Post by thread:[PIC] C issue

Decided to (finally) tinker with C programming this weekend, and using the default install of Hi-Tech C (Lite) that came with MPLAB 8.66. I created a basic piece of code to turn on some LED's (multiplexed discrete and 7-segment) on an existing board I have, and it takes a LONG time to start up -- in the order of several minutes, if at all. One a couple occasions it worked right away. So I'm guessing the oscillator is not starting up properly...? I also tried setting the config bits from the Config Bits menu rather than the code and it worked (code ran immediately on programming) the first time, but I can't reproduce that any more. Also tried explicitly turning on Vdd and releasing from reset, but no difference. And the board works perfectly with assembly code. What am I missing here? For reference, this is the code I'm using (16F883).... #include <pic.h> __CONFIG(1, FOSC_INTRC_NOCLKOUT & DEBUG_OFF & WDTE_OFF & LVP_OFF & BOR_ON & CP_OFF & CPD_OFF & FCMEN_ON & IESO_ON & MCLRE_OFF & PWRTE_ON); __CONFIG(2, BOR4V_BOR40V & WRT_OFF); #define FOSC 8000000L /* 8 Mhz internal clock */ void main(void) { INTCON = 0; OSCCON = 0x71; /* 8 Mhz internal clock */ TRISA = 0b00001011; TRISB = 0b11100000; TRISC = 0b00000000; /* PORTC is all outputs */ ANSEL = 0b00000011; /* RA0,1 are analog inputs */ ANSELH = 0; ADCON1 = 0b00010000; /* Left justified, External Vref+ */ OPTION_REG = 0; /* Port-B pullups active */ WPUB = 0b01100000; VRCON = 0; /* Turn off */ T1CON = 0; CM1CON0 = 0; CM2CON0 = 0; CM2CON0 = 0x02; /* Turn off comparators */ PORTA = 0b11111111; /* Turn on all columns */ PORTB = 0b11111111; /* Turn on all rows */ PORTC = 0b10101010; /* Turn on all rows */ /* TODO -- testing... just stop here for now */ for (;;) { } } Cheers, -Neil.

Neil, just to test add external crystal and turn off internal also turn off brawn out reset. see what happens. AA ________________________________ From: PICdude <spam_OUTpicdude3TakeThisOuTnarwani.org> To: .....piclistKILLspam@spam@MIT.EDU Sent: Mon, June 6, 2011 8:52:55 AM Subject: [PIC] C issue Decided to (finally) tinker with C programming this weekend, and using the default install of Hi-Tech C (Lite) that came with MPLAB 8.66. I created a basic piece of code to turn on some LED's (multiplexed discrete and 7-segment) on an existing board I have, and it takes a LONG time to start up -- in the order of several minutes, if at all. One a couple occasions it worked right away. So I'm guessing the oscillator is not starting up properly...? I also tried setting the config bits from the Config Bits menu rather than the code and it worked (code ran immediately on programming) the first time, but I can't reproduce that any more. Also tried explicitly turning on Vdd and releasing from reset, but no difference. And the board works perfectly with assembly code. What am I missing here? For reference, this is the code I'm using (16F883).... #include <pic.h> __CONFIG(1, FOSC_INTRC_NOCLKOUT & DEBUG_OFF & WDTE_OFF & LVP_OFF & BOR_ON & CP_OFF & CPD_OFF & FCMEN_ON & IESO_ON & MCLRE_OFF & PWRTE_ON); __CONFIG(2, BOR4V_BOR40V & WRT_OFF); #define FOSC 8000000L /* 8 Mhz internal clock */ void main(void) { INTCON = 0; OSCCON = 0x71; /* 8 Mhz internal clock */ TRISA = 0b00001011; TRISB = 0b11100000; TRISC = 0b00000000; /* PORTC is all outputs */ ANSEL = 0b00000011; /* RA0,1 are analog inputs */ ANSELH = 0; ADCON1 = 0b00010000; /* Left justified, External Vref+ */ OPTION_REG = 0; /* Port-B pullups active */ WPUB = 0b01100000; VRCON = 0; /* Turn off */ T1CON = 0; CM1CON0 = 0; CM2CON0 = 0; CM2CON0 = 0x02; /* Turn off comparators */ PORTA = 0b11111111; /* Turn on all columns */ PORTB = 0b11111111; /* Turn on all rows */ PORTC = 0b10101010; /* Turn on all rows */ /* TODO -- testing... just stop here for now */ for (;;) { } } Cheers, -Neil.

