LPC1102, World's smallest 32-bit ARM MCU in a 5mm2 pckg

On Apr 29, 12:02=A0pm, An Schwob in the USA <schwo...@aol.com> wrote:
> On Apr 29, 5:47=A0am, rickman <gnu...@gmail.com> wrote:
>
>
>
> > On Apr 23, 8:34=A0pm, An Schwob in the USA <schwo...@aol.com> wrote:
>
> > > On Apr 23, 5:12=A0pm, linnix <m...@linnix.info-for.us> wrote:
>
> > > > On Apr 23, 4:55=A0pm, "langw...@fonz.dk" <langw...@fonz.dk> wrote:
>
> > > > > On 24 Apr., 01:37, linnix <m...@linnix.info-for.us> wrote:
>
> > > > > > On Apr 23, 4:32=A0pm, malcolm <malcolm...@gmail.com> wrote:
>
> > > > > > > On Apr 24, 9:45=A0am, An Schwob in the USA <schwo...@aol.com>=
 wrote:
>
> > > > > > > > Hi,
> > > > > > > > ever heard about a 32-bit microcontroller this tiny? I am f=
amiliar
> > > > > > > > with the Silabs devices that are available in similar packa=
ges but
> > > > > > > > this time it is an ARM Cortex-M0 and it has more I/O and me=
mory than I
> > > > > > > > thought.
> > > > > > > > 10-bit ADC, 4 timers, UART, SPI, 32K Flash and 8K SRAM, rea=
lly not bad
> > > > > > > > for such a grain of silicon.
>
> > > > > > > > Check it out.http://www.mcu-related.com
>
> > > > > > > > An Schwob
>
> > > > > > > Yes, it's small, in two dimensions...
> > > > > > > However, deployed it will need a minimum layer count,
>
> > > > > > There are only 4 pads in the center. =A0Double layers should wo=
rk, but
> > > > > > you need plugged via.
>
> > > > > > > so that Y axis comes at some cost.
> > > > > > > If it bumps the layer count on your board, it's suddenly very
> > > > > > > expensive.
>
> > > > > > > Can they fit the die into SO16N, or QSOP16 - and so
> > > > > > > offer a single-sided solution ?
>
> > > > > > > -jg
>
> > > > > I think you could route the inner pads out between the outer, wit=
h 4-5
> > > > > mil design rules
>
> > > > > -Lasse
>
> > > > I usually limit myself to 7 mil rules. =A0I wonder what the center =
pads
> > > > are. =A0If unnecessary I/Os, you can ignore them. =A0If ground pads=
, you
> > > > can join them with an outer pad. =A0Anyone got inside info?
>
> > > Hi,
>
> > > I was looking for more info too but could not find anything more
> > > specific on the NXP website. Expect to see more next week because NXP
> > > announced to show something at the ESC in San Jose.
>
> > > An Schwob
>
> > This part is not for low cost boards. =A0It is for small devices.
> > Sometimes, but not always the same thing.
>
> > Rick
>
> Rick,
>
> you are absolutely right. You can not use an ultra low cost facility
> because the device need at the very least a double sided PCB, may be
> more layers and a more sophisticated machinery than a soldering iron
> and low cost labor. The lowest cost devices are the LPC1111, selling
> for 65 cent @ 10k according to the press release.http://www.mcu-related.c=
om/architectures/35-cortex-m3/92-lpc1100
> Still a bit tough for hand soldering but low cost equipment will do an
> excellent job with these chips.
>
> An Schwob

Since when is hand soldering cheaper than pick and place???  Maybe if
you are building just one, even then I have been asked for CNC
information since they just don't want to do hand work.  The cost in
using these parts comes from the small via sizes required... at least
I am assuming.  Is there a breakout that does not require drills
smaller than 12 mil?  When I try to get boards made with 10 mil vias
the board makers have "rounded" that up to 13 citing a +- 3 mil
tolerance spec.  of course I don't do business with that company
anymore, but the point is smaller drills cost extra and that is a
recurring charge.  I can't say anything about the number of layers
since I haven't tried to route this part yet.  But I'd bet it is hard
on a two layer board.

It may be that the assembly house charges extra for mounting these
parts too.  I've seen extra charges for pin pitch below 20 mil (0.5
mm) on leaded parts.  I'm not sure where they start charging extra for
fine pitch BGA type parts.

But then if this size part makes a design possible that otherwise
wouldn't be, it doesn't matter if the unit costs a few cents extra.

Rick

---snip----
> > Rick,
>
> > you are absolutely right. You can not use an ultra low cost facility
> > because the device need at the very least a double sided PCB, may be
> > more layers and a more sophisticated machinery than a soldering iron
> > and low cost labor. The lowest cost devices are the LPC1111, selling
> > for 65 cent @ 10k according to the press release.http://www.mcu-related=
.com/architectures/35-cortex-m3/92-lpc1100
> > Still a bit tough for hand soldering but low cost equipment will do an
> > excellent job with these chips.
>
> > An Schwob
>
> Since when is hand soldering cheaper than pick and place??? =A0Maybe if
> you are building just one, even then I have been asked for CNC
> information since they just don't want to do hand work. =A0The cost in
> using these parts comes from the small via sizes required... at least
> I am assuming. =A0Is there a breakout that does not require drills
> smaller than 12 mil? =A0When I try to get boards made with 10 mil vias
> the board makers have "rounded" that up to 13 citing a +- 3 mil
> tolerance spec. =A0of course I don't do business with that company
> anymore, but the point is smaller drills cost extra and that is a
> recurring charge. =A0I can't say anything about the number of layers
> since I haven't tried to route this part yet. =A0But I'd bet it is hard
> on a two layer board.
>
> It may be that the assembly house charges extra for mounting these
> parts too. =A0I've seen extra charges for pin pitch below 20 mil (0.5
> mm) on leaded parts. =A0I'm not sure where they start charging extra for
> fine pitch BGA type parts.
>
> But then if this size part makes a design possible that otherwise
> wouldn't be, it doesn't matter if the unit costs a few cents extra.
>
> Rick

Hi Rick,

that was a misunderstanding. I was talking about low cost equipment
for the LPC1111. Never said or wanted to imply that hand-soldering is
the lowest cost. Just to make sure there is not another
misunderstanding, I totally agree with you!

An Schwob