Hi the_intern,
First off, you are correct about ch0-ch31. For the NI-6289 port 0 has 32 lines. Correlated digital I/O is still only for the first port, which on some board is ch0-ch7, but for your board it is the 32 lines.
When I was discussing resolution, it was not for simply sending out data or an adress at a particular rate. You discussed earlier that you wanted to measure the width of a pulse. This is why I was bringing up accuracy. If you are simply sending out data at 8MHz then your signal can change high or low at that rate. If you were to toggle the signal high/low at each clock edge you would effectively create a 4MHz square wave.
However if you want to measure the width of a pulse you typically want to sample as fast as possible, i.e. 50MHz. The accuracy is further discussed in this
Tutorial. It is unlikely that the frequency/period to be measured aligns with the signal that is being measured. By this I mean the two signals will not be synchronized. Therefore it is possible to catch an extra edge or miss one. For example the signal could last for 5.5 pulses of the signal that it is being compared to for measurement. But since we can only count edges, we will either count 5 or 6 pulses. Therefore the measurement can be off by a full pulse of your clock frequency. I mis-quoted that it would be a multiple of two, this is actually the amount by which the measurement could vary (i.e. if a signal varies by +/-1 unit, the total variation is 2 units.).
You would not want to measure the width of a signal with a frequency that is similar. If you did, then your error could be close to 100%. By using a much faster signal (i.e. 100 times faster) you can decrease the error (i.e. 1%).
If you are just reading and writing values on the lines, synchronous to an 8MHz clock which is already provided, either device would work. If you want to measure signal widths or place in variable delays of a short duration, the higher frequency board would be a better fit.
Regards,
Jennifer O.
