DAQ reload/restart was performed at about 1315 PST today. The previous ini file was backed up as c1pem20120309.ini in the /chans/daq/working_backups/ directory.
I set the following to record:
The two JIMS channels at 2048:
[C1:PEM-JIMS_CH1_DQ] Persistent version of JIMS channel. When bit drops to zero indicating something bad (BLRMS threshold exceeded) happens the bit stays at zero for >= the value of the persist EPICS variable.
[C1:PEM-JIMS_CH2_DQ] Non-persistent version of JIMS channel.
And all of the BLRMS channels at 256:
Names are of the form:
On monday I intend to look at the weekend seismic data to establish thresholds on the JIMS channels.
256 was the lowest rate possible according to the RCG manual. The JIMS channels are recorded at 2048 because I couldn't figure out how to disable the decimation filter. I will look into this further.
The PEM model has been modified now to include a block called 'JIMS' for the JIMS(Joint Information Management System) channel processing. Additionally I added test points inside the BLRMS blocks that are there. These test points are connected to the output of the sqrt function for each band. I needed this for debugging purposes and it was something Jenny had requested.
The outputs are taken out of the RMS block and muxed, then demuxed just outside the JIMS block. I was unable to get the model to work properly with the muxed channel traveling up or down levels for this. Inside the JIMS block the information goes into blocks for the corresponding seismometer channel.
For each seismometer channel the five bands are processed by comparing to a threshold value to give a boolean with 1 being good (BLRMS below threshold) and 0 being bad (BLRMS above threshold). The boolean streams are then split into a persistent stream and a non-persistent stream. The persistent stream is processed by a new library block that I created (called persist) which holds the value at 0 for a number of time steps equal to an EPICS variable setting from the time the boolean first drops to zero. The persist allows excursions shorter than the timestep of a downsampled timeseries to be seen reliably.
The EPICS variables for the thresholds are of the form (in order of increasing frequency):
The EPICS variables for the persist step size are of the form:
I have set all of the persist values to 2048 (1 sec.) for now. The threshold values are currently 200,140,300,485,340 for the GUR1X bands and 170,105,185,440,430 for the GUR1Y bands.
The values were set using ezcawrite. There is no MEDM screen for this yet.
PEM model was restarted at approx. 11:30 Mar. 7 2012 PST.
Yesterday I adjusted the preweighting of my IIR fit to the transfer function of MC2, and also managed to reduce the number of poles and zeros from 8 to 6, giving a smoother rolloff. The bode plots are pictured here:
The predicted IIR subtraction was very close to the predicted FIR subtraction, so I thought these coefficients would lead to a better online filter.
However, the actual subtraction of the MCL was not as good and noise was injected into the Y arm.
The final comparison of the subtraction factors between the online and offline data showed that the preweighting, while it improved the offline subtraction, needs more work to improve the online subtraction also.
I measured the transfer functions in the delay line cables, and then calculated the time delay from that.
The first cable had a time delay of 1272 ns and the second had a time delay of 1264 ns.
Below are the plots I created to calculate this. There does seem to be a pattern in the residual plots however, which was not expected.
The R-Square parameter was very close to 1 for both fits, indicating that the fit was good.
I re-measured the transfer function for Cable B, because the residuals in my previous post for cable B indicated a bad fit.
I also realized I had made a mistake in calculating the time delay, and calculated more reasonable time delays today.
Cable A had a delay of 202.43 +- 0.01 ns.
Cable B had a delay of 202.44 +- 0.01 ns.
I applied a bandpass filter to the accelerometer huddle data as a pre-filter. The passband was from 5 Hz to 20 Hz. I found that applying this pre-filter did very little when comparing the PSD after pre-filtering to the PSD with no pre-filtering. There was some improvement though, just not a significant amount. For some reason, it also seemed as though the second accelerometer improved the most from pre-filtering the data, while the first and third remained closer to the unfiltered noise. Also, I have not yet figured out a consistent method for choosing passband ripple and stopband attentuation, both of which determine how good the filter is.
My next step in pre-filtering will be determining a good method for choosing passband ripple and stopband attenuation, along with implementing other pre-filtering methods to combine with the bandpass filter.
I updated the bandpass filter I was using, finding that having different stopband attenuations before and after the passband better emphasized the area from 3 Hz to 20 Hz. I chose a low passband ripple but high stopband attenuation to do this. My passband ripple was 2 dB, the first stopband was 25 dB, and the second stopband attenuation was 40 dB. As can be seen in the filter Magnitude plot, this resulted in a fairly smooth passband and a fairly step dropoff to the stopband, which will better emphasize the region I am trying to isolate. My goal was to emphasize the 3-20 Hz region 10-30 times more than the outside regions. I think I accomplished this by looking at the Bode plot, but I may have chosen the second stopband attenuation to be slightly too high for this.
The front panels for the ALS delay line box came in last week. Some of the holes for the screws were slightly misaligned, so I filed those and everything is now put together. I just need to test both front panels to determine if the SMAs should be isolated or not.
Koji had also suggested making the holes in the front and back panel conical recesses so that flat head screws could be used and would counteract the anodization of the panel and avoid the SMAs being isolated. I think if we did that then conductivity would be ensured throughout the panel and also through the rest of the box. I also think one way we could test this before drilling conical recesses would be to test both front panels now, as one has isolated SMAs and one has conductive SMAs. If the anodization of the panel isolated the SMA regardless, we could potentially figure this out by testing both panels. But, would it also be that it is possible that the isolation of the SMA itself does not matter and so this test would tell us nothing? Is there a better way to test if the SMAs are being isolated or not? Or would this be more time consuming than just drilling conical recesses as a preventative measure?
I updated my bandpass filter and have included the bode plot below in Figure 1. It is a fourth order elliptic bandpass filter with a passband ripple of 1dB and a stopband attenuation of 30 dB. It emphasizes the area between 3 and 40 Hz.
Below, I applied this filter to the huddle test data. The results from this were only slightly better in the targeted region than when no pre-filter was applied.
