I've done a first pass at trying to arrive at a mode-matching solution for the X-end table once we swtich the lasers out. For this rough calculation, I used a la mode to match my seed beam (with z = 0 being defined as the shutter housing on the current position of the Innolight laser head, and the waist of the beam from the NPRO being taken as the square-root of the X and Y waists as calculated here), to a target beam which has a waist of 35um at the center of the doubling oven (a number I got from this elog). I also ignored the optical path length changes introduced by the 3 half-wave plates between the NPRO and the doubling oven, and also the Faraday isolator. The best a la mode was able to give me, with the only degrees of freedom being the position of the two lenses, was a waist of 41um at the doubling oven. I suppose this number will change once we take into account the effects of the HWPs and the Faraday. Moreover, the optimized solution involves the first lens after the NPRO, L1, being rather close to the second steering mirror, SM2 (see labels in Attachment #2, in cyan), but I believe this arrangement is possible without clipping the beam. Moreover, we have a little room to play with as far as the absolute physical position of the z=0 coordinate is - i.e. the Lightwave NPRO head can be moved ~2cm forward relative to where the Innolight laser head is presently, giving a slightly better match to the target waist (see attachment #3). I will check the lenses we have available at the 40m to see if a more optimal solution can be found, but I'm not sure how much we want to be changing optics considering all this is going to have to be re-done for the new end table... Mode-matching code in Attachment #4...
I looked in the optics cabinet to see what lenses we have available, and re-ran the mode-matching calculation to see if we could find a better solution - I'm attaching a plot for what looks like a good candidate (optimized mode-matching efficiency for the X mode is 100%, and for the Y mode, it is 97.98%), though it does involve switching "L1", which is currently a 175mm efl lens, for a 125mm efl lens. I've also indicated on the plot where the various other components are relative to the optimized positions of the lens, and it doesn't look like anything is stacked on top of each other. Also, the beam width throughout is well below 4.7mm, which is the maximum cited width the Faraday can handle, as per its datasheet. "L1" doesn't quite get the waist of the beam to coincide with the geometrical center of the Faraday, but I don't think this is requried? Also, I've optimized the mode matching using the measured X width of the beam (red curve in Attachment #1), and have overlaid the calculated Y width of the beam for the optimized position of the lenses (red curve in Attachment #1). The target waist was 35um at the center of the doubling oven, which the X profile achieves, but the Y profile has a width of 32 um at the same point.
In all the calculations, I've not accounted for possible effects of the HWPs and the Faraday on the beam profile....
Tonight we embarked on the laser swap. In short, we have gotten ~210mW through the faraday doubler, but no green light is apparent. The laser outputs ~300mW, so it's not exactly a work of art, but I still expected some green. More work remains to be done...
Gautam took numerous photos of the table before anything was touched. One lens was swapped, as per Gautam's plan. The innolight laser and controller are on the work bench by the end table. The lightwave is on the table and on standby, and is not hooked up to the interlock mounted on the table frame, but instead one below the table directly next to the controller. The ETMX oplev laser is turned off.
Steve pointed me to an old elog by Zach where he had measured the waist of the 1W Innolight NPRO. I ran a la mode with these parameters (and the original optics in their original positions prior to last night's activities), and the result is in reasonably good agreement (see Attachment #1) with my initial target waist of 35 um at the center of the doubling oven (which I presume coincides with the center of the doubling crystal). The small discrepancy could be due to errors in position measurement (which I did by eye with a tape measure) or because I did not consider the Faraday in the a la mode calculation. However, I wonder why this value of 35 um was chosen? In this elog, Kiwamu has determined the optimal waist size to be 50um at the center of the doubling crystal. Nevertheless, as per his calculations, the doubling efficiency should be non-zero (about 1% lower than the optimum conversion efficiency) at 35um or 70um, so we should be able to see some green light as long as we are in this fairly large range. So perhaps the fact that we aren't seeing any green light is down to sub-optimal alignment? I don't think there is a threshold power for SHG as such, its just that with lower input power we expect less green light - in any case, 200mW should be producing some green light... From what I could gather from a bunch of old elogs by Aidan, the Raicol PPKPT crystals have dimension 1mm x 1mm x 30mm (long axis along beam propagation), so there isn't a whole lot of room for error perpendicular to the direction of propagation... I wonder if it is possible, for the initial alignment, to have the top cover of the doubling oven open so that we can be sure we are hitting the crystal?
