Continued work from klog26902.

Abstract:

We diagonalized ETM TM, IM, and MN actuators with the scripts, which was used before for diagonalization.
TM H1/H2 and IM H1/H2/H3 balances seemed changing a lot (more than 10%).

Detail:

We balanced ETMX actuators on TM, IM, and MN.
balancing was done by scripts, which I developed before.
The balancing method is as follows:

1. Send excitation signals from ETMX_LKIN_{PIT,YAW} oscillators to each coil (10Hz sine signals for TM an 6Hz sine signals for IM and MN).
2. Demodulate TM OpLev signals at that frequencies and calculate the amplitude of OpLev signals at that frequency.
3. Compared the amplitudes excited by each actuators.
4. Caluculate the good gains for each actuators accroding with the obtained amplitudes.

The scripts are not well managed well, s they are in my personal directry (/users/ushiba/scripts/python/).
"coilbalance_TM.py" is for TM actuator balancing.
"coil balance_sample.py" is for IM and MN actuator balancing

Figure 1, 3 , 5 shows the COILOUTF filter banks of TM, IM, and MN before actuator balancing, respectively.
Figure 2, 4, 6 shows the COILOUTF filter banks of TM, IM, and MN after actuator balancing, respectively.
MN actuator balancing doesn't change so much but TM and IM, especially TMH1/H2 and IMH actuaotrs, efficiency changes more than 10%.

Note:

I found the bug of the code of coilbalance_TM.py: somehow the gain that scripts shows are swapped (H1/H2 and H3/H4 gain ratio are swapped).
I will fix the bug later.