When I pre-filtered the mode cleaner data and then used an IIR wiener filter, I found that the results did not differ much from the data that was not pre-filtered. I'm not sure yet if I'm targeting the right region of this data with my bandpass filter, and will be looking more into choosing a better region. Also, I am only using certain regions of ff when calculating the transfer function, and need to optimize that region also. I uploaded the code I used to make these plots to github.
I was looking at the new seismometer data and plotted the coherence between the different arms of C1:PEM_GUR1 and C1:PEM_GUR2. There was not much coherence in the X arms, Y arms, or Z arms of each seismometer, but there were within the x and y arms of the seismometer.
I think the area we should focus on with filtering is lower ranges, between 0.01 and 0.1, because that it where coherence is most clearly high. It is higher in high frequencies but also incredibly noisy, meaning it probably wouldn't be good to try to filter there.
Ignacio and I downloaded data from the STS, GUR1, and GUR2 seismometers and from the mode cleaner and the x and y arms. The PSDs along the arms have the most noise at frequencies greater than 1 Hz, so we should focus on filtering in that area. The noise levels start dropping at around 30 Hz, but are still much higher than is seen at frequencies below 1 Hz. However, because the spectra is so low at frequencies below that, Wiener filtering alone injected a significant amount of noise into those frequencies and did not do much to reduce the noise at higher frequencies. Pre-filtering will be required, and I have started implementing a pre-filter, but with no improvements yet. So far, I have tried making a bandpass filter, but a highpass filter may be ideal in this case because so much of the noise is above 1 Hz.
I tested both of the front panels (conductive and isolated SMAs) with the ALS Delay Line Box by driving extremely close frequencies through the cables. By doing this, we would expect that a spike would show up in the PSD if there was crosstalk between the cables.
In the plots below, for the conductive panel, the frequencies used were
X Arm: 22.329 MHz Y Arm: 22.3291 MHz
For the isolated panel, the frequencies were
X Arm: 22.294 MHz Y Arm: 22.2943 MHz
This gives a difference of 100 Hz for the conductive panel and 300 Hz for the isolated panel. Focusing on these areas of the PSD, it can be seen that in the Y Arm cable there is a very clear spike within 30 Hz of these differences when frequencies are being driven through both cables as opposed to the signal being in only the Y Arm. In the X Arms, the noise in general is higher when both cables are on, but there is no distinct spike at the expected frequencies. This indicates that some sort of crosstalk is probably happening due to the strong spikes in the Y Arm cables.
Previously, I had gotten the same results for the conductive and the isolated front panels. Today, I sanded off the anodized part on the back of the conductive front panel. I checked afterwards with a mulitmeter to ensure that it was indeed conductive through all the SMA connectors.
I drove a frequency of 29.359 Hz through the X Arm cable and 29.3592 Hz through the Y Arm cable, giving a difference of 200 Hz. Previously, there would only be a spike in the Y Arm at the difference, while the X Arm did not change if the Y arm was on or off. Now that the panel is fully conductive, a spike can also be seen in the X arm, indicating that crosstalk may possibly be happening with this panel, now that the spike corresponds to both the X arm and Y arm. These results are only after one set of data. Tomorrow I'll take two more sets of data with this panel and do a more in depth comparison of these results to what had been previously seen.
Koji had suggested that I sync up the two function generators to ensure that they have the same base frequency and so that crosstalk will actually appear at the expected frequency. After syncing up the two function generators, I drove the following frequencies through each cable:
X: 29.537 MHz Y: 29.5372 MHz
X: 29.545 MHz Y: 29.5452 MHz
Each time, the difference between the frequencies was 200 Hz, so if there was crosstalk, a spike should appear in the PSDs at 200 Hz when frequencies are being driven through both cables simulataneously, but not when just one is on. We very clearly see a spike at 200 Hz in both the X arm and the Y arm with the conductive SMAs, indicating crosstalk. For the front panel with isolated SMAs, we see a spike at 200 Hz when both frequencies are on, but it is much less pronounced than with the conductive SMAs. It seems as though there will be crosstalk using either panel, just less with the isolated SMAs.
After testing both the Conductive and Isolated front panels on the ALS delay line box using the actual beatbox and comparing this to the previous setup, I found that the conductive SMAs improved crosstalk the most. Also, as the old cables were 30m and the new ones are 50m, Eric gave me a conversion factor to apply to the new cables to normalize the comparison.
I used an amplitude of 1.41 Vpp and drove the following frequencies through each cable:
X: 30.019 MHz Y: 30.019203 MHz
which gave a difference of 203 Hz.
In the first figure, it can be seen that, for the old setup with the 30m cables, in both cables there is a spike at 203 Hz with an amplitude of above 4 m/s^2/sqrt(Hz). When the 50m cables were measured in the box with the conductive front panel, the amplitude drops at 203 Hz by a factor of around 3. I also compared the isolated front panel with the old setup, and found that the isolated front panel worse by a factor of just over 2 than the old setup. Therefore, I think that using the conductive front panel for the ALS Delay Line box will reduce noise and crosstalk between the cables the most.
Today I finished fitting the transfer function to a vectfit model for seismometers T240_X and T240_Y, and then used these to filter noise online from the mode cleaner.
The Bode plot for T240_X is in figure 1, and T240_Y is in figure 2. I made sure to weight the edges of the fit so that no DC coupling or excessive injection of high frequency noise occurs at the edges of the fit.
I used C1:IOO-MC_L_DQ as the first channel I filtered, with C1:IOO-MC_L_DQ(RMS) for RMS data. I took reference data first, without my filter on. I then turned the filter on and took data from the same channel again. The filtered data, plotted in red, subtracted from the reference and did not inject noise anywhere in the mode cleaner.
I also looked at C1:LSC-YARM_OUT_DQ and C1:LSC-YARM_OUT_DQ(RMS) for its RMS to see if noise was being injected into the Y-Arm when my filter was implemented. I took reference data here also, shown in blue, and compared it to data taken with the filter on. My filter, in pink, subtracted from the Y-Arm and injected no noise in the region up to 10 Hz, and only minimal noise at frequencies ~80 Hz. Frequencies this high are noisy and difficult to filter anyways, so the noise injection was minimal in the Y-Arm.
