(Jax, Alexa, Arnaud, Chris, Sheila, Daniel, Stefan)

We implemented the DC relief to ETMX for the ALS lock

- Implemented a IPC link from H1:ASL-Y_ETM_LONG to H1:SUS-ETMY_L3_ISCINF_L_IN1.
- Added dewhitening filter to H1:ALS-Y_REFL_SMON FM1 (Zeros 10Hz, 10Hz; Poles 100Hz, 100Hz).
- Zerod the electronic offsets in the CM board B and slow path (H1:ALS-Y_REFL_SERVO_COMOFS=0.505V, H1:ALS-Y_REFL_SERVO_SLOWOFS=-1.675V, H1:ALS-Y_REFL_SMON_OFFSET=-1645).
- Added 1/f filer in stages (Z:P: 10:1:0, 1:0.1, :1) and gain of -0.03 in H1:ALS-Y_ETM_LONG filter bank.
- SUSETMY is in "distributed feed-back mode". We feed back only to the top mass.
- H1:SUS-ETMY_L3_ISCINF_L and H1:SUS-ETMY_M0_LOCK_L have a flat gain of 1.

We also killed a 1.3Hz oscillation in the light from the fiber by putting a 20dB,Q=3 resonant gain into the H1:IMC-L path. All this, together with another alignment tweak, made for very stable locks: while I am writing this log the arm stays locked for 20min at a time while the relief integrator is ticking away (Tidal?). I set the limiter H1:ALS-Y_ETM_LONG_LIMIT to 10000 cts to prevent kicking the Quad too hard - this eventually breaks the lock. 

Sheila and I quickly took a peak, but there is no green beam coming out at the corner yet. Not too surprising,.. next we need to align the IR beam to hit the ETM cage.

Things to do with low priority:
- Decouple force to pitch (& yaw) at DC - the alignment degrades slightly for O(1e4) counts of M0 length drive. But its likely good enough for the HIFO work.