Yep, I got it from C:\Program Files\HI-TECH Software\PICC\9.81\include. There is also a directory for 9.80, but that did not have the F883, so I used the include from the 9.81 directory. And the executable is C:\Program Files\HI-TECH Software\PICC\9.81\bin\picc.exe anyway. The include file defines "CONFIG1", but not sure where I should use that. I found the syntax of __CONFIG (1, ...) from an online example. And why would it not work if I tell it to use the config bits from the menu/manual selector rather than the code? Cheers, -Neil. Quoting Oli Glaser <oli.glaserKILLspamtalktalk.net>: {Quote hidden}> On 06/06/2011 16:52, PICdude wrote: >> What am I missing here? For reference, this is the code I'm using >> (16F883)... >> >> > > Are you sure the config bits are right? > I just checked in the include file for the 16F883 I have for an early > version of PICC I have on here (9.70) and for the internal oscillator > (no clock out) you put INTCLK. > In case you don't know, all the bits are defined in install > directory->include->pic16f887.h (right at the bottom) > > > >

For now I can't do that -- no crystals where I'm at, and this board is not setup for a crystal anyway. It's a custom board I was using to develop another app. Cheers, -Neil. Quoting Andre Abelian <.....abelian.andreKILLspam.....yahoo.com>: {Quote hidden}> Neil, > > just to test add external crystal and turn off internal also turn > off brawn out > reset. > see what happens. > > AA > > > > > ________________________________ > From: PICdude <EraseMEpicdude3spam_OUTTakeThisOuTnarwani.org> > To: piclistspam_OUTMIT.EDU > Sent: Mon, June 6, 2011 8:52:55 AM > Subject: [PIC] C issue > > Decided to (finally) tinker with C programming this weekend, and using > the default install of Hi-Tech C (Lite) that came with MPLAB 8.66. > > I created a basic piece of code to turn on some LED's (multiplexed > discrete and 7-segment) on an existing board I have, and it takes a > LONG time to start up -- in the order of several minutes, if at all. > One a couple occasions it worked right away. So I'm guessing the > oscillator is not starting up properly...? > > I also tried setting the config bits from the Config Bits menu rather > than the code and it worked (code ran immediately on programming) the > first time, but I can't reproduce that any more. > > Also tried explicitly turning on Vdd and releasing from reset, but no > difference. And the board works perfectly with assembly code. > > What am I missing here? For reference, this is the code I'm using > (16F883)... > > > > #include <pic.h> > > __CONFIG(1, FOSC_INTRC_NOCLKOUT & DEBUG_OFF & WDTE_OFF & LVP_OFF & > BOR_ON & CP_OFF & CPD_OFF & FCMEN_ON & IESO_ON & MCLRE_OFF & PWRTE_ON); > __CONFIG(2, BOR4V_BOR40V & WRT_OFF); > > #define FOSC 8000000L /* 8 Mhz internal clock */ > > void main(void) > { > INTCON = 0; > > OSCCON = 0x71; /* 8 Mhz internal clock */ > > TRISA = 0b00001011; > TRISB = 0b11100000; > TRISC = 0b00000000; /* PORTC is all outputs */ > > ANSEL = 0b00000011; /* RA0,1 are analog inputs */ > ANSELH = 0; > ADCON1 = 0b00010000; /* Left justified, External Vref+ */ > > OPTION_REG = 0; /* Port-B pullups active */ > WPUB = 0b01100000; > > VRCON = 0; /* Turn off */ > T1CON = 0; > CM1CON0 = 0; > CM2CON0 = 0; > CM2CON0 = 0x02; /* Turn off comparators */ > > PORTA = 0b11111111; /* Turn on all columns */ > PORTB = 0b11111111; /* Turn on all rows */ > PORTC = 0b10101010; /* Turn on all rows */ > > /* TODO -- testing... just stop here for now */ > for (;;) > { > } > } > > > Cheers, > -Neil. > > > > >