I am starting work on the PSL table at the 40m. My goal is to lock the laser coming from the nearby table to the FP cavity and get a measurement of the response to a temperature step on the surrounding can.
I have to mode match the beam to the cavity. Specifically, I have to mode match to the beam coming from the PMC through the EOM to the polarizing beam splitter. Yesterday David and I measured the beam width at various distances (from a particular lens through which the beam traveled), and I fit that data using MATLAB to find the beam's waist size and location. However, I'm not convinced that the fit is any good, since we only took measurements at five spots and they had large error bars.
Here is the fit I obtained using fminsearch. The horizontal beam width measurements were smaller than the vertical width measurements, suggesting that the incoming beam was elliptical. I fit the data for each set of measurements separately and got two waist locations. The red trace is the fit for the horizontal width and the blue represents the vertical width of the beam. Averaging the two fitted waist locations and sizes gives
vert z_0= -1760 mm (waist location)
horiz z_0= -1540 mm (waist location)
vert w_0 = 0.286 mm (waist size)
horiz w_0 = 0.275 mm (waist size)
avg z_0= -1650 mm
avg w_0 = 0.281 mm
Here is the code I used:
I defined the function spotsize.m and then made a function gaussbeam.m that called it with input parameters and returned the least squares error. I then wrote another function twobeamfits.m that ran fminsearch to minimize the least squares error and made the above plot. I've pasted the code below.
function omega = spotsize(z_0, w_0, z)
function sse = gaussbeam(params,xvals,yvals)
%This f'n takes as its inputs
%three parameters (w_0, z_0, and lambda),
%a vector of x-values (distances),
%and an associated vector of y-values (spotsizes),
%It then generates a vector of fitted y-values by applying
%an exponential approach function (single pole), with the given parameters,
%to the x-values.
%It then returns the sum of the squares of the entries of the difference
%between the fitted y-vector and the actual y-vector
fityvals=spotsize(z_0, w_0, xvals);
error=(fityvals - yvals);% .*xvals;
% sse stands for sum of squares error
function [outputs] = twobeamfits(guesses, dists, vert, horiz)
%This f'n takes as its inputs
%two starting guess parameters (w_0 and z_0),
%a vector of distances (x-values),
%and two associated vectors of measured beam radii,
%the radius measured along the vertical axis
%and the radius measured along a horizontal axis (y-values).
%It then calls the gaussbeam f'n for each set of y-values and minimizes its output (sum of squares error)
%using the fminsearch f'n. It outputs the fit parameters it settles on.
%It then plots the input data, the fitted curves, and the residuals
fitvert=spotsize(vertparams(1), vertparams(2), dists);
spoterror=[.1, .1, .1, .1, .1]; %uncertainties, all in mm
fithoriz=spotsize(horizparams(1), horizparams(2), dists);
errorbar(dists, vert, spoterror, 'x')
errorbar(dists, horiz, spoterror, 'r*');
plot(points,spotsize(vertparams(1), vertparams(2), points));
plot(points,spotsize(horizparams(1), horizparams(2), points),'r');
xlabel('Distance z (mm)')
title('Gaussian Beam Fits')
ylabel('Spotsize w (mm)')
legend('Vertical Spotsize','Horizontal Spotsize','Vertical Fit',...
legend('Vertical Fit Residuals','Horizontal Fit Residuals',...
Later on I may repeat some measurements and try to gain more certainty in my fit. In the mean time I will use this beam profile for mode matching.
I found a mode matching solution to match the beam coming to the PSL table from the AP table so that I can lock the laser beam coming onto the PSL table to the reference cavity on the table. I determined that at the polarizing beam splitter, I want a beam with a q=(147+25.1i)mm (w0=58mm). This came from applying the ABCD matrices for three distances,
to a beam with q0 = 406.4i mm (w0=0.371 mm at the PMC).
I obtained the following mode matching solution, which I will try to implement on the PSL table:
The beam I have has waist 0.281 mm at -2.74 m (I set my origin at the polarizing beam splitter--the spot where I want my beam to match the beam coming from the PMC, so all waists are behind that point). These numbers come from the beam-profiling and MATLAB-fitting I did (see 5015).
The solution I chose was: f = 1145.6 mm at -0.95 m and f = 572.7 mm at -0.62 m. This may need to be changed however, if I need to add in some beam steering, which would increase the path length traveled by the beam.
I've been working on the PSL table to put together a setup so that I can measure the reference cavity's response to a temperature step increase at the can surrounding it. My first step was to mode match the beam coming from the AP table to the cavity.
I implemented my mode matching solution. I ended up using a different one from the one I last elogged about. Here is the solution I used:
Two lenses: f = 1016.7.6 mm at -0.96 m and f = 687.5 mm at -0.658 m. (I set my origin at the polarizing beam splitter--the spot where I want my beam to match the beam coming from the PMC, so all waists are behind that point). Below is what it should look like.
What I did on the table:
Here's a picture of the PSL table with the lenses and mirror I added. The beam is redirected by a mirror and then a polarizing beam splitter. Past the beam splitter is another lens (f=286.5 mm), which was already in place from the mode matching of the beam from the PMC to the reference cavity.
Here is a block diagram of my intended experimental setup:
I am going to try to lock the laser to the cavity given my preliminary mode matching and then go back and improve it later. My next step is to find a frequency range for dithering the voltage sent to the PZT. To do this I will:
I ended up having to switch to a different mode-matching solution, because I was unable to find the f = 572.7 mm lens. See my next elog entry (5069).
I am using a PDA255 photodiode to measure the power outputted by the NPRO beam on the PSL table. (I'm going to then use a network analyzer to measure the amplitude response of the PZT to being driven at a range of frequencies. I'll detect the variation in in response to changing the driving frequency using this PDA255.)
The PDA255 has an active area of 0.8mm^2 and a maximum intensity for which the response is linear of 10mW/cm^2. This means that a beam I focus on the PD must have a power less than 0.08 mW (and even less if the spot size is smaller than the window size).
I used a power meter to measure the beam power and found it was 0.381 mW.
