I turned the monitor switch to fb0 on the rack, and saw nothing come up on screen. I presume this means it needs a hard reset?

DO WE HAVE ANY IDEA  WHY FB0 SEEMS TO NEED A HARD RESET AT A RATE OF ABOUT 1 / MONTH? THIS SEEMS TERRIBLE. MAYBE WE ARE BUMPING SOME RACK CONNECTION FREQUENTLY? Is this why we are burning through hard drives like toast?

Today we moved some of the optics away from the blue box to make room for the mode matching.

We also found a 1" diameter optic in one of the cabinets with a transmission of 0.81 % (S-pol). This makes a nice match with the 0.57% transmission of the 2" diameter output coupler and so its now installed. We'll later replace it with a good 2" diameter mirror.

We also moved all of the optics after the laser to align with the screw holes and are trying to make the beams go along the screw hole lines.

Alastair has remeasured the laser beam profile after the EOM and the PBSs using the WinCamD beam scanner. Results are being produced.

 The Particle Counter is a GT-526 from Met One. Datasheet on the web here.

The manual is here.

I was unable to get the data off of it using my Mac and Terminal. I tried to get the data off of it using Frank's wonderfully helpful elog entry, but was still unable to. When retrieving data using the 'Comet' software on the little laptop, the window just says '0% complete' and continues counting. After 12 minutes, I gave up.

The setup of the counter seems to be that it adds up for 10 seconds and then waits for 590 seconds. The display on the front panel is in units of 'CF".

The current reading (while Alastair and I are working on the table) is:

0.3 micron     30000 CF

0.5 micron      3000 CF

1.0 micron      1700 CF

I'm assuming that we multiply by 6 to get to the commonly used CFM (# of particles per cubic foot per minute).

Around 5:30PM today, Alastair commented on how the 2 HVAC systems seemed to be fighting. Recently, the physical plant guys installed a second AC sensor (stay tuned for the elog and photos from Aidan). Apparently, these systems don't feed the same controller. So I decreased the temperature on the new South wall sensor by 1.5 F down to 68 F. The one in the NW corner of the lab is unchanged.

Ah.. right. I remember that this possibility dawned on me when I went upstairs to calculate finesses that day. I meant to tell you that I didn't check that the polarization was still S---sorry.

I tried some more mirrors:

Mirror 7 : CVI PR1-1064-99.5-1037 for 0degrees, 1".  Transmission 1.28% for 45S.  Taken from the PSL lab.  Mirror is dirty.

Mirror 8 : CVI PR1-1064-98-1037 for 0degrees, 1".  Transmission 69% for 45S.  Taken from PSL lab.  Mirror is dirty.

Mirror 9 : VLOC FM-25.4-6.35-1FS-1010-1064-2, 1".  Transmission 101ppm.  Taken from 40m.

One possible configuration would be to use Mirror 7 as the input and Mirror 1 as the output.  That gives us a finesse of around 330 and a contrast defect of 0.13.

I re-measured the mirror transmissions for 45S for all the candidate mirrors for the cavity.  The Y1 45P mirrors that were borrowed from the 40m have very low transmission for 45S.  I think that when they were measured before, the polarization had been rotated at the wave plate just after the faraday, and this is what gave the higher transmission values for the 45P mirrors.  I rotated it back to S using one of the good polarizers and power meter.  I have labelled each mirror with a number and the measured transmission.

Mirror 1:  BS1 - 99 - 45S, 2".  Old input mirror from the gyro.  Power in 115.4mW.  Power transmitted 0.657mW.  Transmission 0.57%

Mirror 2: Y1 - 45S, 2".  Old output mirror from the gyro.  Power in 115.3mW.  Power transmitted 19uW.  Transmission 165ppm.

Mirror 3: Y1 - 45P, 2".  From the 40m.  Power in 115.1mW.  Power transmitted 13uW.  Transmission 113ppm.

MIrror 4 Y1 - 45P, 2".  From the 40m.  Power in 115.3mW.  Power transmitted 20.2uW.  Transmission 175ppm.

Mirror 5: Y1 - 45S - 6.00m, 1".  Old curved mirror from gyro.  Power in 114.9mW.  Power transmitted 15.3uW.  Transmission 133ppm.