Hi Neil, My guess is the CONFIG statements are wrong. HiTech has changed the format of them at least four times in my memory. It's also pretty much left to the user to figure out. In the old version I use you would not have the "1" and "2". Just having the statements on separate lines is what it wants. But that's my old version. Cheers, Bob On Mon, 06 Jun 2011 08:52 -0700, "PICdude" wrote: > > __CONFIG(1, FOSC_INTRC_NOCLKOUT & DEBUG_OFF & WDTE_OFF & LVP_OFF & > BOR_ON & CP_OFF & CPD_OFF & FCMEN_ON & IESO_ON & MCLRE_OFF & PWRTE_ON); > __CONFIG(2, BOR4V_BOR40V & WRT_OFF); -- http://www.fastmail.fm - The way an email service should be

On Mon, Jun 6, 2011 at 5:32 PM, Oli Glaser <@spam@oli.glaserKILLspamtalktalk.net> wrote: > Are you sure the config bits are right? > I just checked in the include file for the 16F883 I have for an early > version of PICC I have on here (9.70) and for the internal oscillator > (no clock out) you put INTCLK. > Just as a side note: Since Microchip acquired HiTech, some names and defines changed in the header files, so old projects would not compile at all. Many people actually turnes off from PICC because of that (no reason changing these without backward compatibility). Tamas > In case you don't know, all the bits are defined in install > directory->include->pic16f887.h (right at the bottom) > > > >

Okay, my bad (which I never doubted anyway), and it's now solved. I just noticed a warning that the "__CONFIG(1, ...)" line(s) had an error. Found in the manual that I just need to use "__CONFIG(...)" twice in a row and it will sequentially assign each to the correct config register (so I don't have to explicitly differentiate between the two, other than putting them in the correct order). Cheers, -Neil. Quoting PICdude <KILLspampicdude3KILLspamnarwani.org>: {Quote hidden}> Yep, I got it from C:\Program Files\HI-TECH > Software\PICC\9.81\include. There is also a directory for 9.80, but > that did not have the F883, so I used the include from the 9.81 > directory. And the executable is C:\Program Files\HI-TECH > Software\PICC\9.81\bin\picc.exe anyway. > > The include file defines "CONFIG1", but not sure where I should use > that. I found the syntax of __CONFIG (1, ...) from an online example. > And why would it not work if I tell it to use the config bits from > the menu/manual selector rather than the code? > > Cheers, > -Neil. > > > > Quoting Oli Glaser <RemoveMEoli.glaserTakeThisOuTtalktalk.net>: > >> On 06/06/2011 16:52, PICdude wrote: >>> What am I missing here? For reference, this is the code I'm using >>> (16F883)... >>> >>> >> >> Are you sure the config bits are right? >> I just checked in the include file for the 16F883 I have for an early >> version of PICC I have on here (9.70) and for the internal oscillator >> (no clock out) you put INTCLK. >> In case you don't know, all the bits are defined in install >> directory->include->pic16f887.h (right at the bottom) >> >> >> >> -

On 06/06/2011 18:44, PICdude wrote: > Yep, I got it from C:\Program Files\HI-TECH > Software\PICC\9.81\include. There is also a directory for 9.80, but > that did not have the F883, so I used the include from the 9.81 > directory. And the executable is C:\Program Files\HI-TECH > Software\PICC\9.81\bin\picc.exe anyway. > > The include file defines "CONFIG1", but not sure where I should use > that. I found the syntax of __CONFIG (1, ...) from an online example. > And why would it not work if I tell it to use the config bits from > the menu/manual selector rather than the code? > Right, I see, you're using PICC 9.81 (I mistakenly thought you were using an early version) According to my include file (for the 883) it is FOSC_INTRC_CLKOUT They are all defined at the top of the file. I wouldn't put that 1 and 2 at the start of the arguments, what are they for? I'm pretty sure last time I read the PICC manual it said just to put the config 1 and 2 (if more than one register) in order and it will know which one is which. It should work if you set the config bits in MPLAB though, so it looks like there is another problem. Does it work in MPLAB SIM?

