??? 10/25/08 15:55 Read: times |
#159352 - To Show My Point Responding to: ???'s previous message |
Here I have built up a simulation of a typical 5V to 3V level translator circuit using NPN transistors that is typical of what you may see in many schematics:
![]() Now here you may see the resulting response of such circuit. Note in particular the long delays seen on the edges related to state transitions where a transistor is turning off. ![]() The delays seen in even this typical circuit are on the order of 2 to 2.5 microseconds. If the bit bang timing being generated happened to depend on signals that were changing at at the same time on the MCU side the delay could easily destroy a setup or hold time at the target side. This needs to be taken into account in the whole implementation strategy. 2.5 microseconds may not sound like much but when you have an interface that wants to clock at say 100kHz this delay ends up being equivalent to 25% if the clock period!! Michael Karas |
Topic | Author | Date |
bit banged spi on ds89C420 for sd card | 01/01/70 00:00 | |
Some notes about the code | 01/01/70 00:00 | |
some more details | 01/01/70 00:00 | |
Have you found the "real" specification? | 01/01/70 00:00 | |
Keil logic analyzer | 01/01/70 00:00 | |
GIGO - verify with the so-called standard | 01/01/70 00:00 | |
I agree ! | 01/01/70 00:00 | |
More comments | 01/01/70 00:00 | |
some answers | 01/01/70 00:00 | |
Have you considered.... | 01/01/70 00:00 | |
3V3 to 5V is easily accomplished ..![]() | 01/01/70 00:00 | |
To Show My Point | 01/01/70 00:00 |