Mirror 6: Y1 - 45S - 6.00m, 1".  Old curved mirror from gyro.  Power in 115mW.  Power transmitted 20.5uW.  Transmission 178ppm.

plz, can someone check what's going on with fb0. I can't login but ping is working. workstations and fb1 are ok

Taking some transfer functions of the PDH boxes for the simulink model.  We keep talking about which one has been modified and which hasn't, so at this point I've labelled them box1 and box2 (label on the front panel) so we'll be able to identify them later.  My understanding is that box 1 has been modified by Rana.

First TF below is box 2.  Gain is 4.8 and boost is off.  This is using the agilent without attenuators.

 Second TF is box 1.  Same settings as above.

The 35W laser and the Gryo NPRO have been shutdown in preparation for tomorrow's conduit installation work. The key for the 35W laser in the lower draw in the first desk attached to a Thorlabs USB flash-driver. The key for the Gyro laser wasn't in the driver. The laser was in use this afternoon and since then someone has completely powered it down and removed the key, I'm assuming in prep for tomorrow morning.

The following is a more-or-less exhaustive photo summary of the sealing work that was recently done in the ATF. The insets indicate the locations of the ducts within the lab.

This is the Box that I built to replace the funky solder free breadboard. It does the intensity stabilization, locking the Mach Zehnder, and has the transimpedance amplifiers for the green PDs.

** Picture to small !!*

Yes it does happen if I touch the BNC shells to the box ground.

• I found that it was a ground loop from using the same power supply on both the ISS PD box and my box, and using the ground from that power supply to ground both boxes.
  • I did not, however see a forest of the line harmonics. I saw a huge white noise, and I am near certain that it was noise entering into my ground and touching the "+" input of my transimpedance resistors.
  • When I put a 50 Ohm terminator on the transimpedance amplifier's "-" input (100k Trans Res), I amplified the heck out of the noise at the +, which allowed me to see this crazy behavior. Fortuitous?
• I put them on separate power supplies and the dark noise level now is not displaying "THE EFFECT"
• I measured a phase noise spectrum open and closed loop, and saw the same behavior as before...
  • I see no suppression of green phase noise when I close the loop on IR
  • I talked with Vladimir for a while about this, and think I might have a handle on how locking the IR might not suppress the green if thermoelastic and thermorefractive noise in the crystal drives the phase noise at all.
  • I will try making the temperature more unstable to see if I see any effect (go closer to open loop)
  • I may try to add some sort of shielding around the ovens to keep air currents from beating on the copper block
  • One oven has "bad wiring" of some sort and I see ~15 mK rms fluctuations IN LOOP over a couple seconds. I will probably want to rewire this
  • I need to think more on what quasi phase matching means, and if the phase mismatch in the quasi phase matched case (the argument of the sinc^2) is just the phase difference between the IR and green field leaving the crystal (modulo pi)
• I shall also add some pics and schematics of the circuitry for later date debugging purposes

Does it happen if you disconnect the IR PDs but touch the outside of the BNCs?

Update on "The Effect" (The Voltage Noise at the Green PD Difference in my circuit changes depending on whether or not I have light on my IR PDs)

I checked the circuit a little more in detail, and I see the behavior at the output of the transimpedance amplifier for both PDs.

• I am using LT1792
• I short + to ground
• I put a 50 Ohm between - and ground
• The feedback is a 100k resistor, and a 100 pF capacitor
• Again, the behavior is a big noise increase when the IR circuitry is going.
• I unplugged the control signals from their respective destinations (PZT Drive box and AOM driver), but the behavior persists.
• With the box off, and the MZ Locked, this noise is (slightly above) about the same as the "MZ Phase Noise" above 10 Hz
• Below 10Hz the phase noise (or not this noise) is higher
• If I disconnect the IR PDs from my box the effect is gone (I'm using the same power supply for the ISS box and my box - I was worried that I was tugging the ground around somehow)
• I pulled EVERY op amp from the IR paths (including the feedback) - just 6 left. The effect persists.
  • The above op amps are the transimpedance amplifiers for each PD, the balancing op amps, and the sum and difference op amps.
• If I connect either one of the IR PDs (with a terminator on the other input) into either of the inputs, the effect is there

I noticed some funny behavior I couldn't explain.

• I take my Green PD inputs off on my box, and replace them with 50 Ohm terminators
• I look at the difference signal downstream
• This difference signal noise level changes by almost two orders of magnitude depending on whether or not I have the IR PDs blocked.
• This seems WRONG
• I think the green voltage noise level shown for with the IR PDs illuminated is higher than the level measured at around the same setup.

