Re: LCD
Casainho
<opendous...@gmail.com> wrote:
> Casainho,
>
> eclipse and your SDRAM code work for me! I ran tests with all ones
> as well. However, I still have not figured out how to use eclipse's
> built-in compile functions with your code. What specific versions of
> gcc and gdb are you using? I am using the Chumby 2008q3-72 GCC and
> the 2010q1-188 GDB.
I always have a open shell where I run make. Eclipse can run make
however for me it can't find the gcc on path, like if path is not set
on Eclipse. So, I just run make myself.
I am using on Linux this one:
http://www.codesourcery.com/sgpp/lite/arm/portal/release1294
> Don't worry about the LCD connector. As you can see from the
> pictures it can withstand reflow and there is double-sided tape near
> the connector's edge.
Ok :-) -- maybe I can find over Internet some howto solder that
cables, or at least some pictures of correct soldered by hand.
>>>If you and Bob are interested in pursuing
>>>the LCD I suggest you order the parts from
>>>www.DigiKey.pt yourself.
>>
>>Ok, I will. But Digikey is slow to arrive here in
>>Portugal, also because I need to pay taxes :-(
TAX here in Portugal is 21% :-( -- but I were talking with Bob and on
UK is 15%. Bob will buy for both of us and after send my part :-)
>>when you have your LCD soldered, please post pictures.
>
> I will but I am still considering using the older LCD pictured which
> will need wire jumpers as the footprint is different. It will not
> look anything like a proper board. I still need to finish the SDRAM
> first.
Sad that we will have different hardware, that will make difficult for
both of us, because of different firmware drivers.
Nice, so SDRAM should work correctly, I hope.
Anyone knows if SDRAM can run at higher than 133MHz?
> By the way, my name is not Matt Opendous. Opendous Inc. is the name
> of my company and it is the combination of the words Open and
> Stupendous. It would be more correct to say "Matt from Opendous Inc."
> in any posts.
Ok ;-) -- Since I don't know more of your full name, I was telling
like that. I will now say that "Matt from Opendous Inc." :-)
Casainho
>>
>> I will but I am still considering using the older LCD pictured which
>> will need wire jumpers as the footprint is different. It will not
>> look anything like a proper board. I still need to finish the SDRAM
>> first.
>
> Sad that we will have different hardware, that will make difficult for
> both of us, because of different firmware drivers.
>
> Nice, so SDRAM should work correctly, I hope.
>
> Anyone knows if SDRAM can run at higher than 133MHz?
In next 3 or 5 years, we will be making our Open Source and DIY netbook :-)
We just need an ARM like this one running at 1GHz or more. If we have
electronics that can be DIY, also the LCD and a better 3D print
technology :-)
Bob
> In next 3 or 5 years, we will be making our Open Source and DIY netbook :-)
>
> We just need an ARM like this one running at 1GHz or more. If we have
> electronics that can be DIY, also the LCD and a better 3D print
> technology :-)
notebook/netbook area, there was a thread on the Pandora forums which
now I can't find. I was just looking about and found this one
http://sharism.cc/specs/ which is not very far from our spec (ex Open
Moko guys).
There are also several companies claiming to be "open hardware" but
more in marketing than in practice, like these guys
http://www.alwaysinnovating.com/touchbook/ but they seem to have a
smart looking product.
That unfortunately is where the open projects come unstuck. I am
skeptical that even commercial printing can achieve the unit cost and
quality that injection moulding can achieve, certainly not with
current technology. People have gotten used to nicely finished
plastics at low cost, and generally turn up their nose at things that
look "homemade". People judge a netbook by its cover ;)
john cooper
> On 12 June, 21:00, Casainho <casai...@gmail.com> wrote:
>
>> In next 3 or 5 years, we will be making our Open Source and DIY netbook :-)
>>
>> We just need an ARM like this one running at 1GHz or more. If we have
>> electronics that can be DIY, also the LCD and a better 3D print
>> technology :-)
>
> I have heard of a few projects doing open hardware stuff in the
> notebook/netbook area, there was a thread on the Pandora forums which
> now I can't find. I was just looking about and found this one
> http://sharism.cc/specs/ which is not very far from our spec (ex Open
> Moko guys).
