The regular periodicity and constant amplitude of the 'zipper' trace shows it is not a real world effect. Some sort of quantisation effect surely. Although the graph shows segmented vertical traces that may be limitations of the printer producing the graph. Is it every other recorded/printed sample ? Notice that in CONF iture's post the red trace in the bottom graph sows three similar similar blip and again it happens to align upwards/downwards to one of the major y axis grid lines. Suspect the plotting software myself.