View topic - problem measuring pulslength

Search

problem measuring pulslength

AVR Freaks Forum Index » AVR (8-bit) Technical Forums » AVR forum

Author

Message

seniorvos

Posted: Mar 09, 2010 - 09:36 AM

Joined: Feb 01, 2010
Posts: 4

Hello freaks,
As a newbie I tried to measure a pulslength. According to the Atmel documentation this should be very easy. But I don't get it working. The only thing my program does is counting interrupts that are obviously generated by the program itself, because it does so even when I ground the inputpin.
I try to measure the length of pulses of about 75 to 300 msec. The accucacy is not important.
The programm displays 4 registers on a LCD which should make visible the pulslength and pulspause. The only thing that happen is counting up followed by a reset, for ever. The speed depends on the clock divide factor.
Who can tell me what is wrong in my program? Thanks!
My program:

Code:

; pulse.asm 

.include "m162def.inc"       ; ATMEGA162 4Mc

.org $0000

   rjmp start

.org $0018

   rjmp TIM1_CAPT       ; timer1 event capture interrupt

.org $0030                      

start:

   ldi r16,low(RAMEND)

   out spl,r16

   ldi r16,high(RAMEND)

   out sph,r16            

;init timer1

   clr r16

   ldi r16,1<<TICIE1    ; set timer1 capture event interrupt

   out TIMSK,r16

   ldi r16,(1<<ICES1)|(1<<ICNC1)|(1<<CS10)|(1<<CS12) ;noise; rising; /1024

   out TCCR1B,r16

   ser r17              ; now rising edge r17=ff

   ser r16

   out DDRC,r16         ; port C output

   cbi DDRC,PC0         ; port C pin 0 capture

   sei

loop:

; this loop displays r18,r19,r20 and r21 on a LCD

;   rcall LCD

   rjmp  loop

;timer input capture interrupt

TIM1_CAPT:

   push r16

   in   r16,SREG        ; save SREG

   push r16

   in   r18,ICR1L       ; read ICR low

   in   r19,ICR1H       ; read ICR high

   sbrs r17,1           ; was rising(r17=ff)?skip

   rjmp pulslength

   rjmp pulspause              

pulslength:                  ; event was end of puls

      mov  r20,r18         ; r20/21 pulslength

   mov  r21,r19                

   clr  r16

   ldi  r16,(1<<ICNC1)|(1<<ICES1)|(1<<CS12)|(1<<CS10) ;nonoise; rising; /1024

   out  TCCR1B, r16 

   ser  r17             ; remember rising r17=ff

   rjmp clearandexit

pulspause:                   ;event end of pause

                        ;r18/19 all on pauselenth

   clr  r16

   ldi  r16,(1<<ICNC1)|(0<<ICES1)|(1<<CS12)|(1<<CS10) ;nonoise, decreasing, /1024 

   out  TCCR1B, r16

   clr  r17          ; remember decreasing r17=00

clearandexit:   

   clr  r16

   out  TCNT1H,r16             ; clear timerreg

   out  TCNT1L,r16             ; first high then low

   pop  r16

   out  SREG,r16

   pop  r16

   reti                        

; PM: include LCD display routine

Bogdancev

Posted: Mar 09, 2010 - 10:47 AM

Joined: May 25, 2008
Posts: 177
Location: Dnipropetrovsk, Ukraine

Can you measure a period, from raising edge to raising edge?
Try first to simplify the program. Just capture only raising edges and show the period (quantity of you clock pulses) on LCD. Try to multiply that quantity with your AVR clock period (multiplied by pre-scaler). Is it the same as period of your input signal?
What pulse source do you use?
Do you have function generator?
Show you circuit.
Finally, you can use INT0 or INT1, they have option to trigger on both falling and raising edges at the same time. You don't need to readjust timer each time.
I would do (and I have done) like this: use INT0 or INT1, adjust it to trigger on both edges. Start the timer. When interrupt occurs: save value, zero timer, determine if it pulse or pause, show on LCD. Then wait for interrupt again.

seniorvos

Posted: Mar 09, 2010 - 01:06 PM

Joined: Feb 01, 2010
Posts: 4

What a pleasant surprise to get an answer so soon! I'l give more details. On my roof there are a number of sun collectors. They are based on heatpipes and genereate plenty of hot water. But this system needs a lot of controle. To mention some: the pump should only operate when needed. The system must be protected against frost and overheating. When there is too much hot water this should be used for the central heating of the house. There must be protection against legionella bacterial infection and so on. All this at this momentis done by hand. I intend to do this on behalf of an AVR microproceesor and therefore started to get used to AVR programming. I use AVR Studio-4 and a STK500. My first try was trying to get a LCD working. This went well. Now I try to get used to the timer and interrupts of the AVR. I took a DCF77 clock device (which gives nice pulses) and wrote a test program (see above).
This moment my first aim is rather to understand why my program brings the AtMega162 to generate event-interrupts all by itself then getting my program working, although that would be very nice.
Thanks for your advise!

Bogdancev

Posted: Mar 10, 2010 - 06:08 AM

Joined: May 25, 2008
Posts: 177
Location: Dnipropetrovsk, Ukraine

Code:

cbi DDRC,PC0 ; port C pin 0 capture

Is this your capture pin?

If yes - it is wrong. In your case it must be PE0(ICP1/INT2)
With this pin modelling in AVR studio goes right. Without there are no interrupts occur at all. Your LCD just shows something unknown.

seniorvos

Posted: Mar 11, 2010 - 07:46 AM

Joined: Feb 01, 2010
Posts: 4

Yes, it was yes indeed! I misunderstood the AtMega 162 documentation. Now I have changed the pin assignment to ICP1 the program works fine although it needs some optimalisation because sometimes the program wrongly sees the length of a pause as if it had the length of the preceeding puls. Thank you for your help!

All times are GMT + 1 Hour

AVR Freaks Forum Index » AVR (8-bit) Technical Forums » AVR forum

AVRFreaks.net

Search