Background

There has been a mystery about the PMC transmission power; it becomes sometimes lowered (but the profile is TEM00), and the reflection power increases (and the profile seems higher order). Although the similar but more awful status (see 16219) was solved by inputting the NPRO thermal offset (see 16236, and maybe 16340) to shift the frequency (maybe) out of the mode-hop range. But still we have observed that the PMC transmit power gets lowered.

Inspection

In the attached figure, I compared the PMC transmission output (LAS-POW_PMC_DC_INMON), the PMC input (LAS-POW_FIB_DC_INMON), and thermal input to the NPRO (IMC-SERVO_NPRO_TEMP_BIAS_OUT16). First, of course, the transmission power seems correlated to the input power; it is ok. With close look to this figure, the PMC trans power sometimes get "stable". This is also a nice time for us because we can have a stable power laser for the IMC (and the main interferometer). In the attached figure, each white-ish region indicate a timing (but not all) of that the PMC trans power is not "stable".

The point here is that the proportional coefficient between the PMC input and the PMC output seems changed. The coefficient is smaller when we have the stable laser than that when the PMC trans is not so stable (well, that's why it seems stable). I may be able to assume that the input power fluctuation during the unstalbe era would be dominated by mode-hopping (partly maybe due to that the higher order mode profiles do not concentrate the power at the center, so the DCPD would get smaller power (?) I am not fully confident, though), while the one during the stable era is dominated by just amplitude fluctuation of the TEM00 input. Anyway, to see such difference, it maybe better to measure the mode hoppings (if any really) at the PD with an appropriate oscilloscope (maybe the DGS would be too slow to catch such hoppings).

Fnally, let's compare the behavior of the PMC input fluctuation and the thermal offset in the attached figure. At a glance, it is hard to find any relation, but  for example check some peak features (red dashed lines in the figure), they are synchronized. Also, it seems there are a certain threshold of the thermal offset, say, -19550 cnts; the PMC trans seems stable when the termal offset is less than this value (maybe not all). I checked this for the other time, and found that the threshold value are not always the same. But it seems each time there is a certain thershold value (Differs with the different lock, different MCL offset values...and so on).

So, if this assumption is right, more stable feedback signal of the thermal offset might be able to also make the PMC output stable in terms of the transverse modes. Another way would be to give more precise offset value to the thermal offset (the current value is 1.2, the unit of which I have nno idea, though) so that such MCL offset would not drift over the threshold.

By the way, these thermal offset fluctuation is well within the PLL loop of the green lasers??? (What is the tolerable frequency range??)