The second polarizing beam splitter in the setup transmits most of the beam power, but reflects 0.04 mW (according to the power meter). I'm going to place the photodiode there in the path of the reflected beam.
Today I placed the PDA255 photodiode on the PSL table to catch the small amount of beam power reflected by the second polarizing beam splitter in my setup. I plugged the PD output to the oscilloscope to measure the voltage output and positioned the PD such that the voltage output was maximized. At best I was able to achieve a 300 mV DC output voltage from the PD, (which seems a bit low, as the PD is specified to go from 0 to 5 V and the specifications say that the response becomes nonlinear after 10 mW/cm^2 and my beam has an intensity of approximately 5 mw/cm^2. I would therefore expect to get more beam power but after over an hour of maneuvering, 300 mV was the highest voltage output I could get).
I am planning, tomorrow afternoon, to take a measurement of the amplitude response of the PZT driving the NPRO laser. I moved the 4395 spectrum/network analyzer to near the PSL table and connected the RF output to an RF splitter. I fed one output of that into the PZT and the other output into the R port on the network analyzer. I fed the PD output into the A port. I plan to measure A/R as a function of driving frequency, sweeping from 10 Hz to 30 mHz.
I also worked to improve the mode matching of the NPRO beam coming from the AP table to the reference cavity. I drove the temperature of the NPRO at 0.100 Hz with an amplitude of 0.300 V, which Koji told me corresponds to a 1GHz change in the laser frequency. The transmission from the cavity is being monitored by a camera connected to a TV monitor, and also by a PD connected to an oscilloscope. I then repositioned the second lens in my mode matching setup in an attempt to increase the transmission peaks from the zeroth order spacial mode and decrease the transmission peaks from higher order modes. I may have improved the mode matching slightly but I was unable to improve it significantly.
The ABSL beam had been blocked so that it wouldn't enter the interferometer. I moved the block so that the beam I've been using is unblocked by the beam going to the interferometer is still blocked.
I positioned a fast lens (f=28.7mm) a little over an inch in front of the PDA255 in order to decrease the spot size incident on the PD. I adjusted the rotation angle of the half wave plate to maximize the transmitted power through the PBS to the cavity and minimize the power reflected to my PD. I then adjusted the lens potion to fix the beam on the PD. The voltage output of the PD is now 150mW, but I have the ability to increase the incident power by rotating the wave plate slightly.
Now all I need is to set up the network analyzer again to record the amplitude response to modulating the PZT from 10 Hz to 30 MHz, reduce the input voltage into the analyzer using a DC block.
I rolled the network analyzer over to the PSL table (on the south side). I'm borrowing the DC block from Kiwamu's green locking setup. I'm going to first measure the amplitude response of a low pass filter to made sure that the analyzer is outputting what I expect. Then I will measure the laser PZT amplitude response. I plan to finish the measurement and return the network analyzer to it's usual location tonight.
Using a PDA255 on the PSL table, I measured the amplitude response of the NPRO PZT, sweeping from 10kHz to 5 MHz.
I took a run with the laser beam blocked. I then took three runs with the beam unblocked, changing the temperature of the laser by 10 mK between the first two runs and by 100mK between the second and third runs.
At the end of the night I turned off the network analyzer and unplugged the inputs. I'm leaving it near the PSL table, because I'd like to take more measurements tomorrow, probing a narrow bandwidth where the amplitude response is low.
On the PSL table, I'm still monitoring the reflected light from the cavity and the transmitted light through the cavity on the oscilloscope. I'm no longer driving the NPRO temperature with the lock-in.
I closed the shutter on the NPRO laser at the end of the night.
I'll log more details on the data tomorrow morning.
The top plot shows a sweep from 10 kHz to 5 MHz of the ratio of the voltage output of the PD detecting power from the NPRO laser beam and the RF source voltage (the magnitude of the complex transfer function). The black trace was taken with the laser beam blocked. For runs 2 and 3 I changed the laser temperature set point by 10 mK and 100 mK respectively to see if there was a significant change in the AM response. The bottom plots shows runs 2 and 3 compared to run 1 plotted in dB (to be explicit, i'm plotting 10 times the base 10 log of the magnitude of the ratio of two complex transfer functions). Changing the temperature seems to have only a minor effect on the output except at around 450kHz, where the response has a large peak in run 1 and much smaller peaks in runs 2 and 3.
The traces in the top plot consist of 16 averages taken with a 300Hz IF bandwidth, 15 dBm source power (attenuated with a 6 dB attenuator) and with 20dB attenuation of the input power from the PD.
Next I'm going to probe a narrow band region where the response is low (2.0MHz or 2.4MHz perhaps) and choose a bandwidth for the dither frequency for the PDH locking.
I've finished using the network analyzer to characterize find a dither frequency for driving the PZT to use in my PDH locking. I found a region in which the amplitude response of the PZT is low: The dip is centered at 2.418 MHz. Changing the NPRO laser temperature by 100mK has no significant effect on the transfer function in that region. I will post plots tomorrow.
I'm finished with the network analyzer. It is unplugged, and the cart is still near the PSL table. (I'll roll it back tomorrow when it won't disturb interferometer locking).
I closed the shutter on the NPRO at the end of the night.
Tomorrow I plan to put together the fast locking setup. I'll drive the PZT at 2.418 MHz. More details to come tomorrow.
I ended up choosing a different dither frequency for driving the NPRO PZT: 230 kHz, because the phase modulation response in that region is higher according to other data taken on an NPRO laser (see this entry). At 230 there is a dip in the AM response of the PZT.
I am driving the PZT at 230 kHz and 13 dBm using a function generator. I am then monitoring the RF output of a PD that is detecting light reflected off the cavity. (The dither frequency was below the RF cutoff frequency of the PD, but it was appearing in the "DC output", so I am actually taking the "DC output" of the PD, which has my RF signal in it, blocking the real DC part of it with a DC block, and then mixing the signal with the 230kHz sine wave being sent to the PZT.
