Hi GNS,
If you are using an E Series card or a 660x card, you should have no problem making a fairly accurate pulse width measurement. When performing pulse width measurements, your maximum error will be +/- 1 pulse of your timebase (assuming the crystal oscillator error is negligible). If you are using a 20MHz timebase, you will have a measurement resolution of 50ns. 2000 pulses would occur to measure a 0.1ms signal. Your maximum measurement error will be roughly 1/2000.
If you are trying to measure a mechanical source, there could be "bouncing" which will affect your effective pulse width. This is a situation in which you have several small glitches that occur before and after your main pulse. This is typical with push buttons etc. Anyway, these gl
itches could seriously alter your measurement.
The best way to test your measurement system is to use 1 of your counters to generate a 0.1ms pulse and have the other counter measure it with it's maximum timebase. This will indicate your accuracy.
Hope that helps. Have a good day.
Ron