Due to the inability so far to damp the 18040Hz mode I have switch the PI damping to the DCPD HF path.

I changed up a few things such as:

• Catches TypeErrors caused by model restarts. I am nervous that this may catch more than I want, so I also added...
• Any error will email me with details. Often times I never even find out that there was a problem until it becomes a bigger one.
• Minor modifications to a few tests .
• Added a test for the PSL AC.
• Changed the way it gets a Guardian node list. I used to have it call "guardctrl nodes" but this would not work on the alarm computer due to different versions of Bash, plus it is just a really clunky way getting that list. Now it just calls a function from the guardctrl client code to get the list directly.

If there is any problem with this version, please don't hesitate to go back one revision.

Attached are plots of the transfer functions measured this morning.  The input modecleaner
was disabled whilst these measurements were obtained.

    Note the frequency servo schematic is D040105

C20F22-1.tif: common gain = 20 dB, fast gain = 22 dB
C20F16-1.tif: common gain = 20 dB, fast gain = 16 dB

C20F22FastMon.jpg: transfer function measured at TP10
C20F22PCMon.jpg: transfer function measured at TP9
C20F22TP8.jpg: transfer function measured at TP8 (should be 10X greater than TP9 measurement)
               common gain = 20 dB, fast gain = 22 dB

C16F22TP8.jpg: transfer function measured at TP8, common gain = 20 dB, fast gain = 22 dB
C16F22TP9.jpg: transfer function measured at TP9, common gain = 20 dB, fast gain = 22 dB

C20F16TP9.jpg: transfer function measured at TP9, common gain = 20 dB, fast gain = 16 dB
C20F16TP8.jpg: transfer function measured at TP8, common gain = 20 dB, fast gain = 16 dB

C20F16TP3Spectrum.jpg: noise measured at TP3 (the mixer monitor, or error point)
                       common gain = 20 dB, fast gain = 16 dB
C20F22TP3Spectrum.jpg: noise measured at TP3, common gain = 20 dB, fast gain = 22 dB

CrossoverSpectrum.jpg: has the two noise spectra overlapped for a visual comparison.

The smoothing factor was set to 0.999 on all CPs. To confirm that the device is still active and responding and not in an "inactive state" (which happens if smoothing factor is set to its limit of 1.0), the LLCV trends over 30 days are posted for the seven CPs (CP3 excluded).

The smoothing factor algorithm (from EPICS integrated into Beckhoff):

if SMOO is 0 or INIT is True, VAL = val
else VAL = val * (1 - SMOO) + Previous_vale * SMOO

*this implements a first-order infinite impulse response (IIR) digital filter with z-plane pole at tau=-T/ln(SMOO), where T is the time between record processing

Attached are screenshots for the past 7 days of optical lever trends for PIT, YAW, and SUM.

This completes FAMIS request 4683.

Peter was in the PSL enclosure and looking at the FSS - his alog will give details.

He's done so I'm starting to lock.