Sounds like a tricky app. There tend to be a lot of frequency components in rotational measurements on geared shafts. These will influence your measurements. Also, inferring wear mass via flank tolerance sounds like a very subtle measurement requiring very high signal to noise. So I think you've got your work cut out for you from a physics standpoint.
In the data acq world, however, maybe I can help. First, the terminology. It sounds like you've got some reference position that produces 1 pulse per rev. You'd like to grab 1 sample of position on each of these pulses. You're calling this pulse the "trigger." However, in the data acq world, that sort of operation would usually call the pulse an "external sampling clock."
Step 1: Make sure you're treating that 1-per-rev pulse as an external sampling clock, not as a trigger.
Next, how's your resolution? An 18k incremental encoder is pretty high-resolution, but is it really enough to detect the small angular difference caused by mass loss through wear? Have you worked out the geometry to know the linear distance (at the pitch diameter) represented by 1 encoder count? For good trending and measuring, it's typical for the measurement device to have 10x the resolution that you need to discriminate.
A different approach to try to gain resolution might be to measure time between pulses. Depending on your board, you can get either 20 MHz or 80 MHz time resolution. Then, depending on your linear speed at the pitch diameter, you can figure out how much linear distance is represented by 1 time count.
-Kevin P.
ALERT! LabVIEW's subscription-only policy came to an end (finally!). Unfortunately, pricing favors the captured and committed over new adopters -- so tread carefully.
