Searching \ for 'ARM7..should' in subject line. ()
Make payments with PayPal - it's fast, free and secure! Help us get a faster server
FAQ page: www.piclist.com/techref/index.htm?key=arm7should
Search entire site for: 'ARM7..should'.

No exact or substring matches. trying for part
PICList Thread
'[PIC] ARM7..should I take the plunge?'
2006\10\27@120206 by Bob J.

picon face
I have several 18F projects under my belt now, all of them written in
assembler.  Life has been good going this route thus far but I have a new
long-term project I am going to be starting on soon and I have been doing
some light reading of the tutorials on ARM7 development over at sparkfun.com.
Once upon a time I purchased a copy of CCS PCWH to do my coding in C and
found that I had less trouble and tighter code writing in assembler, and I'm
sure it would be no different on the ARM7.  I'm afraid the new project might
get too cumbersome to do in assembler.  I'm curious to hear about the
experiences of those here who have migrated their designs to the ARM7 and
what are the pitfalls of using the LPCxxxx devices from NXP...  Not much has
been posted here about the PIC24's so I am curious to hear some opinions on
the advantages/disadvantages/trade offs/comparisons of the ARM7's and the
new PIC24's.  If I do go down the ARM7 path I would like to stick with GCC
and the free tools available, and not spend thousands on compilers.  One
thing I do know that I will miss is the simulator.  Anyhow...it seems that
the ARM chips are getting to be popular.

Regards,
Bob

2006\10\27@123606 by Marc Nicholas

picon face
Hey Bob,

I've been considering the same thing to replace Parallax's SX chips.

I'm currently designing my first ARM-based board around an LPC2103.

Have you look at ARM assembly in detail? There's some interesting "quirks".

-marc

On 10/27/06, Bob J. <spam_OUTrocketbobTakeThisOuTspamgmail.com> wrote:
{Quote hidden}

> -

2006\10\27@134814 by Andre Abelian

flavicon
face
Bob,

I never used ARM7 I am sure it is as good as pic some
times even better then pic but the main question
is how do you know ARM7 is better then PIC18?
here is why I stopped using assembly

1. takes long time
2. each company has their own assembly forever I have to learn assembly
3. updating, changing the chip effects other subroutines timing etc.
4. lots of things needs to remember "errata sheets etc"
5. others do it in C and get it done much faster then the assembly guy.
6. it is good to know assembly and is good for small projects.
7. program memory is cheap so optimization is not important
8. by the way it is good to mix asm with c like "sleep" "nop" etc.


Andre

 

{Original Message removed}

2006\10\27@143157 by John Chung

picon face
Yeah ARM is definitely a must have when the
requirements gets tough. I have not programmed for ARM
yet but the output from the GCC should be decent
enough to be as good as most commercial compilers
thanks to it's architecture. You may want to join
LPC2000 on yahoogroups to determine how well is
LPCxxxx for your requirements. Btw I dislike the
packaging....... The main thing about ARM is it's
stability and realibility which I haven't heard
anything OUTSTANDING.

John

--- "Bob J." <.....rocketbobKILLspamspam@spam@gmail.com> wrote:

{Quote hidden}

> --

2006\10\27@143823 by Wouter van Ooijen
face picon face
> Have you look at ARM assembly in detail?

Yes

> There's some interesting "quirks".

which?

Wouter van Ooijen

-- -------------------------------------------
Van Ooijen Technische Informatica: http://www.voti.nl
consultancy, development, PICmicro products
docent Hogeschool van Utrecht: http://www.voti.nl/hvu


2006\10\27@150216 by Wouter van Ooijen

face picon face
> Once upon a time I purchased a copy of CCS PCWH to do my
> coding in C and
> found that I had less trouble and tighter code writing in
> assembler, and I'm
> sure it would be no different on the ARM7.  I'm afraid the
> new project might
> get too cumbersome to do in assembler.

I smell a contradiction here. Personally I think most projects are
easier in C than in assember, *if* timing and companctness are not major
issues. ARM7 or PIC14/16 are the same in this aspect, although PIC
assembler is probably easier to learn than ARM assembler, but a lot more
difficult to use (I give both PIC and RAM assembler classes).

> I'm curious to hear about the
> experiences of those here who have migrated their designs to
> the ARM7 and
> what are the pitfalls of using the LPCxxxx devices from
> NXP...

When coming from the PIC world:
- exact timing is much more difficult
- 3.3V I/O (5V tolerant)
- no DIP chips

But there are lots of positive points
- fast!!
- lots of peripherals (two UARTs seems to be standard)
- debugger is cheap (if you can live with a parallel-port wiggler)
- bootloader is build-in
- GCC compiler available

> One thing I do know that I will miss is the simulator.  

There are good simulators. I used them only for CPU simulation, I don't
know what they can do with the peripherals.

