture

Want:
Capacitative Multitouch
C-tylus with dedicted slot
Wireless HDMI
GPS



#3
I think it should be a 10" touchscreen.
The screen would be covered in 0.5" bezel on the horizontal ends and 1.5" bezel on the vertical ends.
The bezel is black/white, continuous (borderless) and covered in glass ... eg iPhone
The horizontal bezel has no features.
Two loudspeakers are centred on the vertical ends.
The vertical bezel should include a camera (on the top right corner), and two customizable buttons oppositely arranged (on the bottom corners), and a home button (on the top left corner).
On the bottom side of device (not bezel) in the left and right corners it should have two "laptop security holes".
On the bottom side of device the HDMI plug is centered.
On the right side of device the headset jack (top corner), AC plug (center) and USB plug (bottom corner) are present.
On the top side of device on the right corner there is a stylus slot that slides towards the right.
On the top side of device the SD slot and microphone hole are centered.
On the left side of device the Volume rocker (top corner), lock button (centre), USB plug (bottom corer) are present.
On the back panel, the battery, removable, is centered and covers a large surface

The tablet can be locked to a desk by the "laptop security holes"
The tablet can be connected to a keyboard+mouse+battery accessory.
This accessory has a swivel hinge on the base and the keyboard+mouse on the top. It attaches firmly via the "laptop security holes" and connects via a USB.
This accessory swivels to form a clamshell (ie closes) or opens up to a maximum angle to double as a kickstand (refer to the "Litl").
This accessory can be attached backwards (keyboard side down) to retain the tablet form but provide extra battery power.
Another accessory is identicle in form but it accomplishes as a dual(touch)screen to provide more screen-estate and enhance the experience by providing a new means of interaction.



#4
Darwin (OSX) is out of the question, I am not Apple, nor do I want to risk legal/Hackintosh problems
Windows 7 is the best OS available but it requires a powerful CPU, has security risks and only runs on x86 CPUs
Windows CE/Mobile/Phone are possible but not appealing OS's due to age and restrictions and the lack of versatility (contarary to x86 Wins).
Linux is great! It is safe, highly functional/adaptable, light and runs on most architectures. The problem is that it is not popular as personal computers as the others and lacks finger friendliness.
However, Android is popular and cross-compatable Linux but lacks the flexibility or compatability to its (Desktop Linux) father.
MeeGo if essentially Android (has all of its advantages) but (none of its weaknesses) is very versatile and can be compatible with Desktop Linux applications.
The clear winner is Windows for powerful device and MeeGo for weaker one.
Since MeeGo has not been officially released it is quite a significant gamble.
So infact the obvious answer is Android vs Windows.

After considering both systems I would prefer Windows 7 which is highly tuned to run quicker and has undergone an innovative custom skin surgery from the ground up.
If I cannot accomplish to get fluid/decent speed and apply a finger-biased skin, it will get the job done and have access to the largest libabry of Apps and constant support.
It is very versatile and should make the diy guys/nerds appeal to it aswell as the casual user (something they are used to/something compatable with their pre-exisisting pcs)
If MeeGo project doesn't gain speed and the adjustments to Windows 7 are inadequet, Android (pref. Froyo) would win by default.



#5
The easiest way to find this out is a step-by-step approach
First, find the capacity of the charge a battery would contain for your device.
Secondly, find the average battery time your device NEEDS for its intents and purposes
Third, list all of the processors that meets the battery life requirements (for the given battery) ... this could be a long list
Forth, cross out ones that cannot support the critical features (eg GPS connectivity, OS support etc)
Fitfh, cross out the ones that are out of your budget.
Now find the one's that offers you the best/great performance and is convinient for you (drivers, software, physical size/anatomy etc)

A) In this case, this slate should hold roughly 5,000mAh (underestimated slightly)
B) I think 7hours is sufficient and mobile for this device
C) I'm not familiar with many CPUs (mainly Intel and ARMs, I'll wait for Exophase to add stuff here) but here's a sample:

5.7hrs with Intel Core i7 640-UM Dual 1.2GHz*
6.2hrs with Intel Core2Duo SU7300 Dual 1.3GHz
8.5hrs with Intel Atom N450 Solo 1.7GHz^
9.0hrs with Intel Atom Z600 Solo 1.9GHz^
9.0hrs with ARM Cortex A9 Dual 2.0GHz*
*similar performance, ^similar performance, All figures guestimated according to target device by exsisting devices or research.

