We reran the single bounce OMC scan now that the RF power levels are back to where they ought to be.

Results

Peak Power in OMC DCPD Sum milliamps
Carrier = 13.656 mA
9 MHz   =  0.159 mA
45 MHz  =  0.223 mA

Estimated Current Modulation Depths
Γ9  = 0.215 rad
Γ45 = 0.255 rad


The old modulation depths from the month of July were

Γ9  = 0.12 rad
Γ45 = 0.20 rad

as seen from our previous single bounce OMC scan.

Mostly a note to self:  When aligning SRY, AS_C pointing isn't trust-worthy.  AS_A_SUM values got to about 5500 cts when the pointing on AS_C was far from center, from 4600 cts when AS_C was centered.  I did SRY with only SRC1 loops closed, and moved SR2 to get better buildup.  We should pico AS_C.

We repeated the PRMI-locked OMC scan.  Power recycling gain is much lower...

Measured 45 MHz Power Recycling Gain with Arms ~ 9

Modeled 45 MHz Power Recycling Gain with Arms = 91

Single Bounce OMC Scan
From our single bounce OMC scan today (see comment above for plot), the 45 MHz sidebands are 7 times more powerful than our July 25th scan:

Today's Single Bounce OMC DCPD Sum 45 MHz peaks = 0.22 mA
July 25 Single Bounce OMC DCPD Sum 45 MHz peaks = 0.03 mA

These sidebands were well-balanced, so I report them together.

PRMI with Arms Scan
However, our PRMI-locked, Arms locked and off resonance OMC scan shows only 2.5 times more power: (see plot 1, all measurements referenced to OMC DCPD Sum milliamps)

Today's PRMI with Arms -45 MHz peak = 1.86 mA
Today's PRMI with Arms +45 MHz peak = 1.25 mA

July 25 PRMI No Arms -45 MHz peak = 0.78 mA
July 25 PRMI No Arms +45 MHz peak = 0.67 mA


This is super weird because we would expect to see 14 times as much power in 45 MHz with arms locked: the factor of 7 from the sidebands, and a factor of 2 from the extra reflectivity of the arms.  
The 45 MHz sidebands are even more unbalanced than before: 59% of the power is in the lower sideband.

I plotted the July 25 and today's PRMI scan on top of one another in plot 2, to see if we could see anything obvious, or we weren't actually looking at the 45 MHz sidebands.  There wasn't anything obvious to me...

PRMI No Arms Scan
I unlocked the arms and misaligned the ETMs and redid the scan. (Plot 3)
When I unlocked the arms, the POPAIR_B_RF18 and RF90 powers both fell to 60% of their "with arms" power when I relocked with the same alignment.  We expect half the power, so this is at least close.


Today's PRMI No Arms -45 MHz peak = 1.10 mA
Today's PRMI No Arms +45 MHz peak = 0.76 mA

This gives us an abysmal transmission gain of 1.67 when compared to PRMI No Arms from before.  I averaged the two sidebands' calculated PRMI gain together:

Measured 45 MHz Power Recycling Gain No Arms ~ 5.3

Modeled 45 MHz Power Recycling Gain No Arms = 48

PRMI Model with Arms
The measured PRMI gains of 9 come from the same PRMI math as before, only with arms we have to use a different Michelson reflectivity/transmission and flip a bunch of signs.  
The model of PRMI with Arms parameters are shown below.

PRMI Sideband Gain with Arms:


Michelson Transmission:

where

, including ITM losses of 60ppm

Hang and I had a look and it turns out the single bounce OMC scan was done with 9.5 watts input power, while the PRMI-locked scans were done with 1.9 watts input power.  
This factor of 5 makes our scan results believable:

Measured 45 MHz Power Recycling Gain with Arms ~ 44

Modeled 45 MHz Power Recycling Gain with Arms = 91

Measured 45 MHz Power Recycling Gain No Arms ~ 27

Modeled 45 MHz Power Recycling Gain No Arms = 48