Yep, that's exactly what I figured out, and it works now. Quoting Bob Blick <spamBeGonebobblickspamBeGoneftml.net>: {Quote hidden}> Hi Neil, > > My guess is the CONFIG statements are wrong. HiTech has changed the > format of them at least four times in my memory. It's also pretty much > left to the user to figure out. > > In the old version I use you would not have the "1" and "2". Just having > the statements on separate lines is what it wants. But that's my old > version. > > Cheers, > Bob > > On Mon, 06 Jun 2011 08:52 -0700, "PICdude" wrote: > >> >> __CONFIG(1, FOSC_INTRC_NOCLKOUT & DEBUG_OFF & WDTE_OFF & LVP_OFF & >> BOR_ON & CP_OFF & CPD_OFF & FCMEN_ON & IESO_ON & MCLRE_OFF & PWRTE_ON); >> __CONFIG(2, BOR4V_BOR40V & WRT_OFF); > > -- > http://www.fastmail.fm - The way an email service should be > >

Yep, saw that in the manual, and all works now. The "(1," and "(2," I had found online. However, I'm bugged that it does NOT ignore the config lines in the code, even if I tell it to use the ones from the menu/manual settings. Cheers, -Neil. Quoting Oli Glaser <TakeThisOuToli.glaserEraseMEspam_OUTtalktalk.net>: {Quote hidden}> On 06/06/2011 18:44, PICdude wrote: >> Yep, I got it from C:\Program Files\HI-TECH >> Software\PICC\9.81\include. There is also a directory for 9.80, but >> that did not have the F883, so I used the include from the 9.81 >> directory. And the executable is C:\Program Files\HI-TECH >> Software\PICC\9.81\bin\picc.exe anyway. >> >> The include file defines "CONFIG1", but not sure where I should use >> that. I found the syntax of __CONFIG (1, ...) from an online example. >> And why would it not work if I tell it to use the config bits from >> the menu/manual selector rather than the code? >> > > Right, I see, you're using PICC 9.81 (I mistakenly thought you were > using an early version) > According to my include file (for the 883) it is FOSC_INTRC_CLKOUT > They are all defined at the top of the file. > I wouldn't put that 1 and 2 at the start of the arguments, what are they > for? > I'm pretty sure last time I read the PICC manual it said just to put the > config 1 and 2 (if more than one register) in order and it will know > which one is which. > It should work if you set the config bits in MPLAB though, so it looks > like there is another problem. > Does it work in MPLAB SIM? > >

Wasn't aware that Microchip acquired HiTech, but for the 16F and enhanced 16F, is there any better option? I'm tinkering for now, but will prob want optimizations at some point and wondering if there's a better less-than-$1200 option out there. Especially since I understand that the 16F version is limited to the 16F, and if I want to use if for the 18F or 24F, I have to buy additional versions for those. Cheers, -Neil. Quoting Tamas Rudnai <RemoveMEtamas.rudnaiTakeThisOuTgmail.com>: > Just as a side note: Since Microchip acquired HiTech, some names and defines > changed in the header files, so old projects would not compile at all. Many > people actually turnes off from PICC because of that (no reason changing > these without backward compatibility). > > Tamas

> > Wasn't aware that Microchip acquired HiTech, but for the 16F and > enhanced 16F, is there any better option? I'm tinkering for now, but > will prob want optimizations at some point and wondering if there's a > better less-than-$1200 option out there. Especially since I > understand that the 16F version is limited to the 16F, and if I want > to use if for the 18F or 24F, I have to buy additional versions for > those. > > Cheers, > -Neil. > > This is dangerously close to the age old religious debate of "which is better". It's been hashed over many, many times and I doubt another cycle is going to find any better answers... I use the CCS compiler (CCSinfo.com). The price is better than most. I'm sure some people will swear by it, others will swear at it. I've managed to use it in projects that sold, so it can't be all bad. -Denn

