P.S. There is a document about the 80MHz VCO box. This may be helpful. 

link to LIGO DCC

2. Weighted screw rod at the bottom for tilting the mirror-holder:

The screw length selected here (2") is not interfering with any part of the assembly.

The 'weights' I have here are just thumb nuts from Mcmaster, so their weight is fixed (1.65g each, btw).

Problem I'd like to solve: Find an assortment of weighted, symmetric nuts with caps on one end to fix position on shaft. 

3. Set-screws on both side of wire clamp to adjust its horizontal position:

Thanks for pointing out the mismatch in travel distance of protrusion and clamp screws. To match them, the clamp screw slot now sticks out of the profile (by 1.5mm). The range of the clamp motion is +/- 3 mm.

Also, here's a screenshot of the slot in the mirror holder:

--

- Excluding the weighted screw rod assembly, the height gap between assembly COM and wire release point is 3.1 mm.

3.
- Do we need this much of extended range of the clamp location? How much range will we need if we use either 3/8 or 1/4 inch mirrors?
- This slot on the mirror holder ring is not machinable.

About the CoM height
- Include the angle adjustment screw and adjust the wire releasing point to have comparable pitch resonant freq to the SOS suspension.

[Keiko / Suresh / Anamaria / Kiwamu]

 The AM components do exist also on the beam after the EOM.

The peaks were found at 11, 29 and 55 MHz, where the PM are supposed to be imposed.

Suresh and Keiko minimized them by rotating the HWP, which is in front of the EOM.

Also Anamaria and I tried minimizing them by adjusting the EOM crystal alignment.

However everytime after we minimized the AM peaks, they grew back in a time scale of ~ 1 min.

Potentially it could be a problem of the HWP and/or EOM alignment.

Since we wanted to proceed the in-vac work anyways, we stopped investigating it and decided to postpone it for tomorrow.

We again adjusted the incident power to 20 mW.

-- P.S.

 The incident power going to MC went down to 7 mW for some reasons. This was found after ~ 6 hours from our works on the PSL table.

We haven't touched anything on the PSL table since the daytime work.

Possibly the angle of the HWP is drifting (why?) and changed the amount of the P-polarizing beam power.

Suresh locked the angles of two HWPs, which are the one just after the EOM and the one after the attenuation PBS.

Indeed it was suspenseful.

We tried finding where the clipping happened, but we couldn't find any obvious clippings.

So we checked centering of the beams on all the optics associated with the AS path, starting from BS, SR3,... to the AS optical bench.

And during the work some of them were recentered.

At the end we found no clipping. To make sure we tested the available range (no clipping range) by exciting the angular motion of BS with AWG (f ~ 1Hz, a ~ 1000).

The beam looked successfully coming out at the most of the angular oscillation point.

I think the precision due to the loop gain uncertainty is something like 0.1% at 0.1 Hz. It's not the issue.

The real issue was the loud motion of MICH, which degrades the coherence of the measurement.

Also last night I tried the fringe hopping technique and gave it up for several reasons.

(uncertainty due to the loop gain)

(Effect from the MICH motion)

[Steve / Kiwamu]

Motivation:

 Since the oplevs were the ones we haven't carefully tested, so the oplevs need to be checked.

This checking is also a part of the suspension optimizations (see the minutes of the last 40m meeting).

 In this work Steve will check two things for all the oplevs :

    1. Noise level including the dark noise, electrical noise and ADC noise to just make sure that the noise are blow the signal levels below ~ 30Hz.

    2. The spectra of the signals to make sure there are no funny oscillations and unexpected structures

Measurement :

  To check the things listed above, we take two kinds of oplves' spectra :

     1. "dark noise" when the He-Ne beam is blocked.

     2. "signals" when the optics are damped by only OSEMs

 We did these checks on the BS oplev today (see the last entry).

All of them are fine, for example the dark noise (including electrical noise and ADC noise) are below the signal levels.

And no oscillation peak was found. Steve will go through all of the oplevs in this way.

The Xarm is currently in its original state, all cables are connected and c1auxex is hosting the slow channels.

Here are the free-swinging spectra for the BS, ETMX, ETMY, ITMX, ITMY, MC1, MC2, MC3, and PRM chambers.  Kiwamu left the suspensions free for 5 hours this weekend, starting at Sat Apr 30 00:15:26 2011.