D) Again I don't know the specifics of each but my device doesn't have anything out of the norm so all the listed ones should be capable
E) I don't know the budget and I'll assume I can afford any of these processors
F) I think I will choose the Intel Atom N450 (1.7GHz Solo). It meets my requirements, gives a decent performance, is widely supported and well priced (maybe). And the Atom gives me the royalty of choosing Windows/MeeGo/Android!



#6
I have two ideas.
I've seen the success of the iPad and forseable competition in the field. I've also noted how much support and attention the MS Courier (recently scrapped) concept had alerted.
In leu of these, my original design is this:

It has a "polished feel" resembling the simplicity of the iPhone/iPad without sockets placed randomly.
It is superior to the second design except in one aspect the Dual-screen.
Due to the hinges' position the gap difference between the two screens is much larger and in the scale of "noticible"


Here's the second idea:

It retains some of the simplistic iPhone esque but to a less standard as the sockets are placed irrationally about (like a laptop).
It compromises design for function (compared to the original) but it allows a very small gap between the Dual screens and increases the experience.

EDIT: My connection is currently slow, I will post the images another day. Sorry for the inconvinience.
EDIT: I've placed them in there but cannot confirm if it did (too slow, doesn't render images) so the gallery is HERE or search for exoslatej1, exoslatej2, exoslatej3 (filenames)

 

[Fzero]

The fact that OPT had done similar projects before, ran an established online store, had a uk phone number I spoke to them on, and had shown themselves on a number of videos were the factors that decided my trusting of them, I think getting peoples trust for these type projects/investments is probabably one of the hardest things, but good luck if you do do this ARM netbook.
I guess it wouldn't be as much an monetary investment though, so that would make [someone like] me more probable to take the 'risk' and not need so much convincing that it's all legit as when handing over a fair amount of money, hundreds of quid.
...and I say risk only as anyone would with an investment on trust over basically the interwebs, nothing personnal to you

I could be interested myself, I recently got a samsung netbook as a cheapish quick buy when my laptop bust and I was without anything to work on, the 10" scree. Is fine for me but more important is the weight, I'd like it lightweight as possible.
I really like the keys on the mac and some others, the spaced out ones.
At least 7+ hr battery
and loud speakers! Watching films outside on them without needing earphones or being right up against the screen to hear, loud.

One thing I would love but I'm sure isn't possible as not seen on any netbook laptop or the likes, BASS!
I guess it's the lack of space bass speakers need? I once got a laptop with some harmen kardon [sp?] speakers that were top of the range apparently, but only good tweet sound, it's a shame you can't get real sound out of anything without pluging headphones or woofer speakers.

 

[Asiyura]

To obtain good bass reproduction, you need to displace large amount of air at low frequency…
It's not doable within the volume of a netbook case full of electronic devices.

My bass's amp is fitted with an aluminum cone of 15" with 3 airpipes on the back of the 608 x 466 x 418 mm box, the 2 loudspeakers of my stereo system have 3 speakers : 1 tweet, 1 mid and 1 9" bass.
My conclusion is : for good bass, you need LARGE woofers, I would say at least 5".

 

[Kangal]

I've actually did some more research and wanted to change a few points on it.
Here's the CPU + battery life update:

10" 720p + C-touch @ normal power draw with 5,200mAh (4-cell)

5.2hrs with Intel Core i7 640-UM Dual 1.2GHz*
5.7hrs with Intel Core2Duo SU7300 Dual 1.3GHz
7.8hrs with Intel Atom N450 Solo 1.7GHz^
8.8hrs with Intel Atom Z600 Solo 1.9GHz^
9.0hrs with ARM Cortex A9 Dual 2.0GHz*
*similar performance
^similar performance


Same setup with the extended 4-cell battery (+5,200mAh) via keyboard/dual screen` = 10,400mAh Total
9.6hrs with Intel Core i7 640-UM Dual 1.2GHz
10.5hrs with Intel Core2Duo SU7300 Dual 1.3GHz
14.4hrs with Intel Atom N450 Solo 1.7GHz
16.3hrs with Intel Atom Z600 Solo 1.9GHz
16.7hrs with ARM Cortex A9 Dual 2.0GHz
`The dual screen has the same battery embedded as the keyboard and device but due to increased power draw from the secod screen, this accessory should provide less battery life than the keyboard accessory but still more than the tablet alone.


