For future reference, just post the pictures instead of putting them into a Word document. If you have more than 3, then zip up your files. Many people will avoid MS Office files since they can contain viruses.
This doesn't look like a noise issue. It is a phase issue. Meaning, when you start acquiring data. If you told the DAQ to acquire the right number of samples to get EXACTLY one period, then you shouldn't see this. But that is really impossible to do.
What you might want to do is change your graph into a chart. The chart keeps a history of the signals it was given. That will give you a better idea of what is happening. I would also recommend performing a Mean and RMS measurement on your waveform to prove that it isn't changing.