I am monitoring the mixer output on an oscilloscope, as well as the transmission through the cavity. I am sweeping the laser temperature using a lock in as a function generator sending out a sine wave at 0.2 V and 5 mHz. When there is a peak in the transmission, the error signal coming from the mixer passes through zero.
My next step is to find or build a low pass filter with a pole somewhere less than 100 kHz to cut out the unwanted higher frequency signal so that I have a demodulated error signal that I can use to lock the laser to the cavity.
DMass and I locked the NPRO laser (Model M126-1064-700, S/N 238) on the AP table to the reference cavity on the PSL table using the PDH locking setup shown in the block diagram below (the part with the blue background):
A Marconi IFR 2023A signal generator outputs a sine wave at 230 kHz and 13 dBm, which is split. One output of the splitter drives the laser PZT while the other is sent to a 7dBm mixer. Also sent to the mixer is the output of a photodiode that is detecting the reflected power from off the cavity. (A DC block is used so that only RF signal from the PD is sent to the mixer). The output of the mixer goes through an SR560 low-noise preamp, which is set to act as a low pass filter with a gain of 5 and a pole at 30 kHz. That error signal is then sent to the –B port of the LB1005 PDH servo, which has the following settings: PI corner at 10kHz, LF gain limit of 50 dB, and gain of 2.7 (1.74 corresponds to a decade, so the signal is multiplied by 35). The output signal from the LB1005 is added to the 230 kHz dither using another SR560 preamp, and the sum of the signals drive the PZT.
I am monitoring the transmission through the cavity on a digital oscilloscope (not shown in the diagram) and with a camera connected to a TV monitor. I sweep the NPRO laser temperature set point manually until the 0,0 mode of the carrier frequency resonates in the cavity and is visible on the monitor. Then I close the loop and turn on the integrator on the LB1005.
The laser locks to the cavity both when the error signal is sent into the A port and when it is sent into the –B port of the PDH servo. I determined that –B is the right sign by comparing the transmission through the cavity on the oscilloscope for both ways.
When using the A port, the transmission when it was locked swept from ~50 to ~200 mV (over ~10 second intervals) but had large high frequency fluctuations of around +/- 50 mV. Looking at the error signal on the oscilloscope as well, the RMS fluctuations of the error signal were at best ~40 mV peak to peak, which was at a gain of 2.9 on the LB1005.
Using the –B port yielded a transmission that swept from 50 to 250 mV but had smaller high frequency fluctuations of around +/- 20 mV. The error signal RMS was at best 10mV peak to peak, which was at a gain of 2.7. (Although over the course of 10 minutes the gain for which the error signal RMS was smallest would drift up or down by ~0.1).
The open loop error signal peak-to-peak voltage was 180 mV, which is more than an order of magnitude larger than the RMS error signal fluctuations when the loop is closed, indicating that it is staying in the range in which the response is linear.
In the above plot the transmission signal is offset by 0.1 V for clarity.
Below is the closed loop error signal. The inset plot shows the signal viewed over a 1.6 ms time period. You can see ~60 microsecond fluctuations in the signal (~17 kHz)
The system remained locked for ~45 minutes, and may have stayed locked for much longer, but I stopped it by opening the loop and turning off the function generator. Below is a picture of the transmitted light showing up on a monitor, the electronics I'm using, and a semi-ridiculous mess of wires.
I determined that it’s not dangerous to leave the system locked and leave for a while. The maximum voltage that the SR560 will output to the PZT is 10Vpp. This means that it will not drive the PZT at more than +/-5 V DC. At low modulation rates, the PZT can take a voltage on the order of 30 Vpp, according to the Lightwave Series 125-126 user’s manual, so the control signal will not push the PZT too hard such that it’s harmful to the laser.
Below are some plots from dataviewer of temperature-step data taken over the past 32 hours. (They show minute trends). I am looking at the thermal coupling from the can surrounding the reference cavity on the PSL table to the cavity itself, and trying to measure the cavity temperature response via the control signal sent to heat the NPRO laser, which is locked to the cavity.
I stepped the temperature set point from 35 to 36 deg. C for the can at 12:30am last night. Then I waited to see the cavity temperature change and the slow actuator (laser heater: TMP_OUTPUT) follow that change.
I was a bit worried about the oscillations that were occuring in the TMP_OUTPUT signal even long after this temperature step was made, but I figured that they were simply room-temperature changes propagating into the cavity, since they seemed to have a similar pattern to the room-temperature variations, and since it is clear that the out-of-loop temperature sensor on the can (RCTEMP) experiences variations, even when the in-loop sensors are recording no variation.
At 8:46pm tonight I stepped the temperature down 2 degrees to 34 deg. C. The step had a clear effect on TMP_OUTPUT. The voltage to the heater dropped and eventually railed at its lowest output. I'm worried that the loop is unstable, although I haven't ruled out other possibilities, such as that a 2 deg. C temperature step is too large for the loop. I will investigate further in the morning.
The lock was lost when I came in around noon today to check on it. The slow actuator was still railing.
1) I got lock back for a few minutes, by varying the laser temperature set point manually. TMP_OUTPUT was still reading -30000 cts (minimum allowed) and the transmission was not as high as it had been.
2) I toggled the second filter button off. The TMP_OUTPUT started rising up to ~2000 cts. I then toggled the second filter back on, and TMP_OUTPUT jumped the positive maximum number of counts allowed.
3) I lost the lock again. I turned off the digital output to the slow actuator.
4) I have so far failed at getting the lock back. My main problem is that when the BNC cable to the slow port is plugged in, even when I'm not sending anything to that port, it makes it so that changing the temperature set point manually has almost no effect on the transmission (it looks as though changing the setpoint is not actually changing the temperature, because the monitor shows the same higher order mode even when with +-degree temperature setpoint changes).
I am trying again to measure a temperature step response on the reference cavity on the PSL table.
I have been working to relock the NPRO to the cavity. I unblocked the laser beam, reassembled the setup described in my previous elog entry: 5202. I then did the following:
1) Monitored error signal (from LB1005 PDH servo), transmitted signal, and control signal sent to drive PZT on oscilloscope.