I thought OLPC would have been a decent bridge to that
end. RedHat is a participant and I had a unit dropped off
in my cube about a year or so ago. It's a commendable
effort by any measure, but to be honest for personal use
I'd probably just drop $300 for an Acer Aspire. And this
is from one so incurably addicted to homebrew I'm typing
this in a home I physically built.
My observation from the trenches is opensource (and perhaps
open technology in general) tend to excel at refining
relatively mature technologies, and only when they attract
sufficient critical mass to drive and propagate the effort.
Take what we're doing here as an example. You wouldn't have
seen it happen 5 years ago because the technology hadn't
matured to the point of suitable SoCs being available and
usable without entering a NDA, though they surely existed.
Even then without critical mass to drive the effort beyond
inception the project is likely to fizzle out.
> There are also several companies claiming to be "open hardware" but
> more in marketing than in practice, like these guys
> http://www.alwaysinnovating.com/touchbook/ but they seem to have a
> smart looking product.
Nice! Although I speculate the fairly recent ipad introduction
may have derailed their charter somewhat. Still they seem
to be fighting the good fight.
> That unfortunately is where the open projects come unstuck. I am
> skeptical that even commercial printing can achieve the unit cost and
> quality that injection moulding can achieve, certainly not with
> current technology.
Sadly I agree -- it can't. Nor can my approach of milling
a case from an ABS blank.
Then again everyone in the better-widget-hitting-the-market-window
race pays the same price to enter the game by leveraging an
existing revenue stream or the promise of one. It's still hard
to make a business case built around open technology where you
intend to fully adhere to its spirit. Yet there are no
shortage of companies which have managed to get themselves in
the FSF's "sue queue" by dipping into the GPL code pool and
not tossing anything back.
I suspect the best general, realistic compromise is struck similar
to Chumby's model. Being a member in good standing of the open
technology community from which they draw, and enough decoupled
IP to maintain a revenue stream.
-john
Bob
> I have heard of a few projects doing open hardware stuff in the
> notebook/netbook area, there was a thread on the Pandora forums which
> now I can't find.
http://www.gp32x.com/board/index.php?/topic/52705-help-needed-replicating-success-of-openpandora-for-a-netbooklaptop
The proposer, lkcl, is not so concerned with open-ness as the fact
that cheap ARM netbooks can be made but manufacturers are not
interested in selling them, the old Intel/Microsoft has something to
do with that I suspect. I think lkcl is trying to get a group together
to create a spec and then have a few thousands units made by a Chinese
supplier. I see that he has already had problems with the Chinese
approach to IP ;)
http://elinux.org/Main_Page seems to be a portal for various embedded
Linux projects.
As far as DIY hardware goes, looks like these guys are nearly there
http://www.liquidware.com/shop/show/BB-BT/BeagleTouch needs add-on
battery pack though, BeagleJuice.
Casainho
> http://www.liquidware.com/shop/show/BB-BT/BeagleTouch needs add-on
> battery pack though, BeagleJuice.
Beagle board is not DIY. That's why I think this imx233 board is so
important, it can be assembled by hand with cheap tools.
And Jhon, I also use one cheap Acer Aspire One, Linux version SSD. On
next August will make 2 years in my hand. Even before this netbook I
had the Asus eee pc 701 :-)
I use a lot this netbook, I had read a lot and learn, I started to
working with ARM thanks to Rockbox project and I had read many PDF
documents form Atmel and Rockbox code, on this netbook, while
travelling by train to work, all days.
And still now I use it a lot for electronics and Internet.
