Individual Interrupts on Port B

[lrw_eng]

I am using the 16LF1946 which allows individual interrupts on every Port B pin. There are defines in the device.h for each port pin to set positive (and/or) negative edge.

HOWEVER -- the compiler does not recognize the defined interrupts for individual pins to be used as isr locations (for example) #INT_RB2_H2L. In fact, I believe the only valid isr location is INT_RB which is the "general" interrupt for any change on the port.

What I've found is that you can define active edges for individual pins by doing individual enables for the desired pins. But using the "general" #INT_RB isr location then ruins the settings by setting all edges valid on all pins whenever you enable_interrupts(INT_RB).

Can you override this "feature" and actually ever enable just 3 or 4 pins on the port as interrupt on change with these supplied calls?

[PCM programmer]

[Ttelmah]

I think the key 'error' is in the heading of the thread. "Individual interrupts on Port B". The chip does not support this. It supports _one_ interrupt only here. INT_RB. What you can have is individual _masking_ of the pins, and which direction they trigger this interrupt in. The enable_interrupt instruction, handles this. So:

enable_interrupts(INT_RB2_H2L);

sets up the masking, so the INT_RB interrupt is only triggered on this edge.

enable_interrupts(INT_RB2_H2L | INT_RB3);

will enable the interrupt to trigger on the high to low edge of RB2, and both edges on RB3.

Now, the chip sets separate flags in the IOCBF register, according to what bit actually triggered the interrupt. On some compiler versions (with varying degrees of success), CCS read this, and test the individual flags to then optionally vector you to separate handlers. Personally I'd say you are much better to do this yourself. Just have a single INT_RB, then read this register, and run the code for the input(s) you want (and remember to clear the bit when finished).

So, whether it will work with separate handlers is version dependant.

Best Wishes

[lrw_eng]

Thanks to all for the clarifications.. The larger part of my confusion was whether making calls to enable INDIVIDUAL pins was taking care of the RBIE bit in the interrupt control register properly.. From the assembly listing above --- it is clear that it does.

So #INT_RB does all of the flag clearing for RBIE.. And I will set all the INDIVIDUAL flags to zero in the IOC flag reg anyway.

I'll post a listing in the other forum when I've got this working.

Thanks again...