Problem with Timer0 in LPC1769

[vycska]

I configure MR0 = CLOCK*500-1 and MR3 = CLOCK*1000-1 [CLOCK is the cpu frequency in MHz, which in my case is 100]. PR=0. Thus, MR0 corresponds to 0.5ms and MR3 to 1ms, that is the time until match.

I then configure MCR to interrupt on MR0 and to reset counter on MR3.

I should get an interrupt periodically every 0.5ms because of MR0, BUT I get interrupt every 1 ms [precisely; I can vary MR0 and MR3 values, but interrupt always happens on the period according to the MR3 value]. What might be explanation of such behavior?

[kerryzhou]

Hi Vytautas C.

   At first, you must make sure your Timer frequency is correct.

   Do you test the official lpcopen code for LPC1769, which you can find it from this link:

LPCOpen Software for LPC17XX|NXP 

     SystemCoreClockUpdate();
     Board_Init();

     /* Enable timer 1 clock */
     Chip_TIMER_Init(LPC_TIMER0);

     /* Timer rate is system clock rate */
     timerFreq = Chip_Clock_GetSystemClockRate();

     /* Timer setup for match and interrupt at TICKRATE_HZ */
     Chip_TIMER_Reset(LPC_TIMER0);
     Chip_TIMER_MatchEnableInt(LPC_TIMER0, 1);
     Chip_TIMER_SetMatch(LPC_TIMER0, 1, (timerFreq / TICKRATE_HZ1));
     Chip_TIMER_ResetOnMatchEnable(LPC_TIMER0, 1);
     Chip_TIMER_Enable(LPC_TIMER0);

     /* Enable timer interrupt */
     NVIC_ClearPendingIRQ(TIMER0_IRQn);
     NVIC_EnableIRQ(TIMER0_IRQn);
‍‍‍‍‍‍‍‍‍‍‍‍‍‍‍‍‍‍‍‍‍‍‍‍‍‍‍‍‍‍‍‍‍‍‍‍‍‍‍

The match value is timerFreq / TICKRATE_HZ1. timerFreq = Chip_Clock_GetSystemClockRate();

Please check the lpcopen code on your side again.

If you still have question about it, please kindly let me know!

Have a great day,
Kerry

-----------------------------------------------------------------------------------------------------------------------
Note: If this post answers your question, please click the Correct Answer button. Thank you!
-----------------------------------------------------------------------------------------------------------------------

[vycska]

Timer clock frequency is definitely correct. I think sufficient evidence for that is that I get interrupt on whatever period I like by writing appropriate value to MR3, e.g. I assign MR3=CLOCK*4567-1 and get interrupt every 4567us precisely, but I want to get interrupt on match with MR0, which I assign a smaller value, e.g. MR0=CLOCK*2500-1, and I configure MCR accordingly, i.e.  MCR = (1<<0) | (1<<10).

I might give my all timer initialization routine [I do not use lpcopen]:

void Timer0_Init(void) {

   PCONP |= (1<<1);

   PCLKSEL0 = (PCLKSEL0&(~(3<<2))) | (1<<2); //pclk=cclk

   T0TCR = 2; //counters disabled and reset

   T0CTCR = 0; //timer mode

   T0PR = 0;

   T0MR0 = CLOCK*500-1;

   T0MR3 = CLOCK*1000-1;

   T0MCR = (1<<0) | (1<<10); //interrupt on MR0, reset on MR3

   IPR0 = (IPR0&(~(0x1f<<11))) | (1<<11); //interrupt priority

   ISER0 = (1<<1); //interrupt enable

   T0TCR = 1; //counters enabled

}

A few notes also:

- I tried to change Timer0 to Timer1 and get the same results;

- T0IR indicates that the interrupt source is MR0, although interrupt happens on the period according to MR3.

I have a hypothesis that interrupt is asserted only when T0TC is reset to zero and I might think of a few tests that would confirm that, but I'm not sure I will do them, because I achieved my goal with other means than with such a usage of Timer as I described and I have to move forward. I wrote here with the expectation that this is a universal error and that someone would duplicate it and get the same result.