Wouter van Ooijen

-- -------------------------------------------
Van Ooijen Technische Informatica: http://www.voti.nl
consultancy, development, PICmicro products
docent Hogeschool van Utrecht: http://www.voti.nl/hvu


2006\10\27@153746 by Bob J.

picon face
On 10/27/06, Wouter van Ooijen <wouterspamKILLspamvoti.nl> wrote:
>
> I smell a contradiction here. Personally I think most projects are
> easier in C than in assember, *if* timing and companctness are not major
> issues. ARM7 or PIC14/16 are the same in this aspect, although PIC
> assembler is probably easier to learn than ARM assembler, but a lot more
> difficult to use (I give both PIC and RAM assembler classes).


No contradiction.  At the time I was using PCWH with the original ICD and a
16F877, had so many problems debugging I gave up and started working in
assembler.  I even bought/tried CCS's ICD and that didn't do much better.
On top of that the compiler code was not well optimized (compared to how I
thought my program should work in assembler).  Needless to say my assembler
code consumed much less program memory.  Right now I have very few problems
debugging assembler with the ICD2.

Regards,
Bob

2006\10\27@192132 by William Chops Westfield

face picon face

On Oct 27, 2006, at 9:35 AM, Marc Nicholas wrote:

> Have you look at ARM assembly in detail? There's some
> interesting "quirks".

Heh.  That's for sure.

Don't forget that you have to go three or four deep in manual
"pointing."  I've been playing with the Luminary Cortex M3 parts
a bit, and you go from data sheet to Cortex architecture document
to ARM7 architecture document to ARM Architecture document before
you get a full picture of what "assembly language" is supposed to
look like (I hope.  I haven't actually found an ARM ARM.  I think
it's a book you have to buy.)

I'm not sure that the gnu assembler (gas) that accompanies the C
compiler
is really well-suited toward writing code in pure assembly language.
It tends to being a very generic multiple-cpu sort of thing whose syntax
doesn't match much of anything else, rather lacking in the features
that the average assembly language programmer expects (like macros.)
I haven't gotten far enough in my searching to see if there's a more
capable open-source assembler for ARM (for instance, there are several
for Atmel AVR.)

And assembler seems foreign to the ARM crowd.  One of the things you
DO find at Luminary is an extensive library of primitives for dealing
with all of their on-chip peripherals (and 200+ pages of documentation
for the library!), all written in C.

BillW

2006\10\27@215357 by Christopher Cole

flavicon
face
On Fri, Oct 27, 2006 at 04:21:30PM -0700, William Chops Westfield wrote:
> Don't forget that you have to go three or four deep in manual
> "pointing."

Absolutely, there's LOTS of swimming to do, due to the increased flexibility.

> ...(I hope.  I haven't actually found an ARM ARM.  I think
> it's a book you have to buy.)

You may... the ARM ARM is available at:
http://search.barnesandnoble.com/booksearch/isbnInquiry.asp?EAN=9780201737196

Or you can download it through a link at Altera's website:
http://www.altera.com/literature/lit-exc.jsp
(scroll down to Third-Party Literature)

> I'm not sure that the gnu assembler (gas) that accompanies the C
> compiler
> is really well-suited toward writing code in pure assembly language.

gas is perfectly fine.  The ARM-Linux folks use it, and several bootloader
projects for ARM use gas.

> It tends to being a very generic multiple-cpu sort of thing whose syntax
> doesn't match much of anything else, rather lacking in the features
> that the average assembly language programmer expects (like macros.)

gas (native and cross) supports macros.

> And assembler seems foreign to the ARM crowd.

Which ARM crowd?  One of the most elegant things about the ARM instruction
set is its well thought-out orthogonal architecture... and its rich register
file... Every ARM developer that I know of is very aware of the instruction
set and its use in assembler routines.  Assembler is certainly NOT foreign to
any true ARM developer.

Take care,
-Chris

--
| Christopher Cole, Cole Design and Development               .....coleKILLspamspam.....coledd.com |
| Embedded Electronics and Software Design                  http://coledd.com |

2006\10\27@220211 by Mike Harrison

flavicon
face

>Don't forget that you have to go three or four deep in manual
>"pointing."  I've been playing with the Luminary Cortex M3 parts
>a bit, and you go from data sheet to Cortex architecture document
>to ARM7 architecture document to ARM Architecture document before
>you get a full picture of what "assembly language" is supposed to
>look like (I hope.  I haven't actually found an ARM ARM.  I think
>it's a book you have to buy.)

You can download it here :
http://www.arm.com/community/academy/eulaarmarm.html


2006\10\27@220328 by Mike Harrison

flavicon
face
On Fri, 27 Oct 2006 20:17:30 +0200, you wrote:

{Quote hidden}

Also IAR do a free C compiler with 32K code limit, plenty for the lower end chips.


2006\10\28@030536 by Wouter van Ooijen

face picon face
> Don't forget that you have to go three or four deep in manual
> "pointing."  I've been playing with the Luminary Cortex M3 parts
> a bit, and you go from data sheet to Cortex architecture document
> to ARM7 architecture document to ARM Architecture document before
> you get a full picture of what "assembly language" is supposed to
> look like (I hope.  I haven't actually found an ARM ARM.  I think
> it's a book you have to buy.)

I don't think that is much different to the newer PIC16F datasheets and
the midrange reference manual.