Some updates on the laser swap situation:
2. Implementing the new solution:
As I check the manual of the Innolight (pg17) and the datasheet of the Lightwave, I wonder if the Quarter Wave Plate that was placed immediately after the Innolight laser head is even necessary now - I assume the purpose of the combination of QWP+HWP was to turn the elliptically polarized light from the Innolight into linearly polarized light before the Faraday. But the Lightwave already produces linearly polarized light. I will check out what is the configuration on the Y-end table...
After the discussion at the meeting, I decided to go ahead and open the top of the oven so that I could get a visual on where the crystal was located - this helped in the alignment, and I was able to get some green light out of the oven. I had to tweak the position of the Doubling oven a little (with the top open) in order to align the crystal to the beam axis. However - I was only able to get ~140uW of green light going into the Faraday. I had measured the power at various points along the beam path recently with the old setup. We used to have ~860uW of green going into the Faraday there. To see if I could improve the situation a little, I checked that the beam was reasonably centered on both apertures of the IR Faraday, and then removed the irides upstream of the doubling oven. These were preventing me from placing the lenses exactly as per the a la mode solution. Once the irides were removed, I moved the lenses to their optimal positions as best as I could with a tape measure to mark out distances. I then further tweaked the position of the doubling oven using the 4 axis stage, monitoring the green power while doing so. The best I could get was ~200uW. Perhaps the positions of the lenses need to be optimized further. I also checked the IR power before and after the IR Faraday - these numbers are ~260mW and ~230mW respectively (I maximized the transmitted power through the Faraday by rotating the HWP, the QWP that was in the beam path has now been removed as the Lightwave outputs linearly polarized light), and compare favourably to the numbers in the old setup. Doing a naive scaling accounting for the fact that we have less power going into the doubling crystal, I would expect ~700uW of green light coming out, so it looks like the mode matching into the doubling crystal is indeed sub-optimal. However, now that things are roughly aligned, I hope the optimization will go faster...
It shipped out for repair evaluation.
Arrived to Hayward,CA 2016Feb16
After carefully tweaking the mode-matching of the IR into the crystal and the four-axis translation stage on which the doubling oven is mounted, I managed to recover 800uW of green power going into the green Faraday. Considering we have ~225mW of IR power coming out of the IR faraday (and roughly that amount going into the SHG crystal), I'd say this is pretty consistent (if not slightly better) with a recent power budget I had made for the X end. The amount of green power we get out of the doubling crystal is very sensitive to the alignment of the crystal to the beam axis. I suspect we could improve the situation slightly if the mode-matching lenses were mounted on translational stages so we could tweak their position, but the current situation on the X endtable does not provide space for this. In any case, I'd say we are at least as good as we were before, and so this should be an adequate fix until the new end-table is installed (though I don't know why we aren't seeing the predicted SHG conversion efficiency of 3-4% as predicted by Kiwamu's calculations, we are getting more like .36% conversion efficiency)...
Because the alignment of the beam before the doubling oven had changed, I had to adjust the steering mirrors to make the green beam go into the green faraday. I had placed a couple of irides for the green beam as a reference of the old path into the arm, and I used these to adjust some of the green mirrors to center the green beam on these. However, I did not observe any flashes in the arm. I will check if we are still mode-matched to the arm, and if the lenses downstream of the doubling oven need to be moved....
800e-6 / 0.225^2 = 0.016
I thought Kiwamu had roughtly 2%/W.
Eric and I spent some time yesterday night trying to recover the green in the arm after the laser swap. The problem essentially was that though I was getting ~800uW of green out of the doubling oven, the mode wasn't clean, and hence, the beam profile looked really messed up just before entering the arm cavity.We got to a point where we thought we were getting a good mode out of the doubling oven (as judged by propagating this beam onto the wall with the help of a mirror). But we were only getting ~400uW of green power. I tried tweaking the alignment of the oven on the 4 axis stage for a while, but was not able to improve the situation much. So I decided to start from scratch:
I am beginning to wonder if this ellipticity is inherent from the IR beam from the laser? My beamscan results suggest that the beam is more divergent in the "P direction" as compared to the "S direction", which is borne out by these photographs. And if this is indeed the case, do we need to add cylindrical lenses to correct this?
Unrelated to this work: The ITMX Oplev seems to have wandered off so the X arm won't lock. I am not realigning the Oplev for now, but am turning the ITMX Oplev servo off for the night.