So it seems

• The green difference noise is lowest when I have no light on the IR PDs and 50 ohm terminators on the green inputs.
• The green difference noise is highest with terminators on the inputs and light on the IR PDs
• The green phase noise is between these two levels when we measure it...

And here is a newer spectrum.

I am having "transient problems". On a 30 - 100 second time scale, I have glitches which pop my spectrum up by about an order of magnitude in most of my measurement band.

• I was able to capture a spectrum of ~6 averages with no transient glitch by taking many many spectra.
• I have no "real" reason to believe that this is not somehow coming from my crystals, so I am wary of ignoring it
• I also noticed <link later post> that my green noise level has some strange behavior originating in the electronics box I built
• This is at the 3*10^-18 level in rms. I can buy another factor of two here if I did subtraction above 10 Hz, but the low frequency stuff is the other half, and (I think that it) isn't coherent between IR / Green
• This was taken with the Agilent 35670A

I realized I hadn't posted a characterization of the ISS I am using, so here that is...

• I blocked each arm of my Mach Zehnder
• I measured the Green Sum and the IR Sum with the new ISS Loop <Link here> open and closed
• I plotted the RIN from these (attached)
• I have also included the voltage noises for reference

I added some EPICS channels to the Hartmann sensor softIoc and then added these to be recorded in the frames.

I then killed the daqd process on fb1 so it would start afresh.

• /cvs/cds/caltech/target/softIoc/HWS.db      - the file with the HWS EPICS channels
• /cvs/cds/caltech/chans/daq/C4TCS.ini            - the .ini file telling the frame builder what channels to record.

 In true CVI fashion they don't have any of the mirrors that we want.  I'm not actually certain what they keep in their warehouse but it certainly isn't optics.

Delivery time is going to be 4weeks (make that more like 6) for both the output mirror (another 2" diameter BS1- 99%) and the beam splitter for the Mach Zehnder (2" BS1 50%).  I'm going to put through the order for them now.

Edit - That's the order through Techmart now (5/18).

fixed the problem with /dev/sdc. fb0 is now working again, both fronend codes running and daqd writing data

 FB0 woes:

Trying to get FB0 back up. The cleaning today seems to have borked it somewhat. I guess it got unplugged / shutdown improperly?

Currently on house N of the thing trying to reboot.

Errors are in sdc1. It periodically comes up and says somethign about the inodes being bad. So far:

"Too many illegal blocks in inode _____"

Clear inode<y>?

and

Illegal block #X in node Y. CLEARED

and back again to

/dev/dsc1 contains a file syste3m with errors, check forced.

lab temperature is almost back to normal. They fixed the problem with the hot water supply for the heating of the cold air.

someone must have changed something. today the temperature dropped by almost 3 Kelvin since 10:00 this morning, but not only in our lab. It's currently chilly 65F in the ATF, brrrr. I turned up the temp to 72F. So don't forget to bring your winter coat if you wanna work in the lab right now

i installed a sensor in the corridor, next to the PSL door, on the right hand side (the only place where i could put it).

The data is C6:PEM-SENS1_xxxx, where xxxx are all the channels we provide like temp, humidity, pressure etc.

Data is valid from now on...

did someone enter the lab this morning already? if yes, can you remember when?

 Phillip  and Tim from Caltech physical plant talked with us today down in the lab about what to do about the HVAC.

They're going to swap out the pre-filters in the air handling room, install another flow meter to check the lab air pressure, and install another temperature sensor in the lab (underneath the south cabinets).

We should record in the ELOG some of the temperature transients we're having now that we think are associated with opening the lab doors. Then we can do a quantitative before and after test to see how things change afterwards.

We probably should also put a temperature sensor out in the hallway to see what's going on out there.

I guess that since it is some sort of CSV file it displays it as text.  Indeed it is the particle counter in the ATF lab. 

I believe this is data from the particle counter that has been on the gyro table. Alastair offered to upload it to the elog for me since I had several things to do this afternoon, and Frank needed to use ASUS(?) for a while.

I deleted Alastair's 8000 line elog spam and have attached it here as a zip file. Its called 'DataForAir.gz', but I'm not really sure what this data is from (i.e. what kind of sensor, what location, etc.)

 I have the Kodak camera in my office while I am transferring pictures of the duct sealing. I will put it back in the lab before I go home.

EDIT: I found a small error in the code. I was not including the right factors of r(j) for the transmitted fields in the line beginning coefs_e(:,j) = ... . This has been fixed now.

