Lpc2000

--- In lpc2000@yahoogroups.com, Owen Mooney <ojm@s...> wrote:
> Enough responses thankyou !!! 
> 
> I don't need any more help thank - I finished it last night with all
lines connected and 1 cap under the chip with short lines to VSS and
VDD lines + a 10uf elsewhere.
> 
> This is as good as a previous LCP2106 design (at only 14 MHz) and I
have dozens of them running in the field. 
> 
> My replies to a few comments:
> 
> 1.   "If you're struggling for space, why not use 0603 or smaller
caps?" 
>  
> I am using 0603 - how else does own get a data logger with 4MBytes
of Storage and an Display driver into 65x35mm (no bigger than the LCD).
> 
> 
> 2.   "I'm not trying to put anyone down, but software people normally 
> needto learn a lot before attempting to do a hardware design."
> 
> Don't be a snob. I did all the circuit digrams for a DSP architecure in 
> 1980 with ttl chips while you were probably sucking on  ....

Hey Owen, if you don't want help, then don't ask.  You said "I'm a
software engineer *learning* hardware".  Now you are saying that you
don't need help.  Please make up your mind.  My comment was clearly
trying to say I did NOT intend to offend.  So next time, if you are
going to be sensitive, please don't put your self down first, or at
least warn us that you don't really want the help.  

BTW, a single 0603 cap of 0.1 uF or similar is clearly not enough to
**properly** decouple a chip like this, especially on a 2 sided board.
 If you don't believe me, ask Philips.  One of the biggest mistakes
engineers (hardware AND software) make is thinking that because
something has worked in the past, that it must be right.  In my 25
years as an engineer I have seen a lot of people eat their words when
their "good enough" approach did not work in a new situation.  

And remember, ground bounce is NOT your friend.   :)