Perhaps related to my work on the endtable: The ETMX oplev MEDM readings seemed to be frozen, though there was red light on the QPD on the endtable. Checking the CDS overview screen, I saw that all models on c1iscex had crashed. I sshed into c1iscex and restarted all the models, but the IOP block remained red. I checked the datetime, and found that this was wrong - so I followed the instructions here, but the "Diag Word" block remains red. I am shutting down the watchdog for ETMX and leaving this as is for now... This seems to have happened before...
I tried aligning the green beam, elliptical as it is, to the arm by using the various steering mirrors after the doubling oven. The following was done:
Given that we were seeing green flashes in the arms, I tried to see if I could get the green locked to the arm in a nice mode. For a start, I tried hooking up the PDH box and LO using the same settings as was being used previously. However, this did not work. I suppose we will have to do the whole AM/PM measurement for the Lightwave as well before we can determine what would be a suitable frequency for the LO. The AM measurement was relatively straightforward, I just repeated the same steps as detailed here. The two attachments show the AM response (one from 10kHz to 5MHz, the other for a narrower range of 100kHz to 1MHz, both with an excitation amplitude of 0dBm). To see if I could guess some sweetspot for operation, I tried setting the LO frequency to the two marked notch frequencies but was unsuccessful in getting the PDH lock going. At the moment, the alignment for the optics that picks off the IR after the doubler and routes it to the fiber are ccompletely misaligned, I will align these and do the PM measurement tomorrow and then we should conclusively be able to say what the appropriate frequency is to actuate on the PZT.
Unrelated to this work: the KEPCO high voltage power supply that drives the green steering mirror PZTs was switched off - I suppose this has been the case since the power outage last week. I turned it back on and reset it to the nominal settings: Vout = 100V, and Imax_out = 10mA, the driver board is currently drawing ~7mA which I judged to be consistent with the values labelled on the unit.
After the discussion at the meeting today, I decided to try and lock the green by sweeping through PZT dither frequencies in the vicinity of 200kHz without worrying about the AM/PM ratio for now. I was able to lock the PDH loop relatively quickly, at an empirically determined PZT dither frequency of 213.873kHz, 2Vpp (the amplitude was copied from the value at the Y-end). For today's efforts, I borrowed the sum+HPF pomona box from the Y-end, I will make a replica given that we are using Lightwave lasers at both ends now. After adjusting the PZT sliders and lenses on the translational stages at the endtable to maximize the green transmission as best as I could, I was able to get GTRX up to about 0.07 - this is far off from the value of ~0.25-0.3 I seem to remember us having with the old setup, even though we have more green light into the arm cavity. I will take a measurement of the loop transfer function to see what sort of bandwidth we have...
I spent some more time today trying to optimize the modulation frequency and amplitude for the X end PDH, and the alignment/mode-matching of the green to the arm. Some notes:
I continued the hunt for a green beatnote today - I decided to take the output from the RF amplifiers sitting on the PSL table and directly connect it to the analyzer in the control room while I swept the temperature of the end laser 10,000 counts on either side of a temperature at which I had taken this measurement - so I expect the beatnote should be found somewhere in this neighbourhood. But I did not see any peaks throughout the sweep. I re-checked that the mode overlap onto the BBPD is reasonable. We have considerably less transmitted green power from the arm now than we did before the laser swap (by a factor of ~3) but I still expected to see some sort of beat signal.
It would be handy to have the IR beat set up as well for this process, but as mentioned in a previous elog, I was getting only ~0.1 mW of IR power incident on the coupler at the end table last week. As I had suspected, tweaking the alignment of the steering optics for the pick-off IR beam after the doubler improved the situation somewhat, and I am now getting about 1mW of IR power incident on the coupler at the end table. But I've not been able to adjust the alignment into the fiber at the end such that I get any IR light at the PSL table.
[Koji, Johannes, gautam]
With Koji's and Johannes' help, I managed to resolve the coupling the pick-off IR beam into the fiber at the X end. I will put up a more detailed elog about how this was done - but in summary, we have about 31% coupling efficiency into the fiber, which isn't stellar, but I felt this was adequate to find a beatnote. Koji also pointed out that the collimation telescope attached to the fiber at the X-end is poorly mounted - this is something to fix when we swap endtables, but this was not addressed right now because if we were to adjust this, we would also have to adjust the mode matching into the fiber.