2) With loop open, swept through 0,0-mode resonance, saw a peak in the transmission, saw an accompanying error signal similar to the signal shown in 5202.
3) Tried to lock. Swept the gain on the LB1005 and could not find a gain that would make it lock. Tried changing the PI-corner freq. from 10 kHz to 30 kHz and back and still could not lock.
4) Noticed that the open loop error signal displayed on the scope was DC-offset from zero. Changed the offset to zero the error signal.
5) Tried to lock again and succeeded.
6) Noticed that upon closing the loop, the error signal became offset from zero again. Turning on the integrator on the LB1005 increased DC-offset.
7) Reduced the gain on the SR560 being used as a low pass filter from 5 to 1. Readjusted the open loop error signal offset on the LB1005.
8) Closed the loop and locked. Closing the loop then caused a much smaller DC change in the signal than I had seen with the larger gain (now around 3mV). RMS fluctuations in error signal are now 1 mV (well within the linear region of the error signal).
9) Noticed transmission has a strange distorted harmonic oscillation in it a 1MHz. (Modulation freq is 230kHz, so it's not that). Checked reflected signal and also saw a strange oscillation--in a sawtooth-like pattern.
I intend to
1) Post oscilloscope traces here showing transmitted and reflected signal when locked.
2) Look upstream to see if the sawtooth-like oscillation is in the laser beam before it enters the cavity:
3) At some point, try to close the slow digital loop, perhaps readjusting the gain.
4) Try to measure a temperature step response.
I decided to go forward and try to close the digital loop in spite of the unexplained oscillations in the transmission.
1) Plugged the 20dB attenuator into the slow port on the laser drive. This pushed the laser out of lock and, for some reason, made the laser temperature stop responding to sweeping the set point manually with the knob.
2) Plugged the output from the digital system into the slow port (with the attenuator still in place).
3) Displayed the beam seen by the camera on a monitor in the control room
4) Stepped the laser temperature using MEDM until finding the 0,1 mode. Locked to that mode.
5) Closed the digital loop (input to slow laser drive attenuated 20dB attenuator). Gain 0.010
6) Loop appeared stable for 30 minutes, then temperature began shooting off. I opened the loop, cleared history, reduced the gain to 0.008, and started it again. Loop appears stable after 15 minutes of watching. I'm going to leave it for a few hours, then come back to check on it and, if it's stable, step the can temperature.
After finishing my last elog entry, I monitored the digital loop's error signal (the control signal for the fast loop) and the output to the laser heater remotely, (from West Bridge), using dataviewer. The ref cav surrounding can temperature was set to 36 degrees C.
With the loop closed and a gain of 0.008, after seeing the output voltage to the laser heater (TMP_OUTPUT) remain fairly constant and the error signal (TMP_INMON) stay close to zero for ~3 hours, I tried to step the temperature. (This was at 2am last night). I was working remotely from West Bridge. To step the temperature I used the following command:
ezcawrite C1:PSL-FSS_RCPID_SETPOINT 35.5
Rather than change the can temperature to 35.5 C, it outputted:
It had set the setpoint to 0 degrees C, which was essentially turning the heater off. I tried resetting it back to 36 and had no luck. I tried changing the syntax slightly.: ezcawrite C1:PSL-FSS_RCPID_SETPOINT=36 and ezcawrite C1:PSL-FSS_RCPID_SETPOINT (36). No success.
I ran over to the 40m and changed the temperature back to 36 manually. The in-loop temp sensor had decreased to 31.5 degrees C before I was able to step the setpoint back up. The system seems to have recovered from this large impulse though, and the laser has remained locked.
(5 hours of minute-trend data)
From left to right:
Top: Out-of-loop can temp sensor; Voltage sent to heat can
Middle: signal sent to heat the laser (TMP_OUTPUT); room temp
Bottom: Error signal for slow loop (sampled control signal from fast loop); In-loop can temp sensor
At 9:30 this morning (7 and a half hours after accidentally setting the setpoint to zero), I came in to the 40m. TMP_OUTPUT was still decreasing but was slowing somewhat, so I decided to step the can temperature up half a decree to 36.5 C.
TMP_OUTPUT responded to the step, but it is also oscillating slowly with room-temperature changes, and these oscillations are on the same order as the step response. The oscillations look like the room-temp oscilations, but inverted. (TMP_OUTPUT reaches maxima when RMTEMP reaches minima). Oddly, there does not appear to be much of a time delay between the room temperature and TMP_OUTPUT signals. I would expect a time delay since there's a time constant for a room-temperature change to propagate into the cavity. Perhaps the laser itself is susceptible to room-temperature changes and those propagate into the laser cavity on a much faster time scale. I don't know the thermal coupling of ambient temperature changes into the laser.
(24-hours of second-trend data)
--If the system can handle it, do a larger temperature step (3 degrees, say), so that I can more clearly distinguish the oscillations with room temp from the step response.
--Insulate the cavity with foam (will in principle make the temperature over the can surrounding the ref cav more uniform and less affected by room temperature changes).
--Insulate the laser? Is this possible?
--Leave the system as is and, as a first approximation, fit the room-temp data to a sine wave and subtract it off somehow from my data to just see the step response.
--Don't bother with steps and just try to get the transfer function from out-of-loop temperature (RCTEMP, which is affected by temperature noise from the room) to TMP_OUTPUT via taking the Fourier transforms of both signals.
I'm flying out tomorrow morning, so I'll either need to figure out how to step the temperature set point of the can remotely, successfully, or I'll need someone to manually enter in the temperature steps for me in the control room.
We opened up the little Thorlabs battery operated PD to see what was inside. Rana took some pictures, and I drew a schematic (attached). It's just a diode, biased with a battery (albeit a crazy 22.5V battery).
Comment by KA: PD is Hamamatsu S1223-01 Si PIN diode.
What a crazy battery. The main point is that it looks like this can be used for reasonable purposes: uses a load resistor on the BNC connector and you can use some pre-amp (e.g. Busby box or SR560) to have a low noise PD readout. You can also use the SR560 in its A-B mode as an 'opamp'. Ground the A input and the use a pole at 1 Hz and make the Output go into the B input through some large series resistor. The BNC from the PD gets Teed into the B input as well.
