It is expected that you don't find a TEM00 each 0.5 MHz frequency shift. As reported in entry #661, since the AOM is put on the green path, the change in the frequency which it induces is compensated by the servo with a change on the IR which is half of the frequency change in the AOM. This means that a shift of 1 MHz in the driving frequency of the AOM corresponds to a shift of 500 kHz in the frequency of the IR light. 

I think that we can now start to tweak the lens position to improve the matching.  It would be also good to do a cavity scan as the one done in entry #776.

[Aritomi, Eleonora P]

Today we managed to align IR into filter cavity and lock both green and IR again. First we locked IR with 109.03569MHz of AOM frequency as last measurement, but we found that TEM00 is much brighter at 109.03535MHz (Pic 1).

The power of each mode is as follows. Pic 2-7 shows shape of each mode other than HG10. Since green phase lock is not stable, we scanned green phase and measured maximum and minimum of IR transmission. Green phase is scanned at 20Hz. 

Now largest mode is TEM00 and main higher order modes are IG20 (Pic 5) and HG10.

I measured END oplev signal with dataviewer. The signal somehow oscillates at around 1.5Hz...

Beam spot fluctuation seems at several Hz. I measured PR and BS closed loop spectrum. This also looks fine.

Both the open loop and close loop end mirror TF look fine to me. There is no oscillation at 0.3 Hz.  Note that striptool used EPIC channels are sampled at 64 Hz, so the oscillation could be an artifact (down-coversion of higher frequency noise?) You can double-check with dataviewer.

From the movie of the transmitted beam it is not clear the frequncy of the oscillation but it likely to be from a steering mirror (BS or PR). It would be useful to take a spectrum of their motion.

[Aritomi, Eleonora P]

We found that green beam spot at trans is fluctuating (Mov 1) and drifting. So we measured oplev signal of each mirror. From Pic 1 we can see that END mirror is fluctuating at around 0.3Hz. We opened END mirror control loop and closed it again, but it didn't change. We also measured open and closed loop END spectrum (Pic 2,3). Open loop spectrum is similar to previous result. Anyway, since beam spot fluctuation is much faster than 0.3Hz, this is not due to END mirror fluctuation.

Mov 1 green beam spot at trans

Pic 1 time series of oplev signal

Pic 2 open loop END spectrum

Pic 3 closed loop END spectrum

Then we maximized IR reflection and could recover 00 flash once, but problem is that green alignment keeps drifting during IR alignment. Beam spot of green reflection at PR viewport is going away mainly in pitch direction and centering of trans beam by BS is also drifting. We're not sure the reason.

[Aritomi, Eleonora P]

Today we tried to improve the alignment of IR to filter cavity with last two steering mirrors on the bench, but unfortunately we lost TEM00 flash. We recovered the reference on PR chamber, but we couldn't find 00 flash (we didn't maximize IR reflection and check second target). Tomorrow we'll maximize IR reflection and try to recover 00 flash and improve the alignment.

[Aritomi, Eleonora P]

When we changed gain of filter cavity lock, glitch appeared in GRMC transmission and variable gain out of MZ like an attached picture and GRMC and MZ lost lock.

You could use some remote "smart" PDUs as done in virgo.

See as example

https://tds.virgo-gw.eu/?content=3&r=15139

(the name of the PDU is ENERGENIE EG-PMS2-LAN, see attached PDF file)

Based on the measurement we did before, we have dark noise of CC PD. We use this PD to lock the green pump phase, at the same time, we have bandwidth lower than 80Hz. So we could use this dark noise to evaluate the phase noise we have for coherent control loop 1. The calibration method is to use pk-pk value when we scan the green phase, which corresponds to radians of pi/2. In the case of that entry, 25mV corresponds to pi/2. The estimated phase noise(RMS) is attached as follows.

While it will be better to measure the spectrum and do RMS integration. And according to Emil thesis, this 200mrad of phase will degrade ~15dB of squeezing to ~5dB. And it seems to be close to the squeezing situation we saw in the past few days.

[Aritomi, Eleonora P, Yuhang]

Today we replaced a green phase shifter with a new thick phase shifter (picture 1). After re-alignment of GRMC, we got good mode matching (picture 2) and succeeded in locking of GRMC with lower gain. Green power is 146mW before GRMC and 90mW after GRMC when MZ is maximized. So transmission of GRMC is 62% which is good.

Lock of green phase with BAB transmission seems stable with this new phase shifter.

We'll measure transfer function of green phase lock loop with new green phase shifter tomorrow.

Eleonora P, Yuhang, and Aritomi

By keeping green always locked, we measured the power of all the higher order modes we could see from camera. The reason to keep green locked is that we have 'two different FSR' for IR. We found a lot of higher order modes. But the main higher order mode is HG10. This means our alignment needs to be improved for yaw direction. The power for each mode is reported as follows.(picture attached insequence without HG10 and TEM00)

Some modes I didn't write name seem so-called ince-gaussian modes. Maybe it comes from the combination of yaw misalignment and mode-mismatch.

We found the second HG10 at 109.4313MHz, which means FSR is 0.499MHz which is in agreement with our expectation.

We will do alignment of yaw for the next step.

Akutsu-san, Tanioka-san, Simon

It has been a long time since I wrote something in this logbook!

Anyway, we (that means AOS) are now relocating the scatterometer, which is still in one of JASMINE's laboratories at the ATC, to our lab on the first floor of ATC.
The reason is mainly that the JASMINE group needs the space in their lab and removed some desks which we used for the scatterometer's PC.

