16Bit timer : ATmega8535

[lord loh.]

I am trying to simulate the 16bit timer on Atmega8535. I have enabled the global interrupt enable and also the Interrupt on Timer1 overflow. My interrupts are also working fine but the TOV1 flag is never set. I put the break points at the right places to see the flag conditions on AVR Studio. If I manually set the flag, the interrupt occurs. But undernatural free running simulation, the OCF1A (Compare Register A match) flag is set. However the OCF1B is never set. I do not understand the reason for this behaviour.

Please help.

The code

main.asm

.EQU Xtal=8
.NOLIST
#include<m8535def.inc>
.include"macros.inc"
.LIST

.CSEG
;Interrupt Vector Table
.ORG 0x000
		rjmp main		;Reset Vector
.ORG 0x008
		rjmp timer1_ovf		;Timer1 Overflow
.ORG 0x00E
		rjmp adccomplete	;ADC conversion Complete.

;Main Subroutine
.ORG 0x015
main:
		ldi r16,HIGH(ramend)	;Init. Stack pointer
		out sph,r16
		ldi r16,LOW(ramend)
		out spl,r16
		bset SREG_I		;Enable Global Interrupt
		rcall loadtimer
		ldi r16,$04		;Enable Timer1 Overflow Interrupt.
		out timsk,r16
		ldi r16,$00
		out tccr1a,r16
		ldi r16,tccr1b_value	;Start Timer1
		out tccr1b,r16


terminate:	rjmp terminate		;END Programme here.

;Timer1 Overflow Interrupt service routine
timer1_ovf:
		nop
		nop
		reti

;ADC Conversion Complete service routine
adccomplete:
		reti

;Other Subroutines
loadtimer:
	;F424 counts @8MHz and prescale 64 accounts for 0.5s
	;0DBD to FFFF accounts for F424
		ldi r16,$0D
		out tcnt1h,r16
		ldi r16,$BD
		out tcnt1l,r16
		ret

macros.inc

//Half Second count Values @ some prescaler settings
.IF (xtal==8)
.EQU COUNT=0x0DBD
.EQU PRESCALE=0x03			;prescale 64
.ELIF (xtal==1)
.EQU COUNT=0x0DBD
.EQU PRESCALE=0x02			;prescale 2
.ENDIF

.SET tccr1b_value=PRESCALE

 

[hjames]

I cut and paste your code into the simulator and set a breakpoint at the first line of the timer overflow interrupt. The first interrupt happens at ~496msec, and takes roughly ~25 seconds to simulate on my relatively recent computer.

The TOV1 flag will only appear for a 1-2 cpu cycles, and is immediately cleared when the interrupt is called.

As for the OC1FA, OC1FB issue, I think that's a bug in the simulator. I guess having 3 seperate events happening at the same time is confusing it a bit.

 

[lord loh.]

The first interrupt happens at ~496msec, and takes roughly ~25 seconds to simulate on my relatively recent computer.

How on earth did you arrive at this figure? 496ms. @8Mhz, I computed my count to give me exactly .5 sec... I am new to AVR studios, but am familiar with a stop clock in MPLAB IDE.

I put my breakpoint at the interrupt vector and still did not find the TOV1 set. What version of AVR are you using? Mine is 4.12 service pack 2 Build 472

As for the OC1FA, OC1FB issue, I think that's a bug in the simulator. I guess having 3 seperate events happening at the same time is confusing it a bit.

Pleasing to know that there is a bug in the simulator and not my code... I discoverd a similar in MPLAB IDE. It was not simulating an interrupt for a particular timer at a particular prescaler.

 

[hjames]

I set a breakpoint at the beginning of the interrupt itself- putting breakpoints into jump tables strikes me as a little hazardous.

As for the timing number:
.5 * 8000000 / 64 = 62500
65536-62500 = 3036 = 0xBDC which is a character swap from the number you have.

personally I don't use the simulator or IDE much - I just downloaded this for the first time today. Most of the time I'm using avr-gcc on a Linux box, and just try to 1) incrementally code and test in hardware 2) get it right the first time

 

[lord loh.]

Oh!

I did 65,535-62500

I'll correct that... Thank you. I am new to AVR and have taken to simulationg... But I am aware of simulation errors. I faced it in PIC where the programme just would not simulate as expected but worked clean on the device.

Thank you.