Why not use Kickstarter to get a batch made?

[Frank26080115]

If there are no flaws in the design, why not make a small batch of
MicropendousX made? I would buy one. If you want to fundraise by
allowing pre-orders using Kickstarter, that'll be great. I would
definitely pre-order one.

Just make a video of the board running a web server from the microSD
card and a USB thumb drive, and then also make a video of the Android
ADK working with it as well. Show you run "make" and then use the
bootloader. I'm sure people would love seeing that sort of demo and
you'll get plenty of attention once people see it. I'm sure you'll get
enough pre-orders if you get on the front page of Hack a Day or
Makezine or something. Or make a deal with a store like SparkFun or
Seeed Studio, you'll get the sales you need if you make the front
page, just because of the volume of traffic they get.

[Opendous Inc.]

Unfortunately, the board does have issues. JTAG drops out randomly,
Ethernet sends out packets only once in a while, and I have fried a
couple of the ICs with ESD even though I am incredibly careful and
have ESD mats and wrist straps. Without a proper GND plane the high
speed IO creates too much GND bounce that even mass amounts of
decoupling capacitors cannot fix as the long GND leads act to increase
capacitor inductance.

I tried the 2-layer thing but have concluded there is not much
chance of getting anything reliable and have moved on to a proper 4-
layer design. Due to the added cost I figure I may as well get HS-USB
and have decided on the LPC18xx. You can follow my progress:
http://code.google.com/p/micropendousx/source/browse/trunk/MicropendousX/Hardware/MicropendousX-LPC183x/
http://code.google.com/p/micropendousx/source/browse/trunk/MicropendousX/Hardware/MicropendousX-LPC185x/

Kickstarter is US-only but I am aware of similar international
services such as indiegogo.com and peerbackers.com.

This project is still moving forward and will be a lot more useful
and reliable when it is ready.

Thanks for the advice. I will make use of it.

[uros]

Thanks for these important details. However we made your card with no
errors so long, but we didn't test ethernet yet! I'm really interested
in going further with this processor for now. Do you have some
documentation about correct way to do isolation in 4 layers?? I'll be
really grateful. Your card has really good features so I think of
redesigning it in 4 layers, and of course sharing it.
I have one mBed card that has the same proc for exemple, and it's 4
layer PCB, so you are right, 4 layer is necessary for this procs.

I was working also on toolchain with Drasko, we think that we can make
makefile more simple and clear to use. For now i'm compiling for your
board with makefile that is used to rebuild mbed project and it's
fairy simple. Did you look at :
http://dev.frozeneskimo.com/notes/compiling_your_own_cmsis_code_for_the_mbed

specially if you are working with other systems than windows (we are
on os x and linux). In every case it's a good start to make a makefile
for this architecture...


Best,
uros

On Sep 12, 7:33 pm, "Opendous Inc." <opend...@gmail.com> wrote:
>   Unfortunately, the board does have issues.  JTAG drops out randomly,
> Ethernet sends out packets only once in a while, and I have fried a
> couple of the ICs with ESD even though I am incredibly careful and
> have ESD mats and wrist straps.  Without a proper GND plane the high
> speed IO creates too much GND bounce that even mass amounts of
> decoupling capacitors cannot fix as the long GND leads act to increase
> capacitor inductance.
>
>   I tried the 2-layer thing but have concluded there is not much
> chance of getting anything reliable and have moved on to a proper 4-
> layer design.  Due to the added cost I figure I may as well get HS-USB

> and have decided on the LPC18xx.  You can follow my progress:http://code.google.com/p/micropendousx/source/browse/trunk/Micropendo...http://code.google.com/p/micropendousx/source/browse/trunk/Micropendo...

[Opendous Inc.]

>However we made your card
>with no errors so long

It isn't a problem using a single interface or a few GPIO pins; only
when you make fuller use of the available functionality. You can make
it work with 2-layers with careful design and if you don't demand too
much from the chip.

>correct way to do isolation in 4 layers

Take a look at the following simple design for an isolator that
takes into account most of the AppNote suggestions related to the
ADuM6401 which offers 5kV of isolation:
http://www.analog.com/en/interface/digital-isolators/adum6401/products/product.html
http://code.google.com/p/opendous/source/browse/trunk/Current_Designs/Isolator/

On Sep 23, 11:28 am, uros <uros.petrev...@gmail.com> wrote:
> Thanks for these important details. However we made your card with no
> errors so long, but we didn't test ethernet yet! I'm really interested
> in going further with this processor for now. Do you have some
> documentation about correct way to do isolation in 4 layers?? I'll be
> really grateful. Your card has really good features so I think of
> redesigning it in 4 layers, and of course sharing it.
> I have one mBed card that has the same proc for exemple, and it's 4
> layer PCB, so you are right, 4 layer is necessary for this procs.
>
> I was working also on toolchain with Drasko, we think that we can make
> makefile more simple and clear to use. For now i'm compiling for your
> board with makefile that is used to rebuild mbed project and it's

> fairy simple. Did you look at :http://dev.frozeneskimo.com/notes/compiling_your_own_cmsis_code_for_t...

>
> specially if you are working with other systems than windows (we are
> on os x and linux). In every case it's a good start to make a makefile
> for this architecture...
>
> Best,
> uros
>
> On Sep 12, 7:33 pm, "Opendous Inc." <opend...@gmail.com> wrote:
>
> > Unfortunately, the board does have issues. JTAG drops out randomly,
> > Ethernet sends out packets only once in a while, and I have fried a
> > couple of the ICs with ESD even though I am incredibly careful and
> > have ESD mats and wrist straps. Without a proper GND plane the high
> > speed IO creates too much GND bounce that even mass amounts of
> > decoupling capacitors cannot fix as the long GND leads act to increase
> > capacitor inductance.
>
> > I tried the 2-layer thing but have concluded there is not much
> > chance of getting anything reliable and have moved on to a proper 4-
> > layer design. Due to the added cost I figure I may as well get HS-USB

> > and have decided on the LPC18xx. You can follow my progress:http://code.google.com/p/micropendousx/source/browse/trunk/Micropendo......