J. Kissel, J. Driggers, J. Wright

While we recover the alignment into to the IMC using the amount of DC light on the IMC REFL PD (on IOT2L) (and eventually MC2 TRANS on HAM3) -- e.g. LHO:88869 -- there's been some confusion about the state of the HAM2 and HAM3 ISIs alignment, and whether that matters. 

Here, I compare 3 times:

    Date Time                       2025-12-19 00:15 UTC                     2026-01-24 01:28 UTC                      2026-01-26 17:26

    Description                     HAM1 in AIR Pre-JAC reference           Post-JAC install, Pre EOM Install        Post-JAC Install, Pre-EOM Install

    IMC State                       LOCKED                                  OFFLINE                                  OFFLINE
    IMC REFL Power [mW]               0.65                                  0.22                                     0.18
    IMC MC2 TRANS Power [mW]        307.0                                   0.4 (confidently dark noise)             0.4 (dark noise)

    MC1 P/Y                         +852.04 / -2229.74                      +874.30 / -2233.84                       +874.30 / -2233.84
    MC2 P/Y                         +582.28 /  -627.99                      +568.40 /  -628.20                       +568.40 /  -628.20
    MC3 P/Y                           +9.23 / -2433.51                        -1.07 / -2430.71                         -1.07 / -2430.71

    HPI HAM2/HAM3 Physical State    locked/locked                           locked/locked                            locked/locked
    HPI HAM2/HAM3 ISO Gain          disabled/enabled(?)                     disabled/enabled                         disabled/disabled
    ISI HAM2/HAM3 ISO State         ISOLATED/ISOLATED                       DAMPED/DAMPED                            ISOLATED/ISOLATED
    ISI Residuals (w.r.t. "they've been that way forever alignment position")
                                       HAM2       HAM3                         HAM2       HAM3                          HAM2       HAM3 
        X [um]                         0.00 /     0.00                         +8.9 /    -56.4                          0.00 /     0.00
        Y [um]                         0.00 /     0.00                         +4.8 /    +36.6                          0.00 /     0.00
        Z [um]                         0.00 /     0.00                         -6.6 /     -6.7                          0.00 /     0.00
        RX (Roll) [urad]               0.00 /     0.00                         +0.4 /     +1.2                          0.00 /     0.00
        RY (Pitch)[urad]               0.00 /     0.00                         +1.4 /      0.0                          0.00 /     0.00
        RZ (Yaw)  [urad]               0.00 /     0.00                         +5.2 /    -28.6                          0.00 /     0.00

In summary -- having the ISI tables DAMPED (Floating) vs. ISOLATED with HEPI physically locked does make a tens-of-micro level shift in alignment of the tables. 
This is evident by the amount the IMC SUS had to move (from 2025-12-19 to 2026-01-24) in order to start recovering even 0.2 [mW] on the IMC REFL DC PD.
On the scale 0.7 [mW], and without changing the SUS positions (from 2026-01-24 to 2026-01-26) it (HAM2 only, of course) makes the difference between 0.22 [mW] and 0.18 [mW] or (0.22-0.18)/0.22 = 20%-ish percent difference.
When checking / chasing in-air alignment of the beam projected into HAM2 with HAM1 optics, we should make sure that the ISIs are ISOLATED, if possible.

To bring the ISIs to ISOLATED, with HEPI Locked:
   - Set the HPI-HAM{2,3}_ISO_GAIN to 0.0 (disabling the HPI controls), and 
   - Requesting SEI_HAM{2,3} Guardians to ISOLATED (which brings the ISI to HIGH_ISOLATED)