| ??? 10/09/10 11:06 Read: times |
#179004 - Obvious, but _you_ singled out propagation delay Responding to: ???'s previous message |
Yet again you try to kick in an open door by claiming something obvious that has already been covered.
As already covered earlier: 1) Make rough tests for inputs/outputs (may miss tristated outputs that requires specific stimuli to leave high-impedance state) 2) Use brute-force or intelligent processing of test patterns to find which chips it can not be. 3) When reaching a leaf of the decision tree, you have zero, one or multiple chips possible. - if zero, the chip is broken, or unknown, or the rule set is buggy - if only one chip is matching, it's meaningful to run through full rule set for that chip - the decision tree can manage to eliminate all-but-one without needing to test the full rule set. And the suggested chip can still be wrong for the reason that the world have more chips with same logic table but who are not part of the tester database. - if multiple chips are found, the chips are so similar that the tester hardware can't discriminate them. * If tester can't load outputs, it can't see difference between 74HC4017 and a 4047 * If tester can't vary voltage to inputs, the tester can't discriminate between schmitt-trigger inputs * If tester can't vary voltage to inputs, it can't separate 74HC00 from 74HCT00. * If tester can't measure output voltages, it can't ... * If tester can't measure propagation delay, it can't ... * If tester can't measure power consumption, it can't ... * ... It has several times been covered that what results you get from a tester depends on what abilities you give it. All depending on amount of time/money to invest in it. And what the expected outcome is. I see a tester as covered in this tread as a hobby project, since the potential users are hobbyists. Potentially a school project, in which case the student can make own decisions on what capabilities to add based on how good "score" they want when the teacher is grading the project. For some reason, you Richard decided to misunderstand the use of an XOR circuit to measure propagation delay. And after having singled out propagation delay, you now inform everyone that just knowing propagation delay isn't meaningful. I would personally put more weight on knowing if inputs have TTL or CMOS logic levels or if outputs have TTL or CMOS logic levels. And that the outputs aren't too weak. For a huge percentage of projects a hobbyist may play with, just about any logic family will be fast enough as long as inputs and outputs are compatible and the circuit doesn't fail from lack of fan-out. |
