| ??? 03/06/07 01:03 Modified: 03/06/07 01:13 Read: times |
#134331 - It looks quite simple to me Responding to: ???'s previous message |
Andy said:
Yes, even the so-called "bible" is unhelpful on this point! You doubt about the "bible"?? Don't you think that the "bible's" authors knew what they wrote? The reason why the "bible's" authors talk about "inputs" is, that Intel, some decades ago, actually wanted to design something that can be both, an "input" or an "output", and this by only setting a control bit to "1" or "0". And I think they didn't anything wrong with the implementation of this idea. Have a look at the 74245, for instance. Here also a port can be either an input or an output, and this by setting a control line to "0" or "1". And like the 8052 port circuitry the input and output circuitry are forever linked together and the output is turned-off by putting it into tristate mode. More, with the 8052 an additional pull-up feature is realized, to prevent unconnected inputs from oscillating and wasting power. Andy said:
Many (most? all?) other processors have a specific register (or registers) that must be specifically configured to make an IO either an Output or an Input. Nevertheless, their inputs and outputs are also forever linked together and the outputs are also put into tristate mode by a control bit. So, where is the difference, that allows you to define, that the one micro has a true input and the other not? The only strange thing with this bi-directional port topology of 8052 is not, that input and output are forever tight together and that it's switched between input and output configuration by putting the output into tristate condition (sorry, high ohmic pull-up condition), but only that it contains a sophisticated pull-up consisting of three differently strong active pull-ups (pFETs) and that the strongest pull-up is turned-on for only two oscillator periods. I again say, that the "bible's" phrase "configuring a port as input" is exactly hitting the nail on the head. You turn-off the output by putting it into high ohmic condition and the port looks like a true input. And it makes no difference, that you need a select bit in the SFR to switch-off the output for this. Such a control bit you will always need, equally how the switching is implemented finally. Andy said:
This is why I (among others) say that it is unhelpful to think of "configuring" an 8051's IO pins as "inputs" - it is far more helpful to think of them as permanently Quasi-bidirectional, but just having the output driver in the 'Off' state. Nevertheless, you finally must call it an input, when the user needs to use it as an input. So, what's wrong to call it an input then? Kai |
| Topic | Author | Date |
| the RESET pin | 01/01/70 00:00 | |
| Reset and Inputs | 01/01/70 00:00 | |
| thanks | 01/01/70 00:00 | |
| No, that's not true. | 01/01/70 00:00 | |
| Picky, Picky, Picky | 01/01/70 00:00 | |
| not picky Lynn, you need to read | 01/01/70 00:00 | |
| I stand corrected | 01/01/70 00:00 | |
| after reading the link | 01/01/70 00:00 | |
| not entirely | 01/01/70 00:00 | |
| I got the point | 01/01/70 00:00 | |
| That's funny... | 01/01/70 00:00 | |
| not exactly | 01/01/70 00:00 | |
| Please read again... | 01/01/70 00:00 | |
| Missing the point | 01/01/70 00:00 | |
| It looks quite simple to me | 01/01/70 00:00 | |
| but it is also an output | 01/01/70 00:00 | |
| There's no excuse! | 01/01/70 00:00 | |
of course not! | 01/01/70 00:00 | |
| I agree | 01/01/70 00:00 | |
| She? | 01/01/70 00:00 | |
| haha | 01/01/70 00:00 | |
| You\\\'re allowed to do whatever you want, | 01/01/70 00:00 | |
| Depends... | 01/01/70 00:00 | |
| so now what has been presented several | 01/01/70 00:00 | |
| I did read | 01/01/70 00:00 |