I made a MATLAB function that will calculate and plot a reflection/transmission spectrum for an arbitrary cavity (any number of mirrors, any distances between them). The m file is attached, and I have pasted the documentation below for reference. I figured it might be useful for this oft-encountered calculation.

Here is a sample output for the aLIGO PMC:

%ARBCAV - ARBitrary CAVity calculator

%Zach Korth - 05/08/2010

%

%arbcav(T,L,loss,lambda,num_pts) takes information about an arbitrary cavity

%and returns a transmission/reflection coefficient spectrum for the cavity,

%as well as the finesse.

%

% Arguments:

%

%   T: vector containing the power transmission coefficients (i.e. T(i) = T

%   of the ith mirror)

%

%   L: vector containing distances between mirrors (i.e. L(1) = distance

%   from mirror 1 to mirror 2; final element is the distance from the last

%   mirror back to mirror 1). Note that for a linear cavity you must

%   treat the return leg as an extra distance (e.g. for a linear 2-mirror

%   cavity of 1-m length, L = [1 1], etc).

%

%   loss: power loss of each mirror--assumed to be the same for all

%

%   lambda: nominal laser wavelength

%

%   num_pts: number of points for scan

%

% Outputs:

%

%   finesse: duh

%   

%   coefs: a matrix of size(num_pts,length(L)) containing the power

%   coupling coefficients between the input coupler (mirror 1) and each

%   cavity mirror (i.e. coefs(:,1) = the cavity reflection coefficient,

%   coefs(:,j) = the transmission coefficient between the input coupler and

%   the jth mirror

%

%   delta_f: frequency vector along which to plot coefs(:,i)

%

 I've calculated the predicted cavity power transmission and reflection coefficients using the T values that I measured the other day. I used an estimated scatter loss of 150ppm per mirror, which is reasonable according to Frank.

I measured the T of each cavity mirror using both S and P light, because I was concerned that there was some funny polarization rotation business going on inside our cavity. Just looking at the result for S, though, I'm pretty sure we can explain the extra light output in an easier way:

When we first calculated the power that should be coming out of the curved Y1S mirrors, we used the nominal Y1 transmission "spec" of 0.5% to calculate the circulating power in the cavity (using the power we measured at the output). We then multiplied this circulating power by the 50ppm spec for the Y1S. In talking to Frank, I learned that CVI gives the 99.5% R spec for the Y1 to cover their asses in case scatter loss is ridiculously high, and that its T is usually right around the 50-100ppm that I measured the other day. This means that the circulating power in the cavity was much higher than we thought, and so the light escaping the cavity through the curved mirrors was reasonable.

This result can be seen graphically above, as the transmission through the curved mirror(s) is of the same order as that through the output coupler. Needless to say, this is a bad impedance matching solution. What we want is to have most of our light escaping through the output (of course!), and a cavity that isn't dominated by losses. For example, if the transmission through the output coupler were truly 0.5%, we would have something that looks like this:

This shows even more clearly why we thought we had too much light coming out of the curved mirrors.

Side note: I think we may have been using an incorrect formula to calculate finesses in the past. From now on, I will use the proper definition

where rho is the fraction of the power remaining in the cavity after one round trip (with no pumping field), i.e., rho = 1 - total round-trip loss. The approximation is good for high finesse.

someone removed the network connector from fb0. My guess is that that happened by accident the last time someone moved the rack. fb0 is not mounted in the rack right now so the main network cable is a bit short. If the rack is moved to far you kill the main network connection by pulling the plug. This of course kills the computer as the RT core and the frontend code is mounted using nfs located on fb1. Same for doing simple commands like ls. The system stops working as soon as you try to access data from a mounted nfs directory which doesn't exist anymore. After reconnecting the network everything was fine. However in order to avoid problems i rebooted the machine...Both frontend codes, atf and psl are back online...

the manual how to use the particle counter is where the rest of the manuals are. If you want to use it read the manual BEFORE you touch it and reconfigure everything so that nothing is working anymore. If you still don't know how to use it or can't find the manual you can ask me or simply google for it. If you only want to read the latest values you don't have to program or even touch the thing at all. Simply connect it to a computer via serial port. Right now it is taking data every 10min or so for a short period of time. Up to 8000 measurements are stored in the device, starting to overwrite the oldest ones when reaching the limit.

The software is installed on the small laptop, the same we use for the WinCam. There is an icon on the desktop, i think the software is called "comet" or so. You have to connect the cable with the USB-to-serial adapter. The cable has a label "particle counter". Plug everything together,  run the software, and from the menu simply choose something like read everything, meaning data and settings. That's it.

