I am not sure if this is a new development or we just never noticed it since 1 step is difficult noticed. But just recently we had a strain gauge mounted to a stage and instead sending a halt command to stop we sent a zero velocity command. We noticed a periodic noise coming from the sensor, which lead us to investigate the stage and that is when we noticed that the sensor was picking up the 1 step movement of the stage. I am able to replicate behavior in NI Measurement and Automation using the 1D-Interaction control. Set it to velocity mode, with a velocity of either 0 steps per second or rpm, press start. Both the position count and position plot will display a 1 step movement.
Based on the specifications of the device (http://www.ni.com/download/labview-development-system-2017-sp1/7099/en/), a velocity command of 0 steps/s is actually unsupported. The velocity range (p. A-2) is 1 to 8,000,000 steps/s. The recommendation is to use the stop command.
Actually the problem was observed to occur even with a non-zero velocity. We saw a one step increase following a certain time, and recurrent. Thus, we tried with a zero velocity and observe it more clearly, but the problem does not come from the fact that the velocity was set to 0.
Just to clarify, even with a non-zero velocity, you are consistently observing a jump of 1 step every 16.4 seconds? Is this with a relatively low velocity?