This is GPS time 988 182 941.  Quick tip: you can do local to GPS time conversions using lalapps_tconvert, which is a lot like tconvert but with special powers.  It is installed on pianosa.

$ lalapps_tconvert Sat Apr 30 00:15:26 2011

988182941

I generated these figures with the attached Python script, measure.py.

Notice that the C1:SUS-ITMX_SENSOR_UL and C1:SUS-MC3_SENSOR_UL spectra fall as 1/f.  Jenne suggested that this might indicate that there is a loose electrical connection.

Also, notice that C1:SUS-ETMY_SENSOR_LR, C1:SUS-ITMY_SENSOR_LL, and C1:SUS-PRM_SENSOR_SIDE are a lot noisier above 10 Hz.

Jenne went through all the suspension racks and pushed all the connectors.

After pushing them, we had a quick look at those spectra and found no funny noise spectrum except for C1:PRM-SENSOR_UL.

We then checked connection around the SCSI cables and eventually found the connection between ADC_card_0 and a SCSI was loose.

We put short standoffs on the ADC card so that the screws from the SCSI can nicely reach to the ADC card. Now everything looks fine.

SUS diagnostic is quite useful !

I believe that the 17 Hz broad structure on SIDE is just because of a bad rotational angle of the SIDE OSEM.

The same structure had been observed on the EMTY_UR, and the structure became narrower after we repositioned/rotated the OSEM yesterday.

My guess is that the SIDE OSEM is now in a place where the OSEM is quite sensitive to the bounce mode

and creating the broad structure due to a bi-linear coupling between the bounce mode and low frequency signals.

I guess the ETMY suspension is still fine. Their OSEM DC voltage and the free swinging spectra look healthy.

It could be a failure in the initial guess for fitting.

 Turns out I was missing a critical step in the process...running makeSUSspectra.m  After I do that, everything is back under control, and ETMY looks fine. 

I'm almost done doing the peak-fitting and matrix inversion for all optics.

I talked with Zach. So this is just a note for the others.

The setup I suggested was totally equivalent with the setup proposed in the entry http://131.215.115.52:8080/40m/1721, except that the PLL PD sees not only 29.501MHz, but also 1kHz and 59.001MHz. These additional beating are excluded by the PD and the PLL servo. In any case the beating at 1kHz is present at the camera. So if you play with the beamsplitter alignment you will see not only the perfect Gaussian picture, but also distorted picture which is resulted by mismatching of the two wave fronts. That's the fun part!

The point is that you can get an equivalent type of the test with fewer optics and fewer efforts. Particularly, I guess the setup would not be the final goal. So, these features would be nice for you.

I resumed the pumping down. It started from 9:55 am.

I stopped the pumping at 14:50 pm because I was going back home. I did the same procedure as Koji wrote down (see here).

The P1 pressure reached 32 Torr.

Koji will take over the pumping shift tonight. 

This was certainly my fault, probably left over from early debugging of my BO switch check script.  I've turned the ITMX whitening all off, to match the other suspensions.

Again, this was my fault.  Sorry.  I just accidentally left this off when I finished yesterday.  Much apologies.  I've turned the ETMX watchdog back on.

It stacked again . We should take a closer look at it.

I went push all the possible connectors for the MC3 shadow sensors including the SCSIs, flat cables and satellite box.

Also I put screws on them so that they won't become loose any more.

As a result UL_PDMON dropped from 0.6 V to 0.490 V and it becomes stable so far.

I didn't strain relief the cables but we must do it at some point before going into the full locking test.

Solved. The power code of c1iscaux was loose.
Has anyone worked around the back side of 1Y3?

I looked into the problem. I went around the channel lists for each slow machines and found the variables are supported by c1iscaux

controls@pianosa:/cvs/cds/caltech/target/c1iscaux 0$ cd /cvs/cds/caltech/target/c1iscaux
controls@pianosa:/cvs/cds/caltech/target/c1iscaux 0$ grep C1:IF *
C1IFO_STATE.db:grecord(ai,"C1:IFO-STATE")

It seemed that the machine was not responding to ping. I went to 1Y3 and found the crate was off. Actually this is not correct.
The key was on but the power was off. I looked at the back and found the power code was loose from its inlet.
Once the code was pushed in and the crate was keyed, the white boxes got back online.