Why Open? because many people hacked their own eee pc 701, mainly to
add USB devices like bluetooh, flash disks, HDD, etc. And because I am
using SDCard for my Linux system and my documents and I would like to
costumize with a larger battery, no fan, ARM, etc.
When eee pc 701 went to market, that ones used SSD disk and I saw
later all going to HDD, just because of windows. That's why I would
like to build my own, to used the technology in which I believe, SSD,
ARM and low power.
Bob
> I found the thread,http://www.gp32x.com/board/index.php?/topic/52705-help-needed-replica...
http://www.advogato.org/person/lkcl/ let's just say lkcl aka Luke
Leighton's grasp on reality may not be too strong. Standing as a
Pirate Party candidate in the election is one thing, but if you write
letters to the Queen and report the government to the local police
station for "crimes of treason" I think you would expect to be invited
to a mental health assessment, rather than be surprised by it.
Either way, it looks like Luke has a lot of stuff on his plate, I hope
he is not getting burn out.
Opendous Inc.
>Open Source and DIY netbook :-)
>
>We just need an ARM like this one running
>can be DIY'ed, an LCD and better 3D
>print technology :-)
You are going to have to adjust your notion of DIY to pursue a
Netbook. All future higher end ICs will be BGA as it is the best IC
package:
http://cache.freescale.com/files/32bit/doc/package_info/PBGAPRES.pdf
The good news is that BGA soldering can be very simple. I have
prototyped about a dozen BGA boards so far and all of them work about
as well as any of my QFP-based prototypes. Cheap DIY rework is
impossible however. All you need is a solder paste stencil and a
cheap convection oven:
http://ohararp.com/Stencils.html
http://www.walmart.com/ip/Black-Decker-Convection-Toaster-Oven-Black/12533802
The initial version of the PROpendous project was based on the
LPC3130 but it was impossible to get RAM and LCD functionality using
just two layers. The following board works in its limited capacity
and uses 0805 components for easy prototyping:
http://propendous.googlecode.com/svn/trunk/PRopendous/Hardware/PROpendous-DIY/PROpendous-DIY_Picture_Small.jpg
I have recently come across only one GHz-class IC in a QFP package.
The Marvell 88F6192 runs at 800MHz and comes in a 216LQFP 0.4mm pitch
package. It costs around $35, so about as much as an i.MX515 which is
supported by Ubuntu. It has advanced features like PCI-Express, dual
Gigabit Ethernet, and dual SATA along with standard features like USB,
GPIO, UART, SPI, I2C, SDIO, I2S, TS/Video, and Flash memory
interface. Unfortunately, although the IC is QFP it only supports
DDR2 memory which is BGA-only. The more advanced 88F6281 is BGA, runs
at up to 1.5GHz, and is the basis for the SheevaPlug(http://
www.plugcomputer.org/) and Open-RD (http://www.open-rd.org/). Both
run Linux and Open-RD is Open Hardware if you ignore the proprietary
file formats. Basically, these are more capable but more power hungry
BeagleBoards.
The biggest obstacle to a DIY Netbook/Laptop is the obscene costs of
LCD displays in small quantities. A decent touch-enabled 7" LCD
(LQ070Y3DG3B) is $150+ at Avnet.com while a 15" LCD goes for $550+.
http://www.sharpledlcd.com/7-inch_Displays.html
Casainho
> Netbook. All future higher end ICs will be BGA as it is the best IC
> package:
> http://cache.freescale.com/files/32bit/doc/package_info/PBGAPRES.pdf
> The biggest obstacle to a DIY Netbook/Laptop is the obscene costs of
> LCD displays in small quantities. A decent touch-enabled 7" LCD
> (LQ070Y3DG3B) is $150+ at Avnet.com while a 15" LCD goes for $550+.
