Brad Ratto 

The purpose of this aLOG is to update the results from the *now* corrected modeling script used to determine the incurred systematic error as a result from the PUM driver models in the different operating states. This aLOG is also intended to support the conclusions made in (LHO:62152), which, quoting Jeff, are:

• The new and old driver response is still valid and still better than before *this* whole exercise. (It begins here: LHO:61729)

• We should NOT be "newly" compensating for the AOSEM response

In order to keep true to Jeff's word, "Brad's going to work through correcting my mistake in the modeling scripts and redo the conclusions from LHO aLOGs 61729 and 61985." Let me now walk through the results from the updated script. (Some notes on the updated script will follow as well)

In similar fashion to LHO:61729, and as a reminder, the plots are produced using data from the following dates:
ETMX-PUM: 2022-01-14
ETMY-PUM: 2022-01-17
 ITMX-PUM: 2022-01-10
 ITMY-PUM: 2022-01-11

This time around, I will only present the results from ETMX. The full list of plots for all the states and TMs can be found here under the folders with the same dates noted above. To begin, we review the "goodness" of our models versus the measurements taken. As a quick reminder State 1 (no switchable response on) and State 2 (highest range, high gain, "Acquire ON" state), and State 3, "Low Pass ON". 

State 1: 2022-01-14_H1SUSETMX_PUMDriver_S1000343_40Ohm_State1_ALL_CoilCompensationComparsion.pdf
State 2: 2022-01-14_H1SUSETMX_PUMDriver_S1000343_40Ohm_State2_ALL_CoilCompensationComparsion.pdf
State 3: 2022-01-14_H1SUSETMX_PUMDriver_S1000343_40Ohm_State3_ALL_CoilCompensationComparsion.pdf

The color legend:
(BLUE) The the measured data, normalized the unity, we are not interested in the DC gain for this analysis, by dividing out the magnitude of the measured data at the lowest frequency point, 0.2 Hz.
(ORANGE) Shows the full zero-pole response, this includes the super-nyquist zeros and poles.
(GREEN) If we were to ignore the super-nyquist zeros and poles and only compensate with the zeros and poles at sub-nyquist frequencies.

It is important to note that we no longer "fold in" the response of the AOSEM in our measurement for models, hence the *only* difference between the full (ORANGE) and the reduced (GREEN) is the exclusion of super nyquist frequency pole and zeros (As noticed in State 2). However, there is an important note to take into account, opening up the plots for State 3 relieves that there is infact a difference between the full and reduced models when there should NOT be (The reason as to why there should NOT be a difference is because there are NO super nyquist frequency poles present in State 3 to get "rid of"). Rather the difference is in how State 3 is constructed in the modeling script. The "full" model for State 3 is taken as given by the fitter, where as the reduced model is constructed from the measurment fits of State 1 and the LP filter, remember State 3 is simply a superposition of "Low Pass ON" and "Acquire OFF" (State 1). The important conclusion to draw from the State 3 model comparision plots is that it is infact "better" in terms of model accuracy to use the superposition of the indvidal states rather than relying on the fitter, see the right panels of the State 3 plots in which the residual is plotted between the measurment and each of the models.

I will now pivot to showing plots of the modeled magnitude and phase systematic error in the PUM actuation transfer function in State 3 and how it then later affects the response function. See bullets (2)-(5) LHO:61729 for the rationale on how this is done. And again, the results for all the other states and TMs are found here.

The plot for the systematic error as seen in the PUM actuation between the two models is seen here: 2022-01-14_H1SUSETMX_PUMDriver_S1000343_40Ohm_State3_etaPUM_Comparison.pdf

And finally in the response function itself: 2022-01-14_H1SUSETMX_PUMDriver_S1000343_40Ohm_State3_etaR_w_etaPUM_Comparison.pdf

As you can see in the plots for the response function the systematic error in the magnitude is no larger than four hundredths of a percent in magnitude and no larger than six hundredths of a degree in the phase. Armed with these plots and results, it can be said the "old" compenstation method still holds wrt to the PUM drivers.

Some comments regarding the script. I made a few changes so that it is *easier* to use, rather than having four seperate copies for each TM, the script now runs using a #TM#_PUM_config.ini file. See the ETMX_PUM_config.txt as an example of what the format is for ETMX below. A few important notes about the format, when listing poles and zeros there must be a space after the comma, complex numbers must be type WITHOUT any spaces (ie 5+4j) and the coils must be listed in the same order for each state and must have the following order: UL, LL, UR, LR. 

The script and the other TM .ini files live here.