Then this becomes a transimpedance circuit readout of the diode, using the current noise of the SR560 as the limit.
Not dead. It just had a HT fault. You can tell by reading the front panel. Cycling the power usually fixes this.
MOPA is back onliine. Rana found that the fuse in the AC power connector's fuse had blown. This was evident by smelling all of the inputs and outputs of the MOPA controller. The power cord we were using for this was only rated for 10A and therefore was a safety hazard. The fuse should be rated to blow before the power cord catches on fire. The power cord end was slightly melted. I don't know why it hadn't failed in the last 12 years, but I guess the MOPA was drawing a lot of extra current for the DTEC or something due to the high temperature of the head.
We got some new fuses from Todd @ Downs.
The ones we got however were fast-blow, and that's what we want The fuses are 10A, 250V. The fuses are ~.08 inches long, 0.2 inches in diameter.
The PMC local oscillator is going a little weird dying. We need to check out why the level is fluctuating so much.
Here's a 6 month plot, where you can see that the lower level keeps getting lower (y-axis is dBm):
This LHO entry from 2008 shows where we first discovered this effect. As Rick Savage and Paul Schwinberg later found out, the ERA-5SM+ amplifier slowly degrades over several years and was replaced for both of the eLIGO interferometers. We have spares in the Blue box and can replace this sometime during the day.
Our PMC LO is made by this obsolete crystal oscillator circuit: D000419. There are many versions of this floating around, but they all have the ERA-5 issue.
Notes to the fiber team:
I am aligning beam onto the RFPDs (I have finished all 4 REFL diodes, and AS55), in preparation for locking.
In doing so, I have noticed that the fiber lasers for the RFPD testing are always illuminating the photodiodes! This seems bad! Ack!
For now, I blocked the laser light coming from the fiber, did my alignment, then removed my blocks. The exception is REFL55, which I have left an aluminum beam dump, so that we can use REFL55 for PRM-ITMY locking, so I can align the POP diodes.
EDIT: I have also aligned POP QPD, and POP110/22. The fiber launcher for POP110 was not tight in its mount, so when I went to put a beam block in front of it and touched the mount, the whole thing spun a little bit. Now the fiber to POP110 is totally misaligned, and should be realigned.
What was done for the alignment:
1. Aligned the arms (ran ASS).
2. Aligned the beam to all the REFL and AS PDs.
3. Misaligned the ETMs and ITMX.
4. Locked PRM+ITMY using REFL11.
The following were modified to enable locking
(1) PRCL gain changed from +2.0 to -12.
(2) Power normalization matrix for PRCL changed from +10.0 to 0.
(3) FM3 in PRCL servo filter module was turned OFF.
5. POP PDs were aligned.
We had a look this morning at the status of the seismometer array, so that we can get it all put together. While we were looking at the Guralp at the Yend, we noticed that it was pointing the wrong way. The North-South nubbins were pointed East-West, so X and Y coming out of the seismometer were backward.
To fix the Yend's Guralp, we powered off the Guralp readout box, rotated the seismometer, re-leveled it, and then turned the power back on. Now X from the seismometer lines up with the X data channel, and similarly for Y.
The Yend Guralp has all of the cabling needed, and is installed on the granite slab. This seismometer doesn't need any more work for now. When we get around to it, we'll need to do some kind of thermal insulation, but other than that, it's good to go.
The Xend will also have a Guralp (Zach still has it in the Gyro lab for now). We have the long cable that should go from the readout box to the slab that we'll need to put into the cable tray. The short cable from the slab's plate to the seismometer is already in place. For this seismometer, we should just need to plop the instrument in place and lay the cable in the overhead cable trays. We should also remove the now obsolete STS-2 cable while we're doing that. So, the Xend seismometer station doesn't need too much work.
The corner station will need more work. Zach made for us the long cable, although he still has it in the Gyro lab, so when we get the seismometer and cable back, we'll need to lay that cable in the overhead trays. The short cable from the slab's plate to the seismometer does not exist yet. We want to make sure that we can feed the finished cable and connector through the hole in the slab, and then we'll solder it up out here on the EE bench. I think this is how Den was doing things. If not, we'll have to do the soldering in-situ, which we don't want. So, for the corner station, we need to make the short cable, lay the long cable, get the T-240 back from Zach and put it on the slab, re-install the readout box that Zach has, etc, etc. We should also make sure that the spaghetti pot fits on the slab, underneath the piece of metal that's sticking out over the slab. We think that it's the same amount of clearance that the Yend pot has, so it should be okay, but we'll check. The O-ring seems to be sitting on the MC2 chamber, so we should remember that.
Neither the Xend nor the corner station had the yellow dog clamps, so we'll have to figure out where Den / Steve have hidden them.
EDIT: We have checked, and the Guralp connector, which is larger than the Trillium connector, fits through the hole in the slab (with some disassembly), so we can solder together the short cable out here on the EE bench, and install it separately. Eeeeexxxxxcelllent.
The PMC transmission was around 0.78 all day, rather than the usual 0.83ish. Rana went out to the PSL table and fixed up the PMC alignment. This should not need to be done very often, so things to check before touching the alignment are FSS / PMC settings (digital stuff). Make sure that the PC RMS (on the FSS screen) is low (at least below 2, preferably below 1), and that the FSS Fast monitor is near 5ish (not near 0 or 10).
This is a capture of PMC REFL's camera after Rana was finished. If it doesn't look this good when you finish then you are not done. Never do PMC alignment without looking at the PMC REFL camera.
The attached trend shows 80 days of PMC REFL and TRANS. The bad alignment stuff started on Sep 21-24 time period. You know who you are.
I wonder what's drifting between the laser and the PMC? And why is it getting worse lately?
We need to work farther on checking out the end transmission QPD electronics situation.
