Post by thread:[PIC] 16F84 & Math

I am new to this list. I am trying to have a little bit of experience with PICs. I have two questions in my first post: 1) How can I do math operation in 16F84. I am using CCS. 2) Whats the best alternative for 16F84 now-a-days. I beleive that this question would have been asked many times, but my focus is on now? Regards, --------------------------------- Ahhh...imagining that irresistible "new car" smell? Check outnew cars at Yahoo! Autos.

Adam Kt wrote: > 1) How can I do math operation in 16F84. I am using CCS. If you use C language, it will take care of all the details for you. You may need to optimize/fine tune your code then, if speed is an issue. > 2) Whats the best alternative for 16F84 now-a-days. I beleive that 16F628/648, IMO. -- Ciao, Dario -- ADPM Synthesis sas - Torino -- http://www.adpm.tk

I can't comment on the CCS compiler, since I don't use it. But, as to the best alternative to the 16f84, I'd say the cheapest chip that does what you want! I often do a sort of all the chips in the family based on price, then work my way down the list until I find the first one that has all the peripherals I need for a particular project. Of course, if there's one we're already using that is suitable and not too much more expensive, I'll go with that to get our volume up and inventory costs down. Harold {Quote hidden}> I am new to this list. > I am trying to have a little bit of experience with PICs. > I have two questions in my first post: > 1) How can I do math operation in 16F84. I am using CCS. > 2) Whats the best alternative for 16F84 now-a-days. I beleive that this > question would have been asked many times, but my focus is on now? > > Regards, > -- FCC Rules Updated Daily at http://www.hallikainen.com - Advertising opportunities available!

> 2) Whats the best alternative for 16F84 now-a-days. every 'best' question needs a contect. what do you want? depending on that context the answer can vary from 'still 16f84, but take the -A version' to 16F648A, 16F88, 16F505, 16F877A, 18Fxxx, 30Fxxxx, to AVRxxxx or even LPCxxxx. And that list is non-exhaustive. Wouter van Ooijen -- ------------------------------------------- Van Ooijen Technische Informatica: http://www.voti.nl consultancy, development, PICmicro products docent Hogeschool van Utrecht: http://www.voti.nl/hvu

> 2) Whats the best alternative for 16F84 now-a-days. The 16F88 is nice in that it's pin-compatible with the '84 and has quite a few peripherals (like an internal oscillator) that can make setup easier. And, since the '84 has typically been the de-facto educational PIC, a jumping up to the 88 is nice. Watch out, though, the sheer amount of multiplexing that's going on in all of the pins can cause some problems, too. The 16F877A is often the development chip of choice because it has lots of pins and more peripherals. Again, it all depends on what you're doing and how you want to go about it. As a hobbyist (and for small projects at work), I keep around the most versatile chips of each series, and begin programming projects with them. If it turns out that I want a smaller package or a different feature set, then I'll go after that. But I'm typically very content with the 16F88 for the small footprint, the 16F877A for the pins, and the 18F2620/4620 for heavier lifting. Matt

On 4/23/07, Matt Bauman <spam_OUTm-baumanTakeThisOuTnorthwestern.edu> wrote: > > > As a hobbyist (and for small projects at work), I keep around the > most versatile chips of each series, and begin programming projects > with them. If it turns out that I want a smaller package or a > different feature set, then I'll go after that. But I'm typically > very content with the 16F88 for the small footprint, the 16F877A for > the pins, and the 18F2620/4620 for heavier lifting. > > Matt I would jump straight to the 18 series for small projects where the small cost difference won't matter.

Note that the F88x family is a closer match to the other newer 16F's such as the 12F863 (8-pin), 16F866 (14-pin) and 16F88 (18-pin). The 28-pin and 40-pin'ers from the 16F88x series fits in nicely in that list. Same internal architecture with nanowatt, 8 Mhz intosc and so on. The 877A (and it's rellatives) are a bit outdated today... Regards, Jan-Erik. Matt Bauman wrote: {Quote hidden}>> 2) Whats the best alternative for 16F84 now-a-days. > > The 16F88 is nice in that it's pin-compatible with the '84 and has > quite a few peripherals (like an internal oscillator) that can make > setup easier. And, since the '84 has typically been the de-facto > educational PIC, a jumping up to the 88 is nice. Watch out, though, > the sheer amount of multiplexing that's going on in all of the pins > can cause some problems, too. > > The 16F877A is often the development chip of choice because it has > lots of pins and more peripherals. Again, it all depends on what > you're doing and how you want to go about it. > > As a hobbyist (and for small projects at work), I keep around the > most versatile chips of each series, and begin programming projects > with them. If it turns out that I want a smaller package or a > different feature set, then I'll go after that. But I'm typically > very content with the 16F88 for the small footprint, the 16F877A for > the pins, and the 18F2620/4620 for heavier lifting. > > Matt >

