By the way, I tried to compare this result with the past one measured in the beginning of January, Aso-san made a power budget 12425 on the AS table (the power distribution on each PD/QPD are not half one by one; why?) 26+13=39mW to AS RF with the condition. Assuming the input power to PRM was 1.5W (I didn't remember, but confirmed with K1:LAS-POW_IMC_DC_OUT16),

1.5W*0.1*1*0.3*0.9 = 40.5mW... hmm this would be consistent with the 39mW.

This is (PRM trans) * (MICH bright lock to AS) * (SRM trans) * (OFI trans) : as far as I remembered, the OSTM was detached from the optical path at that time.

So at least at that time, such a 50% loss seemed not existing, but note that maybe this measurement included all HOM.

Expecting 8.1 mW but measured 3.3 mW means unexpected optical loss of ~60%.
This is actually consistent with the similar measurement done for AS RF back in November (see klog #11762).

More details about the clipping measurement.

PIT

RIN=3e-3 [1/sqrt(Hz)]@5Hz, with 3000 cnt excitation to OMMT2.

QPD1 saw 0.8 cnt p-p beam spot motion, corresponding to 64um.

This corresponds to 47um beam spot motion at the entrance of OFI.

Here, we assumed the distance between OMMT2 and OFI/QPD1 to be 2.5m/3.4m.

Coupling constant from OFI vibration to RIN is 3e-3/47um = 64 [1/m/sqrt(Hz)]

YAW

RIN=4e-3 [1/sqrt(Hz)]@5Hz, with 300 cnt excitation to OMMT2.

QPD1 signal was 0.25p-p => 20um.

Beam spot motion @ OFI: 15um

Coupling constant from OFI vibration to RIN is 4e-3/15um = 267 [1/m/sqrt(Hz)]

Coupling of the clipping to DARM

Using the seismic noise model below and vibration to RIN conversion factor of 300, we can estimate the contribution of the clipping noise to DARM if the OFI moves with seismic.

Seismic noise in KAGRA: 1e-9/f^2 m/sqrt(Hz)

clip_noise = 300*1e-9/f^2*8/C,

where C is optical gain and 8[mW] is the DC output of the DC readout PD.

The projected noise is way below the current noise level.

Hi Yoichi,

Assuming that the coupling coefficient was derived from some measurements in the ITMX single bounce configuration, I might suggest you taking a look at the same thing in full lock for completeness. We know that the spatial beam distribution is extended with the interferometer fully locked and this may change the noise coupling scenario. Although, I am just saying it without a quantitative argument.

non

I estimated the ground motion at OMC leg from our accelerometer data.

• left : amplitude spectrum of acceleration
• right : calculated displacement (vertical)

From this result, the clipping noise can be larger than ASO-san's model about order 2.

[Details]

• place : see http://klog.icrr.u-tokyo.ac.jp/osl/?r=13136
• channel :  K1:PEM-SENSOR_RACK_OMC1_DSUB2_OUT_DQ
• time  : 1267056018 - 1267057042)
• calibration factor : 0.1 (amp), 610uV/counts, 300mV/(m/s2)
• ADC noise : 0.006 count/rtHz (plot)
• Sensor noise : 2.5e-6/f**0.27 m/s2/rtHz (by eye)  <- from a comparison with a seismometer (plot), the slope at <20 Hz can be understood as a sensor noise.

[Ushiba, Kokeyama, Aso]

OMC REFL DCPD (on the air AS table, thorlabs PDA) was installed and and connected.

The power of the OMC REFL beam at the periscope is 23mW for 3500 cts at the OMC REFL DCPD. It is higher than the estimation.

While, the estimation is following with the IMC output of 2.6W, and with the beam reflecting from X arm:
Estimated OMC REFL power = 2.6 * 0.1 (PRM) * 0.9? (estimated arm reflectivity) * 0.25 (double BS) * 0.3 (SRM) * 0.9 (OFI) * 1 (OSTM) * 1 (OMC) = 18 mW

The OMC REFL beam is stronger than the estimation...?

OMC model was modified
Since the OMC REFL DCPD channel is not in the model, one filter bank for it was added. The omc model was done "make" and "make install." It need to be restarted at the next opportunity. Currently it is K1:IOP-OMC0_MADC0_TP_CH12.

OMC REFL camera is online
You can see the rejected HOMs!

AS air power
The power of the AS air beam is 220uW at the periscope where as LSC_AS_PDA1_DC_IN1 is 11.8 cts.

We also measured the powers at AS PDA1 and PDA2. They are 60uW and 40uW respectively. I am not sure what the current calibration factors are applied on LSC_AS_PDA{1,2}_DC_OUT, as they don't seem to show the powers in front of the PDs.

This estimation is maybe the coupling with clipped power and seisimic noise. However, we should also consider the coupling the scattered light generated with clipping and the main beam. However, the latter is normally quite difficult to predict theoretically. Wahimi-kun's analyisys is one of answeres including scattered light effect modulated by seismic noise.