> http://www.sharpledlcd.com/7-inch_Displays.html
Well, let's see how things will be in next 2 years. For now I am
really happy to have the imx233 board, I am sure I can play and enjoy
a lot with our current Open Source and DIY technology :-)
john cooper
> You are going to have to adjust your notion of DIY to pursue a
> Netbook. All future higher end ICs will be BGA as it is the best IC
> package:
> http://cache.freescale.com/files/32bit/doc/package_info/PBGAPRES.pdf
In retrospect I think that writing was on the wall about a
decade or so ago.
> The good news is that BGA soldering can be very simple. I have
> prototyped about a dozen BGA boards so far and all of them work about
> as well as any of my QFP-based prototypes. Cheap DIY rework is
> impossible however.
I can't say I've had to reball a bga device thus far, but the
claimed success reported by others doesn't discourage me
from the task. Both solid balls and solder paste approaches
seem to have fairly promising success rates. But in the case
of paste I'd hazard the stencil void volumes and paste solids
content are fairly critical. Still spending an hour reballing
a $5 device isn't justifiable in general.
The snag I'd encountered with prototyping bga devices was that
of finding an affordable PC fab service. Even a comparatively
relaxed 0.8mm ball pitch was beyond the economy class board fab
services. Running vias in pads might be technically possible
but getting the vias reliably closed isn't. The silver lining
here being even the econo board fabs will need to follow the
median technology so lack of economical bga support is likely
just a transient.
> http://www.walmart.com/ip/Black-Decker-Convection-Toaster-Oven-Black/12533802
Oh man. I was just in MalWart a few days ago looking at that
exact model. A convection oven for that price caught my eye.
The baffles surrounding the heating elements have progressively
spaced apertures (tighter pitch at the wall ends) apparently
intended to balance the heat output relative to the element
centers. If you've actually tried that unit to reflow (or
better yet probed it with a thermocouple) I'd be interested to
hear your results.
The clunker I'd been using, while being a convection oven,
has fatigued elements which have developed visibly cool spots.
> I have recently come across only one GHz-class IC in a QFP package.
> The Marvell 88F6192 runs at 800MHz and comes in a 216LQFP 0.4mm pitch
> package. It costs around $35, so about as much as an i.MX515 which is
> supported by Ubuntu. It has advanced features like PCI-Express, dual
> Gigabit Ethernet, and dual SATA along with standard features like USB,
> GPIO, UART, SPI, I2C, SDIO, I2S, TS/Video, and Flash memory
> interface. Unfortunately, although the IC is QFP it only supports
> DDR2 memory which is BGA-only. The more advanced 88F6281 is BGA, runs
> at up to 1.5GHz, and is the basis for the SheevaPlug(http://
> www.plugcomputer.org/) and Open-RD (http://www.open-rd.org/). Both
> run Linux and Open-RD is Open Hardware if you ignore the proprietary
> file formats. Basically, these are more capable but more power hungry
> BeagleBoards.
Having thrown in the towel long ago on competing with the resources
of the Taiwanese National PC fabrication machine, I can only applaud
the effort to design a non-x86 notebook/palmtop. History painfully
demonstrated how an architecture ill-suited for the task can be
compensated by an army of engineering effort hell bent on making it
somehow work. So I just buy the x86 stuff, pave it over with linux,
pull a bag over my head, and try not to think about it too much.
> The biggest obstacle to a DIY Netbook/Laptop is the obscene costs of
> LCD displays in small quantities. A decent touch-enabled 7" LCD
> (LQ070Y3DG3B) is $150+ at Avnet.com while a 15" LCD goes for $550+.
> http://www.sharpledlcd.com/7-inch_Displays.html
Which given the volume in which lcd panels are being consumed
by the PC world, the cynic in me wonders whether that's the
price point where they can justify catering to that market
segment or have set the price so dauntingly high as to discourage
use in such volumes from ever being cost effective.
