??? 12/20/06 16:33 Modified: 12/20/06 16:35 Read: times |
#129831 - No calculations, just a compromise Responding to: ???'s previous message |
Richard said:
Was this the result of some set of calculations, and, if so, on what was it based? I just picked 10K as an arbitrary value, and it seemed to work. How did you come up with that value? I've used 10 Meg, 1 Meg, 100K, 47K, 4.7K, and even 1K in various applications, with other MCU's. Very weak internal active pull-up of a typical 80C52 is about 100k, weak pull-up about 10k and strong pull-up is about 100Ohm, highly depending on actual variant, of course. The idea of adding external 10k pull-up resistors is, to have at least something in the range of the weak pull-up always being present at the port line, following the motto "better than nothing". Ok, you can calculate a bit: Typical stray capacitance of port lines in stand-alone applications (the micro with sparse external circuitry) is about 20...30pF, forming a rise time of about 500nsec with 10k pull-ups, which equals about the maximum input signal rise time of 74HC(T)MOS. Smaller pull-ups were wishable, but then port load currents and by this total supply current consumption will rise... By the way, there are 20kOhm pull-ups internally of 74LS373. Nevertheless, it's a plain mystery to me, why your circuit only works with added pull-ups! This makes no sense, because even the 8751H uses push-pull outputs when accesing external memory, so, not at all needing any external pull-up resistors. Kai PS.: Sorry for being late |