Adam, On Mon, 23 Apr 2007 08:46:47 -0700 (PDT), Adam Kt wrote: > I am new to this list. Welcome! I hope you have some spare time, because reading this can become a full-time hobby! :-) > I am trying to have a little bit of experience with PICs. > I have two questions in my first post: > 1) How can I do math operation in 16F84. I am using CCS. Can't help there, I'm not a C-person. > 2) Whats the best alternative for 16F84 now-a-days. I beleive that this question would have been asked many times, but my focus is on now? This is like asking for directions in London - if you ask 12 people you'll get at least 13 different answers... My advice would be the 16F88, as it's pin-compatible with the '84 and since you're learning you'll find a lot of examples using the '84. You will have to tweak the software (especially the initialisation, to turn off the new default devices that the '88 has and make it look similar to the '84 - things like AtoD and comparators tend to be on but you need them off to use '84-designed software) but at least the circuits will work as they are (mostly)! 16F628 and '648 are also pin-compatible, and have fewer things to disable, but these days I think they are getting more expensive than the '88 for less capability. Have a look at http://www.amqrp.org/elmer160/lessons/index.html - it was mostly written some time ago so uses the '84 as its target, but at least read Appendix A even if you aren't going to follow the lessons, because there's a lot of useful stuff in there. Cheers, Howard Winter St.Albans, England

On Mon, 2007-04-23 at 14:10 -0300, CristÃ³vÃ£o Dalla Costa wrote: > I would jump straight to the 18 series for small projects where the small > cost difference won't matter. At the very least. Personally I'd recommend people go straight to the 30F parts. There are several 28pin DIP 30F parts available (the package size I find best for most of my hobbyist projects). The sheer number of features and huge code and ram spaces make these parts REALLY easy to work with. You don't have to worry about squeezing your code into the part, or being able to run fast enough. Alot of people shy away from the 30F and 33F parts because the letters "DSP" are associated with them, but if you don't use the DSP instructions, this PICs are very much like all the other pics. I treat them as really big and powerful pics, haven't even played with the DSP features yet. TTYL

Adam Kt >> 2) Whats the best alternative for 16F84 now-a-days. >> You could consider a 16F883. Is cheap, fast and powerful. yeah, it has more pins but... you will love it. Wouter van Ooijen wrote: > every 'best' question needs a contect. what do you want? depending on > that context the answer can vary from 'still 16f84, but take the -A > version' to 16F648A, 16F88, 16F505, 16F877A, 18Fxxx, 30Fxxxx, to AVRxxxx > or even LPCxxxx. And that list is non-exhaustive. > > Wouter van Ooijen > > I think 16F887 can be a real replacement for the 16F877A. Is cheaper (it costs the half or less), it has internal oscillator that can run from 31Khz upto 8MHz. It has more AD channels, usart with autobaud, nanowatt technology.

> I think 16F887 can be a real replacement for the 16F877A. depends on the context. if existing www support (project pages) is a requirement the 16F84, 16F628 and 16F877 (in that order) are probably still the favourites. OTOH if the OP was looking for a PIC on pure technical grounds probably none of those would be on it. That's why I aswked for context... Wouter van Ooijen -- ------------------------------------------- Van Ooijen Technische Informatica: http://www.voti.nl consultancy, development, PICmicro products docent Hogeschool van Utrecht: http://www.voti.nl/hvu

Herbert Graf wrote: {Quote hidden}> On Mon, 2007-04-23 at 14:10 -0300, CristÃ³vÃ£o Dalla Costa wrote: > >> I would jump straight to the 18 series for small projects where the small >> cost difference won't matter. >> > > At the very least. > > Personally I'd recommend people go straight to the 30F parts. There are > several 28pin DIP 30F parts available (the package size I find best for > most of my hobbyist projects). > > The sheer number of features and huge code and ram spaces make these > parts REALLY easy to work with. You don't have to worry about squeezing > your code into the part, or being able to run fast enough. > > Alot of people shy away from the 30F and 33F parts because the letters > "DSP" are associated with them, but if you don't use the DSP > instructions, this PICs are very much like all the other pics. I treat > them as really big and powerful pics, haven't even played with the DSP > features yet. > > TTYL > > This sounds like fine advice... my question for you though is what do you code with... is there a 'student' version of C30? As a hobbyist, the C18 toolsuite with MPLab and PIC18F parts makes a whole lot of sense. Is there a similarly priced suite available for the PIC24, PIC30 and PIC33 parts? What do you use? I guess I could ASM it, but that does not make the PIC3* parts more attractive.... Rolf