-john
Opendous Support
>that of finding an affordable PC fab service. Even a
>comparatively relaxed 0.8mm ball pitch was beyond the
>economy class board fab services
The LPC3130 board I designed is 2-layer with 8mil-8mil-20mil
trace-clearance-minimum_drill for cheap PCB prototyping. I could not
add RAM and FLASH as many of the signals are occluded but the board
does work as all the main power and communication signals can be
escape routed. For a 0.8mm pitch IC 12mil pads work fine and
companies like OHARARP.com can do 10mil solder paste stencil pads. I
have another board based on the LPC3130 that is a fully routed 2-layer
with 6mil-6mil-12mil board which I ordered from GoldPhoenixPCB for
$169 for 155sq.in ($99 + $20(6mil traces) + $30(12mil vias) +
$20(immersion silver RoHS finishing)). It is possible to do BGA
prototypes real cheap. Pool-PCB can do 16sq.in 6-layer boards for
<$150.
>> http://www.walmart.com/ip/Black-Decker-Convection-Toaster-Oven-Black/12533802
>
>Oh man. I was just in MalWart a few days ago looking at that
>exact model. A convection oven for that price caught my eye.
>The baffles surrounding the heating elements have progressively
>spaced apertures (tighter pitch at the wall ends) apparently
>intended to balance the heat output relative to the element
>centers. If you've actually tried that unit to reflow (or
>better yet probed it with a thermocouple) I'd be interested to
>hear your results.
I am still looking into this and other cheap convection ovens. I
have not yet bought anything. I have been using a much cheaper
toaster oven which has a hot spot near the center. It has been
adequate for reflowing small one-off boards. For larger boards I will
require a convection oven.
>>A decent touch-enabled 7" LCD (LQ070Y3DG3B)
>>is $150+ each for 5 at Avnet.com
>Which given the volume in which lcd panels are being
>consumed by the PC world, the cynic in me wonders whether
>that's the price point where they can justify catering to that
>segment or have set the price so dauntingly high as to
>discourage use in such volumes from ever being cost effective.
You can get LCDs cheaper from smaller companies but you lose the
certainty of availability and quality. The large LCD companies are in
effect denying the DIY community from experimenting with novel ideas.
The assumption that only large bulky companies can innovate seems
incorrect to me. DIY'ers are stuck with creating solutions that
cannot scale in manufacturing or overly expensive prototypes. It is
unfortunately very difficult to justify expensive prototypes when you
just want to try an idea out. Or, maybe the distributors cannot
justify large orders. I would really like to see
distributor-available ~7" touch-screen LCDs for less than a new 24"+
LCD monitor. If you find one for <$50 please let me know.
Opendous Support
Oven for reflow.
The results are better than a standard toaster oven but still nothing
like a proper reflow oven. There is still a hot spot in the center.
I have been able to successfully manufacture some boards with it.
john cooper
> I finally got around to testing the Black & Decker Convection Toaster
> Oven for reflow.
>
> The results are better than a standard toaster oven but still nothing
> like a proper reflow oven. There is still a hot spot in the center.
> I have been able to successfully manufacture some boards with it.
>
>
Yea the char pattern is heavy in the center but probably
workable. Looks like the radiant energy hitting it isn't
too well balanced. What I might try to even it out
would be to suspend an section of window screen in front
of the top/bottom elements to even it out. Heating a
board with radiant energy vs. convection always seemed dicey
to me given the wildly different absorption possible between
components. Even if the majority of heating is via convection,
I'd still probably avoid reflowing a large board which extends
into the extremes of the oven interior as the air velocity is
probably minimal in the corners.
Interesting timing. I just picked up one of these last
weekend after passing by them in MalWart for months and
figuring they would be discontinued eventually. I'll
follow up with my experimentation as soon as I have at it.
Thanks!
-john
Casainho
After he did used the cheapest 40W CO2 laser to burn the unwanted
black paint from the PCB. In the end, he etched the board as usually.
He got very good results!!! There are more people doing this. See
here:
And that "cheap laser" from China costs for me 800€ without taxes.
That laser would also cute stencils of plastic...