Just in case I burtrestored these slow channels by the snapshot at 6:07am on Sunday.

I was working around 1Y2 and 1Y3 when I wired the DAC in the c1lsc IO chassis in 1Y3 to the tip-tilt electronics in 1Y2.  I had to mess around in the back of 1Y3 to get it connected.  I obviously did not intend to touch anything else, but it's certainly possible that I did.

I modified a set of the automated MC locking scripts which are dedicated for the low power MC.

Currently there are three scripts like the usual MC locking scripts:

(1)mcup_low_power, (2) mcdown_low_power and (3) autolockMCmain40_low_power.

I ran those scripts on op340m as usual and so far they are running very well. The lock acquisition is quite repeatable.

I hope theses scripts always bring the lock condition to the same one and hence the LOCKIN signals don't change by every lock.

- To run the script

  log into op340m and run autolockMCmain40m_low_power

The leveling was still okay. The MC mirrors were realigned and now they all are fine.

We will go ahead for the vertex alignment and extraction of the pick-off beams.

Here is a summary of the spot measurement.

They are the DC responses.

I put the resonant frequencies that Leo reported in the wiki to obtain the DC response.

The resonant frequencies I used are :

  f_BS = 0.957 Hz

  f_ITMX = 0.966 Hz

  f_ITMY = 0.988 Hz

Also I assumed that all the Q-values are 5 due to the damping.

The procedure you wrote down as a standard is right.   I explain reasons why we didn't do such way. 

For our situation, we can rotate the polarization angle of the incident beam by using a HWP in front of the Faraday.  

This means we don't have to pay attention about the PBS_in because the rotation of either PBS_in or the HWP causes the same effect (i.e. variable transmission ). This is why we didn't carefully check the PBS_in, but did carefully with the HWP.

Normally we should take a maximum transmission according to a instruction paper from OFR, but we figured out it was difficult to find a maximum point. In fact looking at the change of the power with such big incident (~1W) was too hard to track, it only can change 4th significant digit ( corresponds to 1mW accuracy for high power incident ) in the monitor of the Ophir power meter. So we decided to go to a minimum point instead a maximum point, and around a minmum point we could resolve the power with accuracy of less than 1mW.

After obtaining the minimum by rotating the HWP, we adjusted the angle of PBS_out to have a minimum transmission.

And then we was going to flip the Faraday 180 deg for fine tuning, but we didn't. We found that once we remove the Faraday from the mount, the role angle of the Faraday is going to be screwed up because the mount can not control the role angle of the Faraday. This is why we didn't flip it.

??? I still don't understand. What principle are you rely on?

I could not understand why you rotated the HWP to the "minimum" transmission
and then minimized the transmission by rotating the output PBS. What is optimized by this action?

Probably there is some hidden assumption  which I still don't understand.
Something like: Better transmission gives best isolation, PBS has some leakage transmission
of the S-pol light, and so on.

Tell me what is the principle otherwise I don't accept that this adjustment is "to get a good isolation with the Faraday".

P.S. you could flip the faraday without removing it from the V-shaped mount. This does not roll the Faraday.

 Actually in the last 40m meeting we discussed the SUS diagnosis and decided not to post the results on the elog.

Alternatively we will have a summary web page which will contain all the information (sensitivity, UGF monitors, etc.,) and will be updated everyday like GEO.

This will be a place where we should post the SUS results every week.

So we don't need to worry about the cron-elog job, and for running the scripts you can simply use one of the lab machines as a cron host.

Once you get the scripts running on a machine as a cron job, it will be the point where we should quit developing the SUS diagonalizer and pass it to the web summary people.

Well, the diagram I drew is true. I also have been confused by this 1/f^2 issue in our PLL.

As Rana pointed out, the open-loop TF should become 1/f^2 over most of the frequency range, but it still remains 1/f above 5kHz for some reasons. 

 Need more investigations.

 At the beginning I tried phase-locking the VCO to the beat note without any external filters (i.e. SR560 see here), but I never succeeded.

