Lpc2000

At 11:15 AM 11/24/03 -0500, you wrote:
>I'm looking at an "even swap" to possibly replace an Atmel AVR processor
>with an LPC2106 on a product in prototype stage.  The AVR is out of gas
>in several respects, and the '2106 looks like a possible project saver.
>We figured out we were in trouble on Saturday; it's now the Monday
>before Thanksgiving, and I am utterly unprepared.  In this context, I've
>got a couple of fundamental questions that this group might answer
>quickly..

I've been working on a SIMMstick for the lp210x (I'm just about to send out 
for the production ready boards).


>1. Which Development/Evaluation boards have you used (i.e., Ashling,
>Nohau, etc.), and how fast were you able to get an environment up and
>running,

I bought the Nohau evaluation board (200 U$ AFAIR) mainly to use as a 
sanity check so that when I ran into issues I could check to see if it was 
my board or something in either the tools or my understanding.  I've using 
GCC and GDB with the McCraigor (sp?) Wiggler.  The biggest time to set up 
was getting the tools compiled and configured.  That part probably took 
about a week.  Some of that was my own stumbling around figuring out the 
process.  As far as the environment goes, I still want to spend a several 
days setting up make, lint and TLIB (revision control) for doing 
professional development but that's mostly because I'm starting a fresh 
setup here, I'd normally have them already mostly configured and I'd only 
have to set up the variances.

I'm building up the linker script (to define the registers) as I go along 
but that adds very little time to the process.  I'll share that when it's 
close to complete.

The GNU tools don't come with an IDE (unless you buy a packaged setup) but 
since I've never seen an IDE with a micro development system that didn't 
actively interfere with the development process so I don't see that as a 
drawback.  I'd far rather work with make and good standalone editor.

I haven't pushed the compiler yet so I don't know where its breaking point 
is but so far it's behaved itself.  The GDB/Wiggler combination is not 
quite a friendly.  McCraigor hasn't developed any support for the LPC210x 
yet and the generic arm support doesn't allow breakpoints in the 
flash.  One of the arm variants it supports might be able to do that but I 
haven't gone searching.  It looks as if Nohau's debugger/JTAG ICE supports 
this more completely but I've no experience with it myself.

If you are at all time limited I would definitely recommend getting one of 
the pre-built GNU setups (maybe from Microcross?), or one of the 
professional compilers.  I believe Nohau sells a development system that 
includes a GNU toolkit.  I went for the building path partly for the 
learning experience and partly I cheapened out.  Basically the choice is a 
time vs money one (and development time cost money too, unless your time is 
cheap you are probably far better off buying).

Programming is straightforward with the windows utility but definitely for 
development not production.  The specs are straightforward enough though so 
it doesn't look as if it would be too difficult to design what ever 
download utility was needed.


>2. Which compiler (ARM, GNU, IAR, etc.) have you used, and how well did
>they work,
>
>3. How has your experience with Phillips been?  Last year, they
>basically walked away from an entire portfolio of RF parts leaving a
>bunch of customers with single-sourced EOL notices.  Are they in this
>market for keeps, or for at least a couple of years?

This is not (unfortunately) unique to Philips.  All you can really do is 
make an educated guess as to the size of the market and the developers 
commitment to it.  After that "you pays your money and you takes your 
chances".  Given the apparent reaction to the chip and the number of 
variants they have already produced, things are looking up but not being 
privy to their boardroom talks.....

Robert Adsett