Your program is could effectively be refactored as a state machine. It's not a bad overall effort, mind you, but you should do some research on state machines (start by searching this forum).
Some general programming things (this is not everything I see, but I'm a bit pressed for time):
Your FOR loop does nothing. It will either iterate once (in which case you don't need it) or zero times (in which case you REALLY don't need it). And why are you ANDing your "start" button value with a T constant? Just use the boolean value by itself. Same thing.
You can eliminate your local variables and use a shift register instead. (If you don't know about shift registers, that would be another terrific thing for you to learn.) You do some odd things with that "revolutions" value -- why the local write variable? just wire the final value straight to the indicator and get rid of the local write -- and some strange things with your "reset" button value. You're converting the boolean to a number (1 or 0), then comparing it to a value of 1...in other words, you're checking to see if the value is T. The boolean value itself already tells you that. Just use the boolean by itself.
You don't need to invert your "cadence" value before inputting it to the "select" function. Wire the boolean value in, and switch the numeric operations (so that the "+1" is input to the F terminal and the previous value is input to the T terminal). Same thing as what you have, but cleaner.
I don't really understand your "time elapsed" loop within another loop. Exactly what are you trying to accomplish with all that? Is your elapsed time indicator supposed to be somehow related to your DAQ operation? Right now it is not.
Now, your DAQ things:
You are using all of the default values for your physical analog channel, which means that the channel is configured for a range of -5V to 5V (which I'm pretty sure is not what you want) and the wire configuration is set to "default". I don't know what board you're using, so I don't know what the default channel config for that board is, but you do need to make sure it matches your actual wiring. Are you wired as differential? Referenced single-ended? You ask about whether you need to ground. The DAQ board does need some kind of reference in order to measure a voltage.
Have you put a multimeter on your analog input terminals to see if you actually do have a voltage there?
Have you set a probe on the wire to your "generated volts" indicator to see what the DAQ board is measuring on that channel?
You are using "1Channel N Samples" but you don't configure a clock so you won't be grabbing N Samples. Use "1Channel 1Sample" instead.
Currently your "set delay" is set to 15msec. You don't need a loop iteration that fast. Set it to 500msec or thereabouts. You'll grab a DC value and update your front panel every 500msec, which should be more than often enough. You're not measuring a rapidly-varying phenomenon.
Hope some of this points you in a direction. As I said, this is not at all a bad effort. It just needs some cleanup and a better architecture. Everyone starts somewhere and you're doing fine.
