??? 02/20/07 11:47 Read: times |
#133303 - byte programming time Responding to: ???'s previous message |
The datasheet says, byte programming time is 8us. Even if this most probably falls into the "typical" values, and even if this appears to be the net programming time and IAP adds a software overhead significantly bigger than that, it's still several orders of magnitude less than the WDT timeout (which is in the order of milliseconds with typical crystals).
Thus, IMHO it is fair to assume that it's safe to feed the dog before calling IAP without the risk of inadvertent WD reset. OTOH, what happens if a watchdog reset would fire in the middle of statusbyte programming? If it is not programmed at all yet, the application would fire up again, so you can activate NoTouch again. And if it's programmed, NoTouch worked and the bootloader gets active right away. Of course, using the watchdog feed in IAP would doubtlessly contribute to the feeling of security... 0.02SKK Jan Waclawek PS. Lynn, may we know WHY are you asking this? |