Today we ran new experiments to characterize AZ dynamics.
The protocol is as follows:
Pb8 starts at position (175.0,45.0)
AZ movement, constant speed, EL is constant (0.020°/s), 10° range, clockwise
Reference signal: sinusoidal, 2 KHz, 18V rms
Angle acquisition: RDC-19200
The following picture shows the evolution of AZ in blue with the target trajectory (dashed red curve), constant speed value is 0.020°/s:
Now a zoom between 350 and 480 seconds:
We can observe fluctuations on the (measured) trajectory. Again, around 410 s, PB8 seems at rest.
We have estimated the distribution of errors for AZ (differences between the actual trajectory and the target):
If we compare both distributions, EL is much more acceptable than AZ…
We could incriminate the resolvers (the electronic processing is the same, we alternate the inputs), next step…