conf_tilt_meter
af_offset, af_scale, lr_offset, lr_scale, stime
Configure the azimuth tilt meter.
Double af_offset
Double af_scale
Double lr_offset
Double lr_scale
Double stime
conf_tilt_meter 0.0, 0:1:00, 0.0, -0:1:00, 0:0:2.7
The 3rd and 4th arguments tell the control-system that the left-right tilt meter also has no zero offset, and that a +1 volt increase in the output of this tilt-meter, corresponds to -1 arcminute change in the clockwise rotational tilt of the azimuth platform, around the horizontal vector that points outwards from the azimuth axis towards the current pointing azimuth.
Finally, the 5th argument in the example, tells the control system to apply a low-pass filter with a 10-90% rise time of 2.7 seconds, to smooth the tilt-meter readings that are used to dynamically correct the telescope pointing.
cal_tilt_meter
command.
In practice, since each one-second of telescope motion is calculated one second in advance, the tilt meter reading that is used to calculate a given second of motion, is taken from 1 second before the start of that motion. So the actual frequency of the ringing in the current system is slower than 1Hz.
To prevent the tilt meters from introducing instabilities,
it is necessary to smooth their readings, to suppress the
transients that provoke the oscillations. This is done using
a simple exponential filter whose time constant can be set
with the conf_tilt_meter
command. The time
constant is defined as the time taken by the filter to reach
90% of its final value, after a step in the tilt-meter
readings.
af_tilt = af_offset + af_scale * voltsThe signs and values of af_offset and af_scale must be chosen such that positive values of af_tilt correspond to tilts that raise the elevation of the telescope bore-sight, and negative values to tilts that lower the elevation of the telescope bore-sight.
lr_tilt = lr_offset + lr_scale * voltsThe signs and values of lr_offset and lr_scale must be chosen such that positive values of lr_tilt correspond to tilts that turn the azimuth platform clockwise around the horizontal vector that points outwards from the azimuth axis towards the current azimuth.
temp_tilt = temp_offset + temp_scale * volts