I then attempted to tune the temperature to find the IR beatnote. While doing so, I noticed some strange features of the controller - there are essentially two display modes relevant to laser crystal temperature, one which allows us to change the setpoint and one which is an actual readback of the temperature (this one can't be adjusted). While tuning the temperature, I noticed that the latter display ("LT") did not change in value. On a hunch, I disconnected the "SLOW" control BNC on the front panel, and voila, I was able to tune the setpoint and observe the measured temperature shift accordingly. I was thus able to find a reasonably strong IR beatnote (-9dBm) at T ~ 44.6 deg C (the beat PD was set to 0dB attenuation, i.e. high gain mode). However, the moment I reconnected the SLOW control BNC, the beatnote vanished (it gradually shifted out of range of the HP network analyzer), and the same thing happens if I terminate the SLOW control BNC connector! I don't understand this behaviour, as the manual says that the range of voltages accepted to this input is +/-10V, so I would assume 0V means do nothing, but clearly this isn't the case, as the beatnote is being shifted in frequency by > 1GHz, and the tuning coefficient is listed as 5GHz/V in the manual. This situation needs further investigation.
Since I had a reasonable IR beatnote setup, I returned the HP analyzer to the control room and tried to see if a green beatnote was present as well - I first ran ASS, then maximized the green transmission using the PZT mirrors, but no beatnote is evident. The contrast isn't great, the ratio of AUX power to PSL power on the green beat PD is something like 5:1, so this probably requires some tuning as well. I will update this elog after today evening's activities...
Summary of work done tonight:
The good news: both green beatnotes have now been found. The problem was alignment on the green beat PD on the PSL table which I fixed. They are about -40dBm in amplitude (compare to -25dBm we used to see). But looking at the phase tracker Q output seems to suggest that there is adequate signal...
The bad news: the ALS noise still looks bad (see attachment)- I think the IR beat for the Y was perhaps marginally better. The beat amplitude for the X beat was optimized on the PSL table with the help of the oscilloscope. There may be some headroom for improvement with the Y beat.
I also did the AM/PM measurement for the replaced lightwave, chose an LO frequency based on this, and took the loop OLTF, plots to follow...
I've been a little behind on my elogs so here is an update of the end laser situation.
IR beat for X-end recovered
AM/PM characterization of newly installed Lightwave
Next steps in recovering ALS and trying to lock again
Since I could not determine how many volts at the LO input of the pomona box input corresponds to how many volts at the laser PZT, I measured the transfer function between these points using the Agilent network analyzer. The measured TF suggests that for a function generator output of 2Vpp, we get approximately 75mrad of phase modulation, which compares reasonably well with the value of 120mrad reported here. I did not attempt to further increase the LO output signal to push this number closer to 120mrad, as with 2Vpp from the function generator we get +7dBm at the mixer, which is what it wants - so I wanted to avoid any attenuators etc...
Attachments #2 and #3
After ensuring that we have appreciable phase modulation, I set out to measure the PDH OLTFs and adjust the gain on the uPDH boxes accordingly. The X end gain is at 6.0, and the Y end gain is at 4.0. Before measuring the Y-end OLTF, I adjusted the steering mirrors to increase GTRY to ~0.45. GTRX remains a paltry 0.05... But the UGFs seem satisfactory..
Finally, I took the ALS noise spectrum for the green beats. The beat note amplitudes on the network analyzer in the control room are still puny compared to what we had, -40dBm for Y and -45dBm for X. But the phase tracker Q values are ~1000 and ~3000 for X and Y respectively, which are pretty close to what these were if memory serves me right. There may still be some room for optimization of the PDH loop gains etc, and we could perhaps look at lowering the gain of the REFL PD at the X end? I also have yet to do the sweep for the 3 temperatures at which we can find a beatnote and park at the middle one...
These spectra suggest we could even possibly try locking? We are approximately a factor of 3 above the reference for X and on par with the reference for Y....
Unrelated to this work: I also realinged the PMC, PMC transmission is now 0.730V up from ~0.65V.
Why is the transmission of X green so low? Perhaps you can phase lock the IR and then scan the X frequency, using the X arm as the analyzer. i.e. put a slow ramp into MC2 to pull the PSL frquency and thus the green frequency. You can record a movie of the scan using the framegrabber and record the green transmission peaks to see how big the mode match is exactly (which modes are so big)
Oct. 15, 2016
Another attempt (following elog 8755) to extract the oven transfer function from time series data using Matlab’s system identification functionalities.
