Hi Raghu,

yes, interrupts of huge 600msec duration are really bad practice and must be avoided in general.

But I understand absolutely not, why you want to wait 600msec inside an interrupt ?

You get only one interrupt on every changing from high to low level. Then store the timer value and leave the interrupt immediately. This schould only need some µsec and nobody can see, that a multiplex interrupt was delayed of about 10µsec.

After the second changing from high to low, you get the next interrupt. Then store this timer value also and set a flag to the main.

Then the main substract both values and voila, the period time was measured.


Peter