> I'm not sure that the gnu assembler (gas) that accompanies the C
> compiler is really well-suited toward writing code in pure assembly
language.

I have not written that much ARM asm, but who needs to? Mixing ams and C
is IMHO a more fruitfull way.

> And assembler seems foreign to the ARM crowd.  One of the things you
> DO find at Luminary is an extensive library of primitives for dealing
> with all of their on-chip peripherals (and 200+ pages of documentation
> for the library!), all written in C.

Mix asm and C!

Wouter van Ooijen

-- -------------------------------------------
Van Ooijen Technische Informatica: http://www.voti.nl
consultancy, development, PICmicro products
docent Hogeschool van Utrecht: http://www.voti.nl/hvu


2006\10\28@143334 by Orin Eman

picon face
On 10/27/06, William Chops Westfield <EraseMEwestfwspam_OUTspamTakeThisOuTmac.com> wrote:
>
> And assembler seems foreign to the ARM crowd.  One of the things you
> DO find at Luminary is an extensive library of primitives for dealing
> with all of their on-chip peripherals (and 200+ pages of documentation
> for the library!), all written in C.

...much of which resolves into a single hardware register read/write!
A lot of the driver library 'magic' is in the symbolic constants in
the header files.

The library is great for getting things going - use the debug version
and it validates the 'constants' you pass in.

Exact IO timing on the output pins is difficult to say the least.
It's not specified and I went through an interesting exchange on the
question of how long it takes between a write to a port and a read
from the same port... it's on one of the Luminary forums.

Orin.

2006\10\28@164031 by Charles Craft

picon face
Design contest.

Luminary site shows free eval kit with entry.
Circuit Cellar site has a link to reduced price ($39) eval kits.

http://circuitcellar.com/designstellaris2006/

http://www.luminarymicro.com/designstellaris2006



{Original Message removed}

2006\10\28@195308 by Xiaofan Chen

face picon face
To a lot of people, ARM7 class MCUs are the natural replacement for the
existing  8051 MCUs. There are good and faster 8051s out there but they
often cost similar or more than the ARM7 counterparts.

However I was trying to select an ARM7 MCU and found out some problems
(comparing to higher-end 8051s and PIC18F)

1) Power consumption: in general higher
2) I/O driving capability: in general weaker
3) Some ARN7 MCUs like TI TMS470 do not have 5V tolerant I/O. They also
need dual supply (3.3V and 1.8V).
4) Packaging: lack of small package MCUs in general

In the end we still decided to use an ARM7 because of the processing power
needed and the general consensus to move to ARM7. 8051s and PICs will still
be used when processing power is not that critical.

2006\10\29@010946 by Vasile Surducan

face picon face
On 10/27/06, Wouter van Ooijen <wouterspamspam_OUTvoti.nl> wrote:
although PIC
> assembler is probably easier to learn than ARM assembler, but a lot more
> difficult to use (I give both PIC and RAM assembler classes).

Hi Wouter,

Assuming I'm dealing with an ARM926-EJ-S core, what do you suggest I
should read, someting fast and clean (except the ARM documentation)
for being able to program it under C (not assembler) ?


> But there are lots of positive points
> - fast!!

unfortunately not enough...

thx,
Vasile

2006\10\29@031659 by Wouter van Ooijen

face picon face
> Assuming I'm dealing with an ARM926-EJ-S core, what do you suggest I
> should read, someting fast and clean (except the ARM documentation)
> for being able to program it under C (not assembler) ?

If that is an ARM9 I am not familiar with it. But in general I would
say: the vendors documentation of the I/O. If you program in C you can
ignore the CPU, except when speed gets an issue, or when you need to add
interrupt wrappers and suchlike, but they are ofrten supplied by the
programming environment.

> > But there are lots of positive points
> > - fast!!
>
> unfortunately not enough...

Not enough positive points or not fast enough? For some selected tasks
(bit fiddeling) a 75 MHz SX might be a better choice than an ARM. But
for calculations and ARM will be much faster than any PIC.

Wouter van Ooijen

-- -------------------------------------------
Van Ooijen Technische Informatica: http://www.voti.nl
consultancy, development, PICmicro products
docent Hogeschool van Utrecht: http://www.voti.nl/hvu


2006\10\29@083459 by John Chung

picon face
Good to hear that PIC still excels for low end work. I
have to admit it is hard to beat a PIC for low volume
work thanks to the DIP package and less requirements.

After some time I would be using ARM and skip Atmel
b'cos Atmel and PIC is pretty much the same although
Atmel can be easier to develop for in C which is a
good thing.

After that Xilinx for FPGA. Painting a broad brush.

John.



--- Xiaofan Chen <@spam@xiaofancKILLspamspamgmail.com> wrote:

{Quote hidden}

> --


'[PIC] ARM7..should I take the plunge?'
2006\11\01@173424 by alan smith
picon face
And if you need to surround it....Actel has some nice FPGA's that have built in ARM7 cores.

---------------------------------
Low, Low, Low Rates! Check out Yahoo! Messenger's cheap  PC-to-Phone call rates.

More... (looser matching)
- Last day of these posts
- In 2006 , 2007 only
- Today
- New search...