| ??? 02/08/08 08:50 Read: times |
#150460 - Not generally true Responding to: ???'s previous message |
Ap Charles said:
then go ahead , I know the limits of electronics But you don't know the requirements of his application (if he even has a specific application in mind). Therefore you cannot just say - without any proviso or qualification - that a buffer must always be used. As both Richard and Stefan have also said, you must look at the requirements of the application, and the specifications of the particular chip, and then decide whether a buffer is necessary or not. In some cases a buffer is necessary; in others, it is not. making MCU taxing systems and not giving breathing space to systems is not my choise of a design Of course not - but you have absolutely no idea whether or not his system is going to be "taxing", do you? I always keep 30% or more margins So, if he has 30% margin (or whatever) without a buffer, then - by your own reasoning - he does not need a buffer, does he? So it is wrong to just say that he does. My suggestion is in compliance with RE No, it is not: Richard said that it depnds upon the application - you just made a rule without any consideration at all for the application. What if he plans to drive LEDs with this port?. If it is a low-current, high-efficiency LED, it could easily be well within the port's capabilities without the use of any buffer! His application is uncertain. Absolutely - which is precisely why it is wrong to give any unqualified response as if it were a universal rule! |