Now we add in hig performance usage with high-end GPU to the keyboard accessory setup.
4.1hrs with Intel Core i7 640-UM Dual 1.2GHz (ATi HD 5650)
4.6hrs with Intel Core2Duo SU7300 Dual 1.3GHz (ATi HD 5650)
8.7hrs with Intel Atom N450 Solo 1.7GHz
9.8hrs with Intel Atom Z600 Solo 1.9GHz
9.2hrs with ARM Cortex A9 Dual 2.0GHz

AND MOST IMPORTANTLY, the battery will no longer be exchangable (you would need to send it in if it dying state) this is to conform to space limitations.
If you need more use time just plug in the accessory (extra battery/keyboard-shell/dualscreen).
You no longer have the option of shutdown/hibernate and switching the battery. Sorry, need to be reasonable!
These are all estimates of course, and while Core i7 or ARM have not even been released, these are within the accepted range of expected battery life (according to the battery). So in theory, this device is theesable but it may need many optimizations to hit those figures. In practice, we have yet to see.



Firstly, the specs are:
N450 is the chosen CPU
1 or 2GB RAM
128GB SSD/NAND/wateva (not enough room for HDD)
Windows 7 optimized smoothly (not laggy) and made a custom UI skin on top from the ground up [switchable to original UI]
GPU is anything that sips battery (little) but can handle flash 720p at 60fps onscreen and external display simultaneously

This will meet all your computing needs plus some gaming like:
Billions of flash games, Some old but fun PC-games (Quake 3, GTA2, AgeOfMyth etc), DOSbox, N64, DSi, PSOne, SNES, GBA, Sega Genesis, C-64, Atari.


Now secondly, There may be an upgraded model introduced with the same specs except for:
Core i7 640-UM
Upgraded 4GB RAM
Same OS (maybe minus some limitation/optimization since the CPU is more capable)
ATi HD 5650 GPU (switchable)

This would be much more expensive, and even harder to be able to fit the new CPU, GPU and the battery in.
But the added bonus is that you have a smooth experience and better multitasking.

The games library is updated from "All of the original model" plus:
Dreamcast, PCSX2, Dolfin (GC+Wii emulator ... but poorly) and any PC game for Windows between Quake 3 and Current-gen games (Upto MW2, Far Cry 2, Dirt 2, Bad Company, NFShift, Crysis @med +30fps). Less demanding games could run fullspeed at high settings, even online ones like WoW or Combat Arms!
So infact, this model is a full-blown computer squished into a thoughtful tablet form!

 

[Kangal]

So did anyone see my tablet concept?

What are your thoughts?
Do you think the market would want a device like that?

 

[RajTakhar]

to put it quite bluntly - plan the thing out
get your suppliers in line
build a prototype
test the shit out of it
Be happy with it
then pre-order RAD dont work (well in pandora's case)

 

[DEAT]

Firstly, the specs are:
N450 is the chosen CPU

[...]

Now secondly, There may be an upgraded model introduced with the same specs except for:
Core i7 640-UM

Congrats, you completely missed the point of this thread and eradicated the whole niche of this product.

 

[Kangal]

Congrats, you completely missed the point of this thread and eradicated the whole niche of this product.

I didn't miss the point/anything. And I never wanted to derail the thread, notice I've made some input to the OP rather than just my jargon. The thing is this "niche" is a very small gap. Smartphones are getting larger and smarter and notebooks are getting dumber and smaller!

Read a page previous. I told them about how I would go about designing a product. I followed up with why I thought a smartbook (ARM netbook) would have a difficult time because of a lack of software (MeeGo aint here yet) and the fierce competition: iPad, netbooks, netvertibles, Windows tablets and Android tablets!
C'mon if I'm a customer I would buy one device to fill that niche not multiple!.

Then I gave my example of filling that niche: a Windows tablet that doubles as a netbook/CULV and also becomes the MS Courier!
(so it's a 3-in-1 omni device)

Sorry, I just had this concept in mind ever since they killed the Courier so I did a little "project" to make it a concept and would really like some feedback. You never know, this could be a future device!

So ...
what do the brilliant minds of GP32X bloggers think of my tablet concept?
What are your thoughts/improvements?
Do you think the market would want a device like that?

 

[Kangal]

Sorry I've hijacked the thread (but given some sweet ideas in the process)

Just wanted to ask has anybody heard of the Touchbook by Always Innovating?

Source
A Review, more like an analysis
Cool use

It's been a while, but I've only stumbled across it now and see it has the same design aspects of my concept, weird huh?