I apparently do not know the MCU in question. Wakeup's can be tricky and often require more detail evaluation to get them working. I suspect that the time pulse method you had suggested does not have this problem?

Does the device require its -INTR input to be low for an extended period of time and then go high to initiate wakeup?

Or is it that the wakeup from power down happens at the onset of a low going pulse on -INTR and then the MCU is held static until the time of the rising edge at which time the code starts in the interrupt routine?

In either case this sounds messy....you set the sleep state somewhere in your code and you wake back up in an Interrupt routine?? Yuk.

Michael Karas