Exact match. Not showing close matches.
'[PIC] PIC24FJ16GA002 vs PIC16F886'
Forrest W Christian
I was just looking through the product list on the microchip site and
came across the PIC24FJ16GA002...
This looks like a really interesting 16 bit part.
The interesting part is that it is only slighty more expensive than the
16F886 which I use extensively, but of course has limits.
Most of the things I'm using the '886 for work perfectly well in it -
however, they of course would be easier to deal with in the PIC24 line
since things like 32 bit counters mean I don't have to deal with writing
code to deal with counter overflows as often, etc.
I am a bit confused about a few things... In the 8 bit pic world, the
program memory is specififed in words, which is a really good measure of
program code size... I can't seem to find any similar information on
the PIC24... That is, is 16Kbytes on the PIC24 similar to 8KWords on
I note the PIC24 is 3.3V only, which is not a problem - my designs are
slowly ending up 3.3V anyways.
What I'm trying to figure out is what exactly (if anything) I am giving
up if I move to the PIC24 part listed in the subject..
On Sun, 2007-12-30 at 17:31 -0700, Forrest W Christian wrote:
> I am a bit confused about a few things... In the 8 bit pic world, the
> program memory is specififed in words, which is a really good measure of
> program code size... I can't seem to find any similar information on
> the PIC24... That is, is 16Kbytes on the PIC24 similar to 8KWords on
> the PIC16.
Comparing program memory size between differing architectures is often
To start off, IIRC the program word size on the 24F is 24bits, so that
16Kbytes is about 5.33kwords. So, with just that, the 24F can "do less".
However, note that we're talking about a 16bit architecture, so stuff
like 16bit arithmetic can be handled naively on the 24F, while on the
16F you need much more code to get the same job done.
On top of this, you've got a hardware mult unit, which can also drop
code size, along with other enhanced features (i.e. nowhere near as much
In the end, my "feel" is that 5.3kwords on the 24F would in most
circumstances allow you to do at least as much as 8kwords on a 16F, but
this all depends on YOUR code.
Only way for you to tell definitively would be to give each a try and
see what you get.
> I note the PIC24 is 3.3V only, which is not a problem - my designs are
> slowly ending up 3.3V anyways.
Me too, in fact it's funny how often I forget that 5V is still out
there. Recently wanted to interface to an LCD. I didn't even consider
whether I needed a 5V supply until I read through the datasheet and
discovered the LCD needed 5V! Fortunately it's IOs are TTL, so I can get
away with running the IOs on 3.3.
> What I'm trying to figure out is what exactly (if anything) I am giving
> up if I move to the PIC24 part listed in the subject..
There are surely some good things you might be giving up, but whether
they are important or not is up to you. Offhand, none of my recent
designs have used anything less then a 24F.
At 07:31 PM 12/30/2007, you wrote:
The "J" parts have no data EEPROM, presumably due to a process
incompatibility, but in many cases you can easily add an external serial
(SPI, I2C or Microwire) EEPROM if NV storage is required. Available pins
might be an issue on such a small package.
The 5V units (eg. PIC30) are generally regarded as being more "bulletproof",
or at least that's what I hear from Microchip's engineers.
Spehro Pefhany --"it's the network..." "The Journey is the reward"
interlog.com Info for manufacturers: speffhttp://www.trexon.com
Embedded software/hardware/analog Info for designers: http://www.speff.com
On Dec 31, 2007 8:42 AM, Herbert Graf <farcite.net> wrote: mailinglist4
> On Sun, 2007-12-30 at 17:31 -0700, Forrest W Christian wrote:
> > What I'm trying to figure out is what exactly (if anything) I am giving
> > up if I move to the PIC24 part listed in the subject..
> There are surely some good things you might be giving up, but whether
> they are important or not is up to you. Offhand, none of my recent
> designs have used anything less then a 24F.
Something to give up:
1. Flash endurance for PIC24 is not as good as the PIC16F including 16F886.
2. EERPOM, PIC24 does not have EEPROMs. Emulated EEPROM from Flash
may or may not be good for you.
3. Typically you will want to use C instead of assembly for PIC24. Of course
some assembly experts may disagree but that it the case for most people.
C30 is said to be quite good. C30 full version is not expensive. C30 student
version is free and does not have code size limit. It is based on GCC but
Microchip adds a non-free Procedure Abstraction optimizer on top pf GCC.
4. There are less people using PIC24. Some people may prefer ARM7TDMI
and Cortex M3 over PIC24, especially for higher pin-count part.
5. ICD2 may be a bit too slow for PIC24 but it might be usable. ICD2 is
good enough for PIC16F. Real-ICE is recommended for PIC24 debugging.
There should be more. But you do gain on serveral things as well.
PIC24 is much more powerful and nicer architecture and C friendlier than
PIC16. The price seems to be right as well.
On Dec 31, 2007 9:00 AM, Spehro Pefhany <interlog.com> wrote: speff
> The 5V units (eg. PIC30) are generally regarded as being more "bulletproof",
> or at least that's what I hear from Microchip's engineers.
That is normally the case. The old PIC16C is quite bulletproof. PIC16F
is worse. The pure 3.3V part will be worse. This has something to
do with the semiconductor process and cost control.
But processor is only partly the issue for the EMC/EMI performance
of the overall system.
More... (looser matching)
- Last day of these posts
- In 2007
, 2008 only
- New search...