 |
 |
| View previous topic :: View next topic |
| Author |
Message |
asmboy
Joined: 20 Nov 2007
Posts: 2128
Location: albany ny
|
| CCP compare, frequency generation for the 18f4520 [ints req? |
 Posted: Tue Sep 13, 2011 9:12 am |
 |
|
The idea being to use CCP1 as a simple divide by 'n', repetitive
frequency generator based on a variable divisor of the internal clock.
in practice the mode i'm struggling to setup is to use timer 3 with a divide by 1 prescale , compared against the ccpr1H / ccpr1L register pair to generate a frequency output on CCP1 - sans interrupt mediation. Just runnign autonomously based on the correct hardware setup for a 50% duty cycle at various frequencies.
so far i can't seem to make "the pig cross the street"
i think i have a combination of 3 different 'lack of understandings' on:
1- if i have timer 3 setup correct
2- if i have ccp1 setup right
3- if there is a CCS function for properly setting ccpr1H / ccpr1L
in the correct order so as to get the divisor i want
and have not been able to find an example file on simple frequency generation
using ccp compare either.
| Code: |
setup_timer_3(T3_INTERNAL|T3_DIV_BY_1|T3_CCP1_TO_2);
setup_ccp1(CCP_COMPARE_RESET_TIMER);
|
Last edited by asmboy on Tue Sep 13, 2011 11:04 am; edited 1 time in total |
|
 |
Ttelmah
Joined: 11 Mar 2010
Posts: 19506
|
|
| Posted: Tue Sep 13, 2011 9:42 am |
|
|
First, you need to be using compare mode, toggle output on match, not the reset timer selection (this still resets the timer.....). Read the data sheet. Register 15-1.
So:
| Code: |
setup_timer_3(T3_INTERNAL|T3_DIV_BY_1|T3_CCP1_TO_2);
setup_ccp1(CCP_COMPARE_INT_AND_TOGGLE);
|
The 'reset timer' mode, doesn't change the output pin.
You need to ensure the CCP pin is programmed as an output (tris bit = 0).
Then setting the CCP registers to give different times, is just a matter of using:
CCP_1 = val_you_want;
Best Wishes |
|
|
asmboy
Joined: 20 Nov 2007
Posts: 2128
Location: albany ny
|
|
| Posted: Tue Sep 13, 2011 11:03 am |
|
|
MUCH improved but i wish that did it.
i still have a bizarre problem with:
| Code: |
#include <18f4520.h>
#fuses HS,NOPROTECT,NOMCLR,NOWDT,NODEBUG,NOLVP,NOIESO,
#fuses NOFCMEN,PUT,NOWRT,NOFCMEN,BROWNOUT,BORV42,
#fuses CCP2C1,NOPBADEN,NOLPT1OSC,noxinst,
#use delay(clock=16588800) //
#include <DISR576.h>
#use rs232(baud=57600, xmit=PIN_C6, rcv=PIN_C7, ERRORS )
#use fast_io(c);
output_c (0); set_tris_c ( 0b11000000 );
setup_timer_3(T3_INTERNAL|T3_DIV_BY_1|T3_CCP1_TO_2);
CCP_1=1023; // set for base freq
setup_ccp1(CCP_COMPARE_INT_AND_TOGGLE);
void showccp(void){ printf("C=%LU\r",CCP_1); }
void setf(void){
CCP_1 = atol(ArgumentString); // 0-65535 string input
showccp();
}
|
when i first call showccp - it reports 1023
when i call setf() - after that - showccp reports any new value i enter for
CCP_1 and keeps on reporting that correct value as often as i ask.
the LST file shows that the proper registers are being loaded.
BUT: PIN_c2, output CCP1 -ONLY generates a constant 31.646 HZ ,
which is my internal clock (HS xtal div 4)
further divided by 131072 ( 65536*2)
on my DSO screen - all i see is a faint flicker in the waveform when i set a new CCP_1 value - but at least that is consistent
do i have to clear an #INT - & mess with registers to make this
work right ?
its as if the working compare register is only set ONCE when i actually change the value - but after first compare success - it all breaks down
and reverts to using 0xFFFF as the compare toggle trigger
i am still missing something important here |
|
|
asmboy
Joined: 20 Nov 2007
Posts: 2128
Location: albany ny
|
|
| Posted: Tue Sep 13, 2011 2:00 pm |
|
|
so as i so much feared -
i had to handle the CCP1IF
#INT and do a set_timer3(0); for every compare --- to generate nice square waves of my choice. ---
too bad the INT overhead effectively crashes the PIC for
CCP_1 values less than about 18.
so as far as USEFUL freq rates - no higher than about 2 khz with a 16.5888 mhz clock seems the order of the day - with less than 200 HZ being the sweet spot.
even at ccp_1=31 - the interrupt density is so excessive - there is no useful CPU time left to do anything else BUT generate signals
i have yet to discover any setup that does NOT require #int , intervention to reset timer3 - to keep freq generation going ;-)) |
|
|
|
|
You cannot post new topics in this forum
You cannot reply to topics in this forum
You cannot edit your posts in this forum
You cannot delete your posts in this forum
You cannot vote in polls in this forum
|
Powered by phpBB © 2001, 2005 phpBB Group
|
FAQ 
Search
Register
Profile 
Log in to check your private messages 
Log in
