MrSky,
First, I'll start by saying that I'm not very familiar with Manchester signal decoding. From the small amount of reading I did, it would appear as though you need to distinguish the relative length of pulses to be able to determine the logic level of the next bit in your bitstream. Change detection does not provide any timing information about the line(s) which is changing. You can set up a digital filter to filter out pulses that do not meet a minimum length, but then these transitions would be missing from your data set all together and would not allow you to reconstruct your bit stream.
I think a better alternative might be to use a counter on the 6320 to do semi-period measurement on your signal. From the X Series User Manual, "In semi-period measurements, the counter measures a semi-period on its Gate input signal after the counter is armed. A semi-period is the time between any two consecutive edges on the Gate input." With this information, I believe that it would be possible to reconstruct the digital bitstream.
If you must use the digital lines on the device, I think you could use hardware timing to 'oversample' the signal in question, then post process the data to determine when edges occurred. If you are interested in timing on the scale of 60-120 uS, I would think that you could sample the digital signal at 100 kHz (10 uS resolution), then process this data to find the timing of digital transitions on the signal. Once you have this, I would think that you would be able to reconstruct the digital bitstream.
Like I mentioned, I haven't tried doing this before so I hope that I'm not off base in my understanding of Manchester decoding. However, I think either of the approaches I mentioned should allow you to retrieve the relative timing of consecutive edges on your signal.
Hope that helps,
Dan