On 06/06/2011 19:37, PICdude wrote: > However, I'm bugged that it does NOT ignore the > config lines in the code, even if I tell it to use the ones from the > menu/manual settings. Glad that it's working :-) You may already know/have tried the following, but just in case: With the above, did you recompile after you set the bits in the settings? If I remember right, you just have to program after setting them, rather than rebuild and program. You get that pop up box telling you that the settings will override your code *until* you rebuild the project.

I would evaluate SoureBoost as well as ByteCraft MPC. Not sure about the pricing and if any of these would meet your requirements, but speak to them and give them a go. Tamas On Mon, Jun 6, 2011 at 7:47 PM, PICdude <picdude3EraseME.....narwani.org> wrote: {Quote hidden}> Wasn't aware that Microchip acquired HiTech, but for the 16F and > enhanced 16F, is there any better option? I'm tinkering for now, but > will prob want optimizations at some point and wondering if there's a > better less-than-$1200 option out there. Especially since I > understand that the 16F version is limited to the 16F, and if I want > to use if for the 18F or 24F, I have to buy additional versions for > those. > > Cheers, > -Neil. > > > Quoting Tamas Rudnai <EraseMEtamas.rudnaigmail.com>: > > > Just as a side note: Since Microchip acquired HiTech, some names and > defines > > changed in the header files, so old projects would not compile at all. > Many > > people actually turnes off from PICC because of that (no reason changing > > these without backward compatibility). > > > > Tamas > > >

On Tue, 07 Jun 2011 16:51 -0300, "Gerhard Fiedler" wrote: > > > It's also pretty much left to the user to figure out. > > That's right. OTOH, looking at the appropriate section in the processor > include file, the respective pages in the processor manual, and the > examples that come with your compiler version (not the internet :), you > can't go too wrong. No, in fact _my_ PICC manual says nothing about the second __CONFIG macro line. Back when I needed to know I had to search the HiTech forum to find it. On the internet :) There have been changes to the manual since, but the change to the __CONFIG macro was a fairly silent introduction. That's the way HiTech has always handled updates, with great email support and out-of-date manuals. The examples that came with the compiler are for old PICs that do not have Config 2. And include files have the definitions of the fuses, not the syntax of the macros that use them. I've been using HiTech since it was beta, so I'm fairly familiar with how their products have evolved and what to do when upgrades break stuff. I don't own any of their new products, but it wouldn't surprise if their documentation hasn't improved. That isn't why people buy their products, but it's probably why some people don't. Bob -- http://www.fastmail.fm - Same, same, but different...

Bob Blick wrote: > On Tue, 07 Jun 2011 16:51 -0300, "Gerhard Fiedler" wrote: >> >>> It's also pretty much left to the user to figure out. >> >> That's right. OTOH, looking at the appropriate section in the >> processor include file, the respective pages in the processor >> manual, and the examples that come with your compiler version (not >> the internet :), you can't go too wrong. > > No, in fact _my_ PICC manual says nothing about the second __CONFIG > macro line. You say at least two things here; one that I can agree with, and one that just doesn't make sense. "in fact _my_ PICC manual says nothing about the second __CONFIG macro line": That may well be, and AIUI I said as much when I started my comment with "that's right". "No": No what? Between "processor include file", "processor manual" and "example code" I don't see how what the PICC manual says or doesn't say has anything to do with what I wrote. Maybe it's just the expression of a negative predisposition? Gerhar

Bob Blick wrote: > On Wed, 08 Jun 2011 08:57 -0300, "Gerhard Fiedler" wrote: > >> Maybe it's just the expression of a negative predisposition? > > No, unlike some people, I'm only a dick when I want to be. Well, you stated here that this is not a reason for the "no", but you didn't really provide the reason why you wrote it. Rather than an attitude, a reason would go a long way. > Is that satisfactory or would you like some more? If you have anything to say about the actual issue, that might be more satisfactory than this here -- but I'm sure you knew this before I wrote it. (To answer your question, you'd of course have to be a tad clearer about what you mean with "some more".) Gerhar

