18F or 16F for newbies in '05 ?

Well, one school of thought is to learn the 'old' stuff as it is time proven and an abundant of information is available... ... the other thread is to learn the 'latest and greatest'...

Problem is , in the middle of learning about the latest, some new fangled 'better than the last' chip hits the market....

As the PICs are based on a 'core' of silicon, it might make better sense to rally delve into the 877 and then when really,really reqired go to the next level....

I'm still cutting code for C84 and F877s today 'cause if it ain't broke you don't HAVE to fix it and the code is rock stable. that being said, someone will argue that I should get out of the 'stoneage' and use the newer stuff.problem is, this stuff makes me good coins,as is, bullet proof and no 'oops a new feature crashes the code....'

For all my needs I have no need to 'migrate' up the ladder. speed is not an issue,features ? well I don't like USB ( not realtime) and the E-net stuff isn't applicable.

Every chip / tool has it's place....I'm happy making coins my way with the 'old' stuff......

Jay

From what I have seen of C compilers for any of the PICs, you need to do wierd things just to make the PIC code come out a certain way. In assembler, you just tell it exactly what to do and it does it.

I would not want to program in assembler for large scale computer (PC/MAC) software, but for microcontrollers, I still like it the best. I really want to know just what is going on microsecond by microsecond.

But then, thats me.

--
Luhan Monat (luhanis 'at' yahoo 'dot' com)
"The future is not what it used to be..."
http://members.cox.net/berniekm

I don't think a C compiler is a cure-all, it does nothing to decouple you from the PIC peripherals/interrupt structure. That said, the 16F series major hangup is banking program memory making it painful to call outside a segment. The 18F "solves" this by making every GOTO or CALL a two-word instruction... sort-of defeating the simplicity.

The extra 18F instructions may look good on paper at addressing the shortcomings of the 16F architecture, but really they solve no real-world problems.

16F is great for "small" problems where you'll never go beyond a single segment (or can divide up the problem between segments once and not worry about it again.)

For real C support and a truly nice orthogonal 16-bit architecture, I think you have to go to the TI MSP430. You don't have to use a "student" compiler to get a free development platform, there are true open-source gcc/binutils toolchains readily and widely available.

Tim.

I mostly agree. My attitude is that if you think you need more memory, you didn't design the system boundaries well :-). (Or you didn't go toe-to-toe with the product manager who thinks you need all 10000 features and modes in your gadget..)

Tim.

I mostly agree. My attitude is that if you think you need more memory, you didn't design the system boundaries well :-). (Or you didn't go toe-to-toe with the product manager who thinks you need all 10000 features and modes in your gadget..)

Tim.

My favorite part - PIC12F675: 8 pins, 6 I/0, 4 analog, 1ma supply, about $1 in modest quantities. Imagine the possiblilites...

--
Luhan Monat (luhanis 'at' yahoo 'dot' com)
"The future is not what it used to be..."
http://members.cox.net/berniekm

I have to disagree about the 'bank switching' problems with the little PICs. I've been in this game since the 4004 and ,well, you just have to get down, read, program,reread, and reprogram..Heck,the beast has ONLY 35 instructions! Compared to a Z80( 100+) let alone a 80xxx, it is child's play to use PICs. Sure there are others that may be easier to program but after a month or so you can cut good tight asm code on a PIC. For 'most' applications you don't NEED to have more memory,more periperals,more speed,more .....whatever.I've 'stuck' with the PICs using both ASM and CCS C as they do what I need and want to do.You'll be better off choosing one of the many and staying with it,instead of bouncing to this one, that one everytime someone says, 'oh this is the greatest chip since slice bread!' Afterall a chip is only as good as it's programmer. Jay

Thank you all!!

I appreciate and value the perspectives of people of experience. I agree that I should learn to master the core in assembly first.

BTW, those $1.00ish PICs are amazing.

If you want to make small/cheap gadgets you will have to use the 14 or

12 bit core: there are no realy small chips in the 18F series.

yeah, and the datasheets keep getting bigger :(

Wouter van Ooijen

-- ------------------------------------

Webshop for PICs and other electronics Teacher electronics and informatics

Would you prefer to live life as a Lion or as a sheep?. Go the compiler route and you'll become just an 'operator' of this months kiddy programmers idea of what a PIC is to be used for. Go join then the massed flocks of other consumers, bleeters and chewers of grass. Never then dare complain that it's tough earning a living, when having to compete against iron man hackers who have come up the hard way, know how to bitbang a UART and can price in a job at a quarter of your quote. PIC Silicon is pretty much all the same. There's just more of it on the larger chips. Nothing is lost and everything to be gained by starting with an 8 pin PIC and learning how to wring it's neck. Once you've lost the will to live, reading the truly appalling, PIC datasheets, figured out the sadistic bit settings, understand the elegance and beauty of a 'GOTO' and can make the device dance to your tune, then you'll be ready for anything (you'll be a man my son :). All else will be an anticlimax. Yes ... Choose then, if you so wish, to use a compiler. But do so now from a position of strength, with the sure knowledge of that which you choose is that which is good and not that which is third rate. regards john