It was because the hold-in range of the PLL was not sufficiently wide, it could stay locked within frequency range of less than +/- 1MHz from the center frequency of 80 MHz.

This is obviously not good, because the beat note typically fluctuates by more than +/- 3MHz in time scale of 1 sec or so.

  So I decided to put an external filter, SR560,  in order to have a larger DC gain and a higher UGF.

Somehow I unconsciously tuned the SR560 to have a pole at 1Hz with the gain of 2000, which shouldn't work in principle because the open-loop will be 1/f^2.

However I found that the PLL became more robust, in fact it can track the input frequency range of +/- 5MHz.

The open-loop TF is shown above. For comparison I plotted also the open-loop TF wehn it's without the SR560.

I checked the frequency of the VCO output when it was phase-locked to a Marconi, it was healthy (i.e. the same frequency as the input signal from Marconi).

 I haven't yet taken any data for the IR fluctuation when the Xarm is locked by the green locking.

You are right, the DC drift was due to a lack of the DC gain. But don't worry about that, because this issue has been solved.

(DC gain issue)

  The lack of DC gain was because I put an IIR filter called ''DC block" that I made. It has 1/f shape below 30mHz and becomes flat above it.

The purpose of this filter was to avoid a DC kick when it starts feeding back to ETMX.

Usually the output signal from the PLL has an offset, typically ~5V, then this offset is also acquired into the ADC and eventually kicks ETMX through the feedback.

So when I took the time series data I enabled the 'DC block', that's why it drifts slowly.

 After taking the time series, I found that without this 'DC block' technique, the lock can be achieved by appropriately subtracting the offsets with epics numerical values.

This subtraction technique, of course, gave me more stable lock at DC.

(open loop transfer function)

Here is the open-loop TF of the arm locking I measured last night:

The IIR filter chain has the following poles and zeros:

     pole 0.1Hz, 1000Hz,

   zero 1Hz, 30Hz

For the fitting I assume that the ETMX pendulum has a resonance at 1Hz with Q of 5. Also I put the cavity pole at 24 kHz, assuming the finesse is 80 at 532 nm.

I just fitted the gain and the time delay by my eyes.

If I believe the result of the fitting, whole time delay is 330 usec, which sounds pretty large to me.

We use Alberto's laser for the Y end Green Locking.

The reason for using Alberto's laser is that some amount of work has already gone into characterising its phase noise.  Ref elog entry 2788

There is a page on the 40m wiki explaining the procedure.

  http://blue.ligo-wa.caltech.edu:8000/40m/OSEM_adjustment_proceedure

Additionally there are several old elogs about the cross-coupling minimization, which can be useful for us:

[1] "SUS ranking from measured data", iLog by Osamy (Aug.22 2005)

[2] "TF from position to sensors for ETMX ", iLog by Osamu (Aug.25 2005)

[3] "Further OSEM tweaking", iLog by Rana (Oct.3 2006)

Yes.  

The medm screen C1SUS_GDS_TP.adl showed some numbers which correspond to that I wrote  on the outputs.

But I still could not get any physical voltage coming from the DACs.

Okay, now the data are attached. At that time I just wanted to say like the follower.

- - -

In the free-swinging spectra around ~0.5Hz, you can see the two resonances, which come from pitch and yaw mode of the pendulum.

Note that, the vertical and the horizontal axis are adjusted to be the same for the two plots in the figure .

And I found that

* the floor levels are almost the same (the factor of about 1.5 or something like that) compared to the past.

* however the peak heights for two resonances are several 10 times smaller than the past.

* this tendency are shown in all of the data (ITMX, ETMX, ETMY).

If such variation of the peak heights is cased by the seismic activity, it means the seismic level change by several 10 times. It sounds large to me.
 

Well, I get the point now. It could be either seismic or change in the suspension Q.

The pendulum memorizes its own state for a period of ~ Q T_pend. (T_pend is the period of the pendulum)
If the pendulum Q is very high (>10⁴), once the pendulum is excited, the effect of the excitation can last many hours.

On the other hand, in our current case, we turned on the damping once, and then turned off the damping.
Again it takes ~Q T_pend to be excited. 

In those cases, the peak height is not yet before in equilibrium, and can be higher or lower than expected. 