The same time series data from elog 8755 was used in Matlab’s system identification toolbox to try to find a transfer function model of the system.
From elog 8755: H(s) is known from current PID gains: H(s) = 250 + 60/s +25s, and from the approximation G(s)=K/(1+Ts), we can expect the transfer function of the system to have 3 poles and 2 zeros.
I tried fitting a continuous-time and a discrete time transfer function with 3 poles and 2 zeros, as well as using the "quick start" option. Trying to fit a discrete time transfer function model with 3 poles and 2 zeros gave the least inaccurate results, but it’s still really far off (13.4% fit to the data).
1. Obtain more time domain data with some modulation of the input signal (also gives a way to characterize nonlinearities like passive cooling). This can be done with some minor modifications to the existing code on the raspberry pi. This should hopefully lead to a better system ID.
2. Try iterative tuning approach (sample gains above and below current gains?) so that a tune can be obtained without having to characterize the exact behavior of the heater.
Oct. 16, 2016
-Found the raspberry pi but it didn’t have an SD card
-Modified code to run directly on a computer connected to the TC 200. Communication seems to be happening, but a UnicodeDecodeError is thrown saying that the received data can’t be decoded.
-Some troubleshooting: tried utf-8 and utf-16 but neither worked. The raw data coming in is just strings of K’s, [‘s, and ?’s
-Will investigate possible reasons (update to Mac OS or a difference in Python version?), but it might be easier to just find an SD card for the raspberry pi which is known to work. In the meantime, modify code to obtain more time series data with variable input signals.
I connected to the serial port using screen (through Terminal) and using Arduino's serial monitor and basically received the same strings that were received through python, so it's not a python issue. Checked the other TC 200 module and was also receiving nonsense, but it was all question marks instead of mostly K's and ['s.
This rules out a few possible reasons for the weird data. Next steps are to set up and configure the Raspberry Pi (which has been interfaced before) and see if the problem continues.
I've been trying to understand the green beat setup on the PSL table to see if I can explain the abysmal mode-matching of the arm and PSL green beams on the broadband beat PDs. My investigations suggest that the mode-matching is very sensitive to the position of one of the lenses in the arm green path. I will upload a sktech of the PSL beat setup along with some photos, but here is the quick summary.
Attachments #1 and 2: Simulated and measured beam profiles for the PSL and arm green beams. The origin is chosen such that both beams have travelled to the same coordinate when they arrive at the BBPD. The agreement between simulation and measurement is pretty good, suggesting that I have modelled the system reasonably well. The solid black line indicates the (approximate) location of the BBPD
Attachment #3: Mode matching efficiency as a function of shift of the above-mentioned fast lens. Currently, after my best efforts to align the arm and PSL green beams in the near and far fields before sending them to the BBPD results in a mode matching efficiency of ~30% - the corresponding coordinate in the simulation is not 0 because my length measurements are evidently not precise to the mm level. But clearly the mode matching efficiency is strongly sensitive to the position of this lens. Nevertheless, I believe that the conclusion that shifting the position of this lens by just 2.5mm from its optimal position degrades the theoretical maximum mode matching efficiency from >95% to 50% remains valid. I propose that we align the beams onto the BBPD in the near and far fields, and then shift this lens which is conveniently mounted on a translational stage, by a few mm to maximize the beat amplitude from the BBPDs.
Unrelated to this work: I also wish to shift the position of the PSL green shutter. Currently, it is located before the doubling oven. But the IR pickoff for the IR beat setup currently is located after the doubling oven, so when the PSL green shutter is closed, we don't have an IR beat. I wish to relocate the shutter to a position such that it being open or closed does not affect the IR beat setup. Eventually, we want to implement some kind of PID control to make the end laser frequencies track the PSL frequency continuously using the frequency counter setup, for which we need this change...
I tried to realize an improvement in the mode matching onto the BBPDs by moving the lens mentioned in the previous elog in this thread. My best efforts today yielded X and Y beats at amplitudes -15.9dBm (@37MHz) and -25.9dBm (@25MHz) respectively. The procedure I followed was roughly:
As per my earlier power budget, these numbers translate to a mode matching efficiency of ~53% for the X arm beat and ~58% for the Y arm beat, which is a far cry from the numbers promised by the a la mode simulation (~90% at the optimal point, I could not achieve this for either arm scanning the lens through a maximum of the beat amplitude). Looks like this is the best we can do without putting in any extra lenses. Still a marginal improvement from the previous state though...