03-15-2016, 08:55 PM
Unfortunately, this is less than ideal. If there are any issues for the readings to get there and they are re-sent, there will always be inconsistent on the actual timestamps, which compromises the data be used for simulations. This topology, unfortunately, won't be useful for us like that with the timestamps being added by the microcontroller. You suggested the EEPROM, but the folks here didn't understand the usefulness of it since there will be no timestamp to be set together to the reading. What were your thoughts here?
Also, could you clarify if the other topology where the sensors are hooked to the I2C Multi POT would address the issue? How fast could we apply the timestamp and keep sampling? Would it be consistent? Unfortunately this way we would need to add microcontrollers all over the place due to the cable sizes, but it is still a better compromise than the timestamp and sampling rat eif that can't be addressed.