So, my suggestion is:
Track the peak height along the long time scale (~10hrs) and compare between the previous one and the current one.
This may indicate whether it is equilibrium or not, and where the equilibrium is.

I updated the last entry about the in-vac work (see here)

On the wiki I summarized about the modification of the PDH box which is currently running on the end PDH locking. 

http://lhocds.ligo-wa.caltech.edu:8000/40m/Electronics/PDH_Universal_Box

The box was newly labeled "G1" standing for "Green locking #1".

 The framebuilder just needed to be restarted to pull in the fixed channel names.  I restarted the framebuilder and now the channels (C1:SUS-MC2_MCL_*) are showing up properly.

 We have been modifying models that need to have their channels renamed to run activateDQ when Joe's post_build_script to is run. 

The trick is to integrate things to get the post_build_script running after every model build (getting it hooked in to the make file somehow).  We're working on it.

 I've added the following epics channels to sus_single_control model using the epicsOutput part:

• OL_SUM
• OL_PITCH
• OL_YAW

These channels are now all available.  I'm not exactly sure how to ensure that they're being trended.  I'll check that tomorrow.

(Koji, Yuta, Kiwamu)

Now the pressure at P1 is 740 torr, which is close to the atmospheric pressure of 760 torr.

We changed the air cylinder twice in this evening, and the last cylinder ran out at about 23:00 pm.

We left it as it is. Steve is going to make a final touch for it tomorrow morning.

1. You are right, the gain for the single resonant circuit was about 9.3 in my measurement.

But the reason why the triple is better than the single resonant circuit comes from the transformer.

The impedance can be degraded by a loss of the transformer, because it got worse after applying the transformer in the past measurement.

Also I definitely confirmed that the circuit had the impedance of 7.2 kOhm at the resonance of 52.9MHz without the transformer.

So it shall give the gain of 12, but did not after applying the transformer.

2.  Yes, I think we need some variable components just in case.

5.  For the impedance matching, I will select a transformer so that 55MHz is matched. In contrast I will leave two lower resonances as they are.

This is because 11MHz and 29.5MHz usually tend to have higher impedance than 55MHz. In this case, even if the impedance is mismatched, the gain for these can be kept higher than 11.

I will post the detail for this mismatched case tomorrow.

An important rule during the in-vac work :

   Do not change the incident axis of the X and Y green beams.

Because of the air flows and unusual pressure in the chambers, the DC alignment of the suspensions become less reliable in terms of the beam pointing reference.

The green beams will be considered as references during the vent.

Okay, I guess I understand what you want to know. I did the following steps.

1. calculated the conversion efficiency as a function of the waist size. I found w~50um was the optimum waist.

  Note: the confocal relation Zr = pi * wo^2/(lamba/n) = L/2  gives us almost the same optimum waist. 

elog #2735

  2.  Did mode matching to get w=50um

elog #2959

elog #3188

  3. calculated the offset

elog #2850

 4. Moved the ovens

THANK YOU, JOE !!! 

You are right. We should change or rotate the mirror mount.

Actually when Suresh and I were putting the mirror we rotated the mount  by 90 deg such that the fat side of the mount is at left had side.

It was because the fat side had been clipping the oplev beam when the fat side is at right.

At that moment we were blocking the green beam to only see the faint IR beam with a sensor card, so we haven't checked the green beam.

Anyway the mount is apparently not good for the green beam.

Did you take the 180 deg shift into your account ?

Since your measurement was done when the loop was closed, there must be an additional 180 deg phase shift (in other words, minus sign).

 I thought I had, but apparently not, and I'd made another error or two in the complex version of my fitting routine. I've fixed them now, thanks! I'll put up the new fitting results tomorrow morning.

Good work for the oplev noise simulations. Here are some comments/questions:

 (A) The noise looks suppressed but the open-loop transfer function doesn't look so good, because it doesn't have sufficient phase margins at the UGFs (0.01 and 10 Hz).

      I guess it is better to have a phase margin detector in your code so that the code automatically rejects a bad phase margin case.

      Actually since the number of data points are finite, the rms detector in the simulation can not always find a sharp loop oscillation.

 (B) The resultant poles and zeros seem canceling each other but the filter still has a structure. Is something wrong ?