Jeff K., Evan G.

We wanted to make the CAL-CS DARM model to be as explicit as possible to avoid issues or confusion involving sign conventions in the model of the DARM control loop in the front end (nevertheless it is still confusing and requires some critical thinking). To that end, we created a schematic diagram to illustrate where sign flips originate from DARM_CTRL to differential length (see T1800456). This document shows the O2 feedback layout (driving all three stages of ETMY) and compares it to the current feedback layout (L1 and L2 on ETMY and L3 on ETMX).

Sign issues that were resolved last week (see LHO aLOG 44766) were added in the digital gain of the DARM ETM analog filter bank, but to be explicit on where sign flips enter, we moved these signs into filter modules with gain(1) or gain(-1), depending on the appropriate sign for that actuator and arm.

As an example, for ETMX L3:

1) there is currently an ESD bias of -9.5 V (with the sign and magnitude ignored in the model because the linearization is bypassed), so we include in FM6 a "biassign" value of -1. If the real bias is flipped to +9.5 V from it's current value then the real DRIVEALIGN gain is also flipped to compensate. Based on previous experience, the calibration team decided against changing the CAL-CS model (see LHO aLOG 29868) for future flips in the ESD bias, and we will continue this convention (e.g., don't be alarmed if ESD bias and DRIVEALIGN gains in CAL-CS don't match reality, but they should match in sign). Also note that if the CAL-CS model is ever fixed so that the ESD bias sign and value is not bypassed, then we would need to remove the biassign filter. 
2) since the ESD is exerts a pulling force, then FM7--"actsign"--gets a value of -1. All ESDs will have actsign=-1 while pushing actuators (OSEMs) get a value of +1. 
3) since we actuate on ETMX, the result is that the x-arm gets shorter, so there is a negative effect on deltaL = L_x - L_y, the value of FM8--"armsign"--gets -1.

We updated L1, L2, and L3 stages for both ETMX and ETMY, and accepted SDF changes. Screenshots are attached for each filter bank.