On the other hand, all of our optics-related stuff is more or less already in that lab where we are going to put the scatterometer in (including the back-scatterometer). So, it seems logical to relocate it there.

The first step in this week was therefore to find a suitable place in the lab, and we decided to use Torii-san's former space for that.
So, we cleaned it up and moved a theoretically usable clean-booth (still without walls) into the free space and a black-painted optical table underneath that booth (see pictures).

By doing all the cleaning-up thing, we discarded a lot of old carton-boxes and plastic garbage.
Now, it looks much more usable.

Next step will be to move the actual scatterometer.

The day before yesterday, we found a water leakage point in the middle station of TAMA arm (north, the arm we are using). See entry #1403. Yesterday I found it was filled fully by the leaked water.

Eleonora P, Yuhang, Matteo, and Aritomi

Several months ago, we have already found the channel for AOM amplification was broken. And the work we did yesterday was without AOM working. Fortunately, Eleonora C remembers where is the old RF amplifier. And today we first implemented the old RF amplifier(ZHL-2). The amplification factor of it is 16dBm. The optimal RF signal for AOM should be 27dBm(while it converts too much to the first order). so usually, we use a lower value. So I give 7dBm from the signal generator and amplify it by 16dBm to have totally 23dBm of driving signal to AOM. The conversion efficiency is about 80% now.

Then, for sure, we need to do the alignment for green again. And then also for IR.

Another very important thing is the amplification of BAB(after OPO transmission). In the beginning, we were thinking to lock it with the same method for CC locking. But we observed noise level brought by the beat between BAB and CC. And this noise is almost 200 times larger than the coherent control error signal we have. So it is not possible to lock with CC beam. By following Matteo's suggestion, we used the leakage power from OPO s-pol transmission through PBS and to p-pol locking PD. Then we feedback this signal after giving an offset. And then we can basically lock the green phase and have a more stable IR beam going inside the filter cavity.

After all the work above, we could lock both green and IR again by changing the AOM driving frequency. We found make higher order modes and they are listed as follows.

The task of tomorrow will be to maximize the TEM00 by moving the last two steering mirrors on the bench for IR. And measure the height of each higher order modes. Also, maximize mode matching. We will follow the entry.

EleonoraP, EleonoraC and Yuhang

We achieved the matching of the BAB (IR probe beam co-aligned with squeezing) into the filter cavity.

In order to align BAB into the FC, we mainly followed the IR alignment procedure we did more than one year ago (entry #646). The main steps of this procedure are to align the cavity for the green, recover the reference on PR chamber, maximize IR reflection from the input mirror, not to look for the beam on the first target and check instead the second target and the end camera.

However, since BAB is only 250uW, it is too weak to see. So we used the green pump to amplify it. The maximum green we can give is obtained setting the offset on the MZ control servo at 4.3 and at that time we have roughly 30mW of IR going to FC. This measurement of power is done while we scanned the green pump phase. So actually we are sending a repeated segment of a sine wave.

By using this 30mW and following the procedure above, we found the flashes of IR. Check attached video.

As you can see in the video, this beam is too bright so, in order to avoid saturation, we decide to reduce the power to around 10mW. In this case, we need to use temperature of 7.202 for OPO, MZ offset of 4.1 and lock p-pol PLL on 150MHz.

After achiving the alignment with the amplified BAB we confirmed that flashes cannot be seen without amplification. (BAB power ~250 uW)

Next steps are to optimize the matching and to use AOM to make green and IR both resonant.

I used one of the few ADC channels still availabe to acquire the filter cavity green transmission. Channel name is K1: FDS-FC_GR_TRA

For the moment it is recorded as an epic channel, since I wanted to display it on the medm screen.

The attached picture show today's lock streches (Even it doesn't seem so, no realigment has been done during this period and the lock was stopped on purpose)

Since the trasmission PD is actually a PSD, I plan to acquire also the X and Y signals in order to  possibly correlate beam motion to suspesion resonances and understand what we should improve.

As entitled. 

It is probably due to the heavy rain of these days. You can check the video. We put a bucket to collect the water.

As reported in entry #1267 we could not setup the new DGS computer wtith gentoo linux HDD, as it was not able to detect the USB.  We bought a PCI express board to be used as USB driver.

Last week I installed it but I still couldn't make the PC to detect USB. I think the board might not be compatible with linux. We should look for another one or consider to change PC.

[Eleonora P, Yuhang]

We installed the IR injection telescope on the bench and we pre-installed the reflection telescope (putting all the optical mounts on the bench without fixing them).

We used the two 750 mm focal length lenses present in the lens box in the lab. Since we saw the beam focusing before the cavity, after the telescope, we thought one of the two focal lengths to be wrong. In fact the second one was actually wrong and it was replaced by another lens of 750 mm focal, located in the lab (Newport KPX121AR.18, PCX LENS, BK 7).

We put HR Laser Line mirrors, BK 7, ALTECHNA Co. Ltd.

In fig 1 the photo of the injection telescope on the bench. We used two of the new small mirror mounts.

We notice that the pipe of the rotary pump close to BS was directly touching the vacuum pipe, transmitting a lot of vibrations.  We have moved the pump of  few centimeters so that the pipe was not touching anymore.

The vibration of BS in the region of 20 Hz seems improved. (See pic 1)