Data is saved in comma-separated  textfile, using current date and time. Sometimes it happens that it can't plot the data later on because the data is corrupt. I don't know why this happens but it simply forgets some characters while reading from the serial port or saves it twice. As it is a simple textfile this is easy to fix.

 saw it yesterday in one of the drawers of the right desk. Sometimes it is covered by other shit...

Have we made any changes since then?

I tried to ssh into fb0, and ws1 doesn't see it. I can login to fb0 via the keyboard on the rack, but it froze when I did an "ls". I don't know if I was not supposed to do that for some silly reason.

I have yet to be able to use fb0 since the crashes started (maybe a month ago?)

 Ah-haha-HA-HAHAHAHAHAHAAAAAA!

Sealing is done.  They left the air on this time while doing it so they could get as many small leaks as possible.  Anyway, that means that the air is still on in the lab and they have finished for good now.

Gyro key is in the top drawer of the center workstation.

The guys are back this morning to reseal the vents. There are some green marks around the place but also what looks like new red ones over the top of the silicon they laid down two weeks ago.

I shutdown the 35W laser using the 'System Off' option on the touch-screen and then turned off the NPRO and removed the key. That key is now in the second drawer down in the first desk. The key for the Gyro NPRO was not in the laser.

There have been no updates since this post, I will remedy that. Hyperlinks to later more detailed elog posts will be added as they pop up.

Summary of situation at time of above post (When Matt Evans, Hartmut, Koji and I had a work party): This was around the time of the first big frontend crash, so we went totally analog

• Locked PMC with an SR560 (30Hz pole and ~1000 gain)
• Added locking to the Mach Zehnder via stuffable breadboard and PZT
• Changed ISS to use the PD SUM as the error signal (because there seemed to be beam path induced sensing noise from using the PMC Trans)
• Calibrated the Mach Zehnder via the Lissajous of the PD difference signals at each color
• Locked the MZ, using the difference in the IR PDs as the error signal
• Looked at the green difference for bounding "excess phase noise"
• The spectra taken showed "glitchiness" and to get a low noise spectrum with ~5 averages, we needed to take ~10+ spectra. (Weird transient glitches)
• Obtained the above linked plot for ~4 or 5 averages.

Calibration from this night

• Green: [min,max] = [-4.7,7.2] = 11.9 Vpp (PDs Very Unbalanced)
• IR: [min,max] = [-8,8.5]  = 16.5 Vpp (PDs Less Unbalanced)

After that, the thinking was to solder everything together and eliminate the general crappiness of the stuffable breadboard (crazy glitchy pickup). The frontend was still in some broken state

• I took some old NIM box from our bridge subbasement E lab and gutted it.
• I took about a week to make the circuit for the PD readout (slightly modifying crappy things as I went) - I will make this a sub entry later -
• The circuit included
  
  • Green transimpedance amplifiers
  • PD Balancing via Potentiometers
  • Sum and Differencing circuits for both wavelengths
  • Mach Zehnder locking servo circuit
  • ISS servo circuit
  • Regulators
• I also added surgical rubber hosing below the box to try to provide a better air seal
  
  • I am somewhat close to the aperture on my plastic box now - maybe a few beam widths
  • I also haven't done any blocking of where the wires go into the box, and may need to stuff this with some sort of foam
• AT SOME POINT IN TAKING THE BOX ON AND OFF I MESSED UP THE ALIGNMENT INTO THE MACH ZEHNDER: I SAW THE KNOB ON AN INPUT MIRROR VISIBLY MOVE
• I spent maybe 10 minutes playing with the knob and did some crude qualitative alignment into the ovens,
• I need to do something much more quantitative b/c one oven is damaged (linkback): loss / beam quality / efficiency is rather sensitive to exactly what optical axis I'm using through the crystal

I then tried to put everything together and get a phase noise spectrum. Koji helped me figure out some set of things I was doing wrong.

