Virgo Logbook

AdV-DET (Commissioning)

bersanetti, gouaty, mwas - 18:59 Monday 05 August 2019 (46597)

Impact of 56 MHz side band on the HOM content of B1s2 and B1 beams

The tests performed today were aimed to better understand the contribution of the 56 MHz side band to the HOM content of the B1s2 and B1 beams.

While the ITF was fully locked, we performed the following scans of the 56 MHz modulation depth:

First we reduced the 56 MHz modulation depth by steps of 1 dB from its nominal value (12 dB) down to 7 dB (we tried to lowered down to 6 dB but the interferometer unlocked). This scan is shown in Fig.1. During this scan the gains and integration times of the B1s2 and B1 cameras were increased in order to better visualize the modes of very high orders. Fig.2 shows the images provided by these cameras when the modulation depth of the 56 MHz is set to 12 dB, and Fig.3 shows the same images for a modulation depth set to 7 dB. Fig.4 and Fig.5 shows the same thing with a log color scale. The modes of very high order do not seem to be impacted by the reduction of the 56 MHz. Instead the central part of the beam seems to be affected, probably because the power in the TEM00 decreases with less 56 MHz side band, but maybe also because lower order modes are affected.
The second scan of the 56 MHz side band was performed from 12dB to 8 dB (and from 8 dB to 12 dB), with lower gains set on the cameras in order to have images less focused on the very high order modes. The scan is shown in Fig.6. Fig.7 and Fig.8 show the images of the beams for a modulation depth set to 12dB and 8 dB respectively, in linear scale. Fig. 9 and Fig.10 shows the same images in log scale. The shape of the beam (mostly B1s2) seems to be affected, but the conclusion is not clear.

More analysis in progress.

Images attached to this report

46597_1565024298_Screenshot from 2019-08-05 14-05-16.png

46597_1565024304_Screenshot from 2019-08-05 14-24-47.png

46597_1565024310_Screenshot from 2019-08-05 14-05-39.png

46597_1565024315_Screenshot from 2019-08-05 14-24-59.png

46597_1565024325_Screenshot from 2019-08-05 14-47-04.png

46597_1565024330_Screenshot from 2019-08-05 14-51-52.png

46597_1565024334_Screenshot from 2019-08-05 14-47-23.png

46597_1565024341_Screenshot from 2019-08-05 14-52-02.png

Comments to this report:

mwas - 10:09 Saturday 10 August 2019 (46637)

Figure 1. Looking at the powers on the PDs allows us to make an estimate of the 56MHz filtering by the OMCs. If we compare the time with 12dB of 56MHz modulation and 9dB, so a factor 2 in change of the 56MHz power. Then the power on B1s1 drops by 4.5mW and on B1s2 it drops by 0.025mW. The two PDs see about the same fraction of the total power of their respective beams (~8%), so the ratio of the two powers give us the filtering by OMC1, the 56MHz is reduced by a factor 180 by OMC1.

Figure 2. Shows the RAM injections on the 56MHz done two weeks ago, looking at 1kHz where the DARM loop should not be shaping the noise seen on B1. On B1s1 the noise is 21e-6 mW/rtHz, on B1s2 it is 0.16e-6 mW/rtHz and on B1_PD2 it is 0.1e-6 mW/rtHz. So a filtering by a factor 130 by OMC1, and by a factor 10 by OMC2 (taking into account that B1s2 sees 8% of the power and B1_PD2 sees 50% of the power). This is for the noise injection with amplitude 12.

Figure 3. The same for the injection with amplitude 3. On B1s1 the noise is at 6.5e-6 mW/rtHz, on B1s2 it is 55e-9 mW/rtHz and on B1 it is 37e-9 mW/rtHz. Both B1s2 and B1 have a shot noise floor at 25e-9. So it corresponds to 50e-9 mW/rtHz on B1s2 and 27e-9 on B1. Which yields similar filtering estimatates, 130 for OMC1 and 11 for OMC2.

Last time this was measured with a line on 56MHz RAM at the end of January at that time the OMC1 filtering was a factor 230 and the OMC2 filtering was a factor 25. So in total the filtering now seems to be a factor 4 worse than in January.

Images attached to this comment

46637_1565423316_Screenshot from 2019-08-10 07-21-13.png

46637_1565423607_Screenshot from 2019-08-10 09-43-16.png

46637_1565424173_Screenshot from 2019-08-10 09-58-02.png

mwas - 7:20 Wednesday 28 August 2019 (46785)

This measurement can also be used to estimate if there is other noises in the B1s1 spectrum than the 56MHz RIN.

Figure 1. Shows the ratio of the B1s1 spectrum for the nominal 12dB 56MHz modulation and for 8dB of 56MHz modulation. The reduction in modulation is by 4dB, so the 56MHz power should change by a factor 10^(4/10) = 2.51. Which is compatible with the reduction observed between 20Hz and 1kHz.

Figure 2. Shows the spectrum of B1s1 at 12dB and 8dB of modulation, and the estimation of the noise that is not scalling as the 56MHz modulation depth sqrt( ((ASD_8dB*2.51)^2 - ASD_12dB^2)/(2.51^2-1)). This is about a factor 7 below the nominal B1s1 spectrum (with large error bars).

When the 56MHz is subtracted from h(t), the B1s1 data is used, and the 56MHz spectrum is a factor ~2.5 below h(t) before subtraction. During the subtraction the 56Mhz is removed, but other noises in B1s1 unrelated to the 56Mhz would be added. Figure 2 shows us that these other noises are a factor ~7 below the 56Mhz, so what is added is a factor 2.5*7=17 below h(t), sufficiently small to be negligible.

Images attached to this comment