On Wed, 08 Jun 2011 13:42 -0300, "Gerhard Fiedler" wrote: > > On Wed, 08 Jun 2011 08:57 -0300, "Gerhard Fiedler" wrote: > > > >> Maybe it's just the expression of a negative predisposition? > If you have anything to say about the actual issue, that might be more > satisfactory than this here -- but I'm sure you knew this before I wrote > it. (To answer your question, you'd of course have to be a tad clearer > about what you mean with "some more".) I'm not asking for an apology, if you'd just shut up that'd be fine. In the future, if you want to send me insults, do it off-list. Bob -- http://www.fastmail.fm - mmm... Fastmail...

Bob Blick wrote offlist: > I don't know why you bother to write me on list. I don't really write "you" onlist. I write to the list. > Do you even use Hi-Tech PICC for the 16F series? Not currently a lot, but yes, I do have a license for quite a few years, through quite a few versions (even though not the latest), and I used it in a few projects over the years. And since we're in questioning: Did you read what I wrote before responding? All three responses of yours seem to indicate that you didn't, at least not completely and/or thoroughly. I wrote: : OTOH, looking at the appropriate section in the processor include : file, the respective pages in the processor manual, and the examples : that come with your compiler version (not the internet :), you can't : go too wrong. You answered "no", followed by a reason that has nothing to do with what I wrote, and when I asked why "no" you answered with off-topic stuff of the personal nature that generally admins make it their job to stop (but of course not in this case...) -- and again what you wrote didn't have much to do with what I wrote. I understand that you are using this compiler a lot, and actively, so of course it would be interesting if you had anything to say about what I actually wrote. Gerhar

Hey Bob! You forgot the <VBG> at the end of your last sentence! Couldn't resist replying - this was a well-timed moment of levity for me. dwayne At 09:45 AM 6/8/2011, Bob Blick wrote: {Quote hidden}>On Wed, 08 Jun 2011 08:57 -0300, "Gerhard Fiedler" wrote: > > > Maybe it's just the expression of a negative predisposition? > >No, unlike some people, I'm only a dick when I want to be. > >Is that satisfactory or would you like some more? > >Bob -- Dwayne Reid <RemoveMEdwaynerTakeThisOuTspamplanet.eon.net> Trinity Electronics Systems Ltd Edmonton, AB, CANADA (780) 489-3199 voice (780) 487-6397 fax http://www.trinity-electronics.com Custom Electronics Design and Manufacturing

At 10:42 AM 6/8/2011, Gerhard Fiedler wrote: >Bob Blick wrote: > > > On Wed, 08 Jun 2011 08:57 -0300, "Gerhard Fiedler" wrote: > > > >> Maybe it's just the expression of a negative predisposition? > > > > No, unlike some people, I'm only a dick when I want to be. > >Well, you stated here that this is not a reason for the "no", but you >didn't really provide the reason why you wrote it. Rather than an >attitude, a reason would go a long way. > > > Is that satisfactory or would you like some more? > >If you have anything to say about the actual issue, that might be more >satisfactory than this here -- but I'm sure you knew this before I wrote >it. (To answer your question, you'd of course have to be a tad clearer >about what you mean with "some more".) Whoa! Stop! I wrote a few minutes ago saying that I thought that Bob's remark was meant to be funny - and I still do. But I now can see that it might be taken as offensive. I'm pretty sure that was not the intent but . . . I'm not an admin, but can I respectfully suggest that this stop right now? Thanks! dwayne -- Dwayne Reid <EraseMEdwaynerspamspamBeGoneplanet.eon.net> Trinity Electronics Systems Ltd Edmonton, AB, CANADA (780) 489-3199 voice (780) 487-6397 fax http://www.trinity-electronics.com Custom Electronics Design and Manufacturing