J. Kissel, K. Izumi, K. Kawabe

Corrections to gamma calculation
After battling through a little bit of complex number analysis in Simulink, I was able to create a revised calculation of the slowly, time-varying, optical gain correction to the IFO's DARM sensing function for the production of calibrated DARM displacement in the front end. What I had originally implemented (see LHO aLOG 17102) had ignored that the change in open loop gain transfer function at the calibration line frequency is a complex number and so what just, in general wrong. 

See attached screenshot 2015-03-05_CAL_CS_GAMMA_LINE1.png for the current implementation, and see the bottom half of LHO aLOG 17102, which is still valid, for what we intend to *do* with the infrastructure.

Notes:
- GAMMA is the ratio of the current DARM open loop gain transfer function at a given calibration line frequency, G, and the same transfer function at a reference time, G0, is ideally real and unity. It is this real part of GAMMA that is fed into the DARM block to correct the fluctuation in optical gain.
- Ideally, the GAMMA should be frequency independent, and we've implemented the correction as though it were frequency independent, i.e. the output of GAMMA calculated for ONE line should be selected in the GAMMA_MATRIX. However, offline, we'll want to *confirm* that the correction is frequency independent, so all four calibration lines have EPICS outputs and Fast Channel Test Points (even though at the moment, we're only planning to turn on two; see LLO aLOG 15870).
- The AMP EPICs input of each line (which again should be synchronized with that line's DEMOD's OSC CLK GAIN) should be in units of DARM_CTRL counts; Joe's matlab script referenced in LLO aLOG 15944 should take care of converting the desired amplitude to get an SNR of 30 in whatever current DARM sensitivity we have.

Updates to CAL_CS_MASTER
- I've cleaned up the DARM block in the CAL_CS_MASTER block, and added descriptive text that describes the output as we're currently intending to use it, i.e. as what's described in LLO aLOG 16421. See 2015-03-06_CAL_CS_DARM.png.
- I've added a DAQ channel list that will now store channels for the auxiliary degrees of freedom which we'll eventually calibrate. The policy is to store all ERR, CTRL, and SUM signals at the full data rate, but only store the SUM in the science frames. I had considered storing the CTRL channels at a lower rate in the commissioning frames, but since they're only used for commissioning the channels, and one almost always wants to put both CTRL and ERR on the same plot, out to the frequency desired for the ERR signal, I leave them sampled at the same rate. Again, this only impacts the commissioning frames. See 2015-03-06_CAL_CS.png
- I've added whitening filters for all of the auxiliary DOFs CTRL, ERR, and SUM channels. If we have dynamic range / double precision issues with DARM, I'm sure we'll eventually have similar problems with these DOFs.
- I've added EPICs readback of all ERR, CTRL, and SUM channels, and changed the naming convention to "MON" instead of "SLOW." This is so we can get displays of these channels on the overview screen which we'll evenutally have to create.

Added HARDWARE_INJ Library Part to Top Level h1calcs.mdl
I've updated the 
${userapps}/cal/common/models/ 
directory, such that we've received the new Hardware Injection library part. It's as described in LLO aLOG 17120, but in summary, the new feature is an ODC bit.