• I realigned the Mach Zehnder (and was able to set the absolute phase difference between the two wavelengths via translational movement of the beam axis)
• The above is not obviously "bad" to do because there is already some trade off between contrast in each wavelength (due to angular dispersion), so it's just something to consider in my choice of (mis)alignment
• I balanced the PDs by balancing the extrema which the Mach Zehnder difference signals swept through (was able balance to ~ 1%)
• I locked the IR difference, and suppressed the error signal
• I looked at the green difference, and it was not suppressed (Plot number 1)
• The phase noise I saw in the green was inconsistent with the first plot linked (before I took everything apart to solder it together)

Calibration from this night I AM NOT POSITIVE THAT I DID NOT CHANGE THE GREEN TRANSIMPEDANCE RESISTOR (it is now 100k), and my balancing effects this up to 30%

• Green: [min,max] = [-3.5,+3.5] = 7 Vpp (59% of previous calibration voltage)
• IR: [min,max] = [-3.6,3.6]  = 7.2 Vpp (44% of previous calibration voltage - these TIAs are the same)

I Currently think that I increased the scatter in my ovens, and I am IR scatter limited on the green PDs. More to come.

I aso Adjusted the Power into the Mach Zehnder

Haven't found anything in the literature supporting a polarization rotation in a planar cavity, with the exception of one article discussing the effect in rotating ring cavities(!). I highly doubt that the earth's rotation would cause such a major effect via this pathway (if it did, then we could build a pretty sensitive gyro just by measuring the polarization angle shift!).

It also seems unlikely that the plane of the high-extinction PBS we used to isolate the S beam for the input should be misaligned with respect to the cavity plane by enough to account for this effect. I was going to crunch some numbers (i.e., calculate the P-finesse of the cavity and thereby determine the expected transmission through the cavity mirrors of some residual P input light), but it is difficult to find data for Y1(S)-S transmission of P. Not to fear, as we can just measure this.

It appears there was a momentary lapse of experimental reason when we all first measured the excess transmission, as there is no record of the measured output in the elog, nor did we measure the polarization of the escaping light. I suppose it was just easier to say "damn those CVI bastards and their crappy crap".

Troubleshooting scheme:

• Measure T_P for Y1, Y1S cavity mirrors, and calculate expected cavity P-belching for a conceivable residual P in the input beam.

If the result seems consistent with observation, then we simply use the cavity itself as a polarization reference for the input beam (as we perhaps should have done in the first place). If this doesn't account for the effect:

• When we get the cavity set up again (after moving it down the table), and inevitably observe the same problem, GET NUMBERS FOR THE TRANSMISSION AND DETERMINE THE POLARIZATION
• Use the aforementioned information to figure out WTF is going on.

 Before doing the full T vs. angle test on the 6-m mirrors we have been using (so that we could send them back to CVI), I figured I would just test them at 45°, outside of the cavity, using the same simple configuration I used to measure the new 9-m mirror. What I found was that---for both 6-m mirrors---the transmission was in the range of 50 - 100 ppm, roughly independent of angular tuning away from 45° on the order of +/- 5°. This is pretty much consistent with the part specifications from CVI for Y1S, especially considering that they have been on our table for a couple months.

So.. why did we observe a disproportionate amount of light escaping these mirrors with the cavity locked? Good question. We suspect that it might have something to do with polarization issues, as a macroscopic polarization shift in the cavity would certainly account for the spurious transmission.

I don't know much about what could cause this, but I am reading. The good news is that our mirrors are all what they are supposed to be (yippidy doo da).

I fully endorse this plan - make sure to send the bad mirrors back to CVI and get credit for them. It'll make the cylindrical lenses free. Also make sure that the R/T data for them is solid.

 Pending a T measurement on the second 9-m Y1S mirror, we will be installing these in the place of the 6-m ones that are in the cavity right now. The reason is twofold:

1. the T of the 6-m mirrors appears to be way off of what it should be, for one reason or another, and
2. our final arrangement has always called for the 9-m mirrors, as they give us better HOM rejection and a lower optimum modulation frequency of f_mod = 26 MHz (vs. the 33 MHz for the 6-m's).

The input waists for the cavity given these mirrors (the average, for spherical MM, and the two orthogonal waists, for cylindrical) are:

• R = 9 m:                                  w_0,avg =  928.6 µm
• R_x = R cos 45° =   6.643 m:        w_0x =  842.9 µm
• R_y = R/cos 45° = 12.728 m:        w_0y = 1019.7 µm

Since we will have to realign the cavity after replacing the mirrors anyway, we are going to slide everything down the table a pinch so that we can make room for the final MMT (the table is just too crowded for a proper IOO solution right now). After this, we will work out both a cheap (read: quick) modematch to the the average, and a full cylindrical modematch. The former will allow us to continue working on the experiment while we wait for parts for the latter to arrive.