> Personally I'd recommend people go straight to the 30F parts. Personaly, I think that's a bit to far for a startup-hobbyist. The lack of online resources is a major factor. > There are several 28pin DIP 30F parts available As today with a number of PIC24 devices. And then you do not have to worry about the "DSP" letters... :-) Regards, Jan-Erik.

part 1 743 bytes content-type:text/plain; (unknown type 8bit not decoded) > > Personally I'd recommend people go straight to the 30F parts. > > Personaly, I think that's a bit to far for a startup-hobbyist. > The lack of online resources is a major factor. > > > There are several 28pin DIP 30F parts available > > As today with a number of PIC24 devices. And then you > do not have to worry about the "DSP" letters... :-) iirc PIC24 chips are all j series with the horrible flash endurance that brings wheras DSPIC33 chips have the conventional long life flash. No virus found in this outgoing message. Checked by AVG Free Edition. Version: 7.5.463 / Virus Database: 269.5.9/773 - Release Date: 22/04/2007 20:18 part 2 35 bytes content-type:text/plain; charset="us-ascii" (decoded 7bit)

On Mon, 2007-04-23 at 16:34 -0400, Rolf wrote: > This sounds like fine advice... my question for you though is what do > you code with... is there a 'student' version of C30? Yup, the student/free version does everything the full version does, except one type of optimization (I think procedural abstraction?). The result is code is around 30-50% bigger then with the full version. For a hobbyist this is rarely a problem. It's also in my experience less "quirky" then C18 is, which means it's FAR less quirky then any of the 16F C options. The beauty as a hobbyist of using a part far beyond what you need is you can use an HLL and not worry as much about being careful about code space, ram space or pin requirements. You can learn how to do things, instead of the frustrating effort sometimes needed in smaller parts at bypassing things to make them fit. For professional projects obviously different issues rule. TTYL

On Mon, 2007-04-23 at 22:48 +0200, Jan-Erik Soderholm wrote: > > Personally I'd recommend people go straight to the 30F parts. > > Personaly, I think that's a bit to far for a startup-hobbyist. > The lack of online resources is a major factor. Not really. Sure, there are tons of 16F84 projects out there, but in my experience having referenced them it can get very frustrating how poorly documented some of them are. Beyond the "hello world" project of blinking an LED, I don't see that much of an advantage to the "online resource" factor (in fact, seeing how poorly some are done it might be better for the beginner to skip them completely instead of giving up in frustration because an online example was just wrong). > > There are several 28pin DIP 30F parts available > > As today with a number of PIC24 devices. And then you > do not have to worry about the "DSP" letters... :-) True, except until very recently there were no DIP versions. Also, many of them are only 3.3V, a BIG problem for a hobbyist IMHO (although most of my recent hobbyist designs are 3.3, beginners should start at 5V simply because of the larger peripheral selection still needing 5V, i.e. LCD modules). Another wonderful advantage of the 30F line is voltages are less important, there is no "LF" version like with the 16F/18F parts. You want to run at 3.3V or 5V? Go ahead, only thing to worry about is max available OSC speed at the lower VDD (again, usually not a problem for a beginner since they'll invariably likely start with internal RC OSC). TTYL

part 1 464 bytes content-type:text/plain; charset=windows-1250; format=flowed (decoded 7bit) peter green wrote: > iirc PIC24 chips are all j series with the > horrible flash endurance that brings wheras DSPIC33 > chips have the conventional long life flash. The datasheet for the "PIC24FJ64GA004 FAMILY" (DS39881B) says : > - 10,000 erase/write > - 20-year data retention minimum Which of them is the problem here ? I could definitely live with both "limits"... Jan-Eirk. part 2 35 bytes content-type:text/plain; charset="us-ascii" (decoded 7bit)

On Tue, 24 Apr 2007, Jan-Erik Soderholm wrote: > peter green wrote: > >> iirc PIC24 chips are all j series with the >> horrible flash endurance that brings wheras DSPIC33 >> chips have the conventional long life flash. > > > The datasheet for the "PIC24FJ64GA004 FAMILY" > (DS39881B) says : > >> - 10,000 erase/write >> - 20-year data retention minimum Those smaller (44 pin and fewer) 24F parts are the exception. All of the other 24F (the 64 pin and larger), and all of the 24H are 1,000 cycle. And note that the 10,000/1,000 specs are "typical"; the "minimum" is 10% of those figures. -- John W. Temples, III