J. Kissel, J. Driggers, S. Dwyer

Annoyed by the constant verbal alarms of SUS ETMX saturating while the arms are controlled by ALS, I took a spectra of the L3 / ESD output request. The spectra revealed two things:
(1) We still should have plenty of head room, with the upper ESD quadrants requesting the largest drive, but that drive still only having an RMS of ~20e3 [ct], out of the 128e3 [ct] DAC range and being quite featureless. It makes no sense that we're constantly saturating.
(2) The upper quadrants are receiving more drive than the lower quadrants, which was suspicious because we traditionally don't drive the ESD in angle.

Upon further investigation, we reminded ourselves that there have been off-diagonal, L2A elements in ETMX's L3 / ESD DRIVEALIGN matrix since 2014 (LHO aLOG 15260). As a quick test, we reduced these elements to zero. It had two effects:
(1) It *didn't* affect the ALS DIFF control at all, and we rode happily through the lock acquisition sequence, and 
(2) It reduced the drive a little bit, so the ESD quadrants saturated a little less. It still saturates pretty regularly though.

As such, Jenne's accepted the removal of these elements in the SDF system, and removed the check for them from the DOWN state of the ISC_LOCK guardian. The other good news about this -- two fewer signs to keep track of when flipping the ESD bias sign on ETMX. I wish it would have cured the constant saturation problem though!

EX EX EX EX EX EX EX EX EX EX EX EX 
EX EX EX EX EX EX EX EX EX EX EX EX
EX EX EX EX EX EX EX EX EX EX EX EX
EX EX EX EX EX EX EX EX EX EX EX EX
EX EX EX EX EX EX EX EX EX EX EX EX
EX EX EX EX EX EX EX EX EX EX EX EX ...