In bullet-point form, we need to:
* Ensure that the Weiss QPD head modifications have been made on these diodes. (cf. Rai W's LLO elogs on QPDs)
* Ensure that the QPD biases are somewhere in the range of 10-15V, not the old 100V. (Because we only need HV to make the capacitance low for RF use. Low voltage means less power dissipation in the head)
* Ensure the Rana/Rob modifications have been propagated to the whitening boards, so that we have full dynamic range. (Steve is looking for the marked up paper schematics)
* Replace signal path resistors with low noise metal film resistors.
* Check QPDs / whitening boards for oscillation (with a scope probe), ensure that we chose an appropriate analog gain.
In thinking about the transimpedances that we want, we thought about the current that we expect. We should get about 100 mW of light transmitted through the ETMs when we have full IFO lock. There is a 50/50 BS to split the light between the QPD and the Thorlabs transmission diode, so we have about 50 mW incident on the QPDs, which is about 13 mW per quadrant. With a sensitivity of about 0.15 Amps/Watt for silicon, this means that we're expecting to see about 2 mA of current per quadrant once we have the IFO fully resonant. We want this to correspond to about 5V, which means we want a transimpedance gain of around 2.5 kOhm.
For the things that need checking, each quadrant has:
Photodiode ------ Gain Switch 1 ----- Gain Switch 2 ------ Gain Switch 3 ------ Variable Gain Amplifier ------- Whitening stage 1 (z @ 4 Hz, p @ 40 Hz) ------- Whitening stage 2 (z @ 4 Hz, p @ 40 Hz)
We want to check on the status of each of these switches, and whether they actually do what they say on the QPD Head screens. Q has checked out and fixed the bit outputs for the whitening stages, but the rest still needs to be checked out. Also note that the Switch 1, Switch 2 and Switch 3 are common to all 4 quadrants (i.e. enabling switch 1 on one quadrant enables it on all quadrants), but the variable gains and the whitening stages are individual for each quadrant.
[ Rana, Jenne]
We remeasured the Yend PDH box.
When we first started, the green couldn't hold lock to the arm - it kept flickering between modes. Changing the gain of the PDH box (from 7.5 to 6.0) helped.
We measured a calibration, from our injection point to our measurement point.
The concept was that we'd take the mixer output, and put that into an SR560, and put the swept sine injection into the other input port of the '560, and use A-B. So, for this calibration, we left A unplugged, and just had the RF out of the 4395 going to input B of the '560. The 600 Ohm output of the '560 went to the error point input on the PDH box (during normal operation the mixer output is connected directly to the error point input). The SR560 was set to gain of 1, no filtering. I don't recall if we were using high range or low noise, but we tried both and didn't really see a difference between them.
We had the 4395 take that calibration out, and then we measured the closed loop gain up to 1 MHz. (Same measurement setup as above, but we connected the mixer out to the input of the SR560 to close the loop, and made sure we were locked on a TEM00 green mode.) Rana used an ipython notebook to infer the open loop gain from our measurement. Our conclusion is that we don't have nearly enough gain margin in our loop. We found the PDH box gain knob at 7.5, and we turned it down to 6.0, but the loop is still pretty borderline. We used the high impedance active probe to measure the error point monitor, since we aren't sure that that point can drive a 50 Ohm load.
We also measured the error point spectra and the control point spectra. Unfortunately, the saved data from the analyzer (no matter what is on the screen) comes out in spectrum, not spectral density. So, we need to check our conversion, but right now to get from Watts power to Volts, we do sqrt(50 ohm * data). We then need to get to spectral density, and right now we're just dividing by the square root of the bandwith that is reported in the .par file. This last step is the one we want to especially check, by perhaps putting some known amount of noise (from an SR785?) into the 4395, and checking that our calibration math returns the expected noise spectrum.
What still needs to be done is to calibrate this into Hz/rtHz. To do this, we were thinking that we should look at the error point on a 'scope while the cavity is flashing.
Anyhow, here is the uncalibrated error point spectrum. Purple is a measurement up to 30kHz, with 30Hz bandwidth. Blue is a measurement up to 300kHz with 300Hz bandwidth. The gain peaking schmutz above 10kHz sucks, and we'd like to get rid of it. We also see the same peak at ~150kHz that Q saw earlier today. We were using the high impedance probe here too.
We have the data for the control point (all the data files are in /users/jenne/ALS/PDHloops/Yend_18Aug2014), but we haven't plotted it yet.
Things that need doing:
* (JCD) Think about this box's purpose in life. What kind of gain do we need? Do we need more / less than we're currently getting? NPRO freq noise is 1/f and is 10kHz/rtHz at 1Hz (this is from a plot of an iLIGO NPRO from Rana's thesis, but it's probably similar). Talk to Kiwamu; the noise budget in the paper seems to indicate that we had some kind of boost on or something. Also, if we need much more gain than we already have, we'll definitely need a different box, maybe the PDH2 box that they have over in WBridge.
* (EQ, priority 1) Measure and calibrate error point noise down to lower freq for both arms. What could we win by putting in a boost? If the residual noise is high, maybe the laser isn't good at following arm, so beatnote isn't good length info for the arm, and we can't succeed.
* (EQ, priority 2) Measure TF of PDH box, and a separate measurement of the Pomona box that is between the mixer and the error point - is that eating a bunch of phase? It's already an LC circuit which is good, but do we really want a 120kHz lowpass when our modulation frequency is roughly 200kHz? Ask ChrisW - he worked on one of these with Dmass.
* (EQ, priority 2ish) Measure TF of Xend PDH loop (unless you already have one, up to ~1MHz).
* (JCD) Make DCC tree leaf for PDH box #17. Take photos of box.
We tweaked the mirror on the AP table to go through the center of the lens in order to get a more circular beam, but it seemed ineffective. So we put an IR card in front of the lens and behind the lens to see if the beam was circular or ovacular, but could not tell. We also moved the camera to see, but still couldn't see a distinct circle or oval. So Mike and Q will do a beam scan tomorrow in both the X and Y directions to see if the beam is circular or not.
Jenne, Koji, Rana
After fixing up the Mode Cleaner a bit more (fiddling more with the MC_align sliders to get the alignment before locking, making sure that it is able to lock), we noticed that the MC Trans path could use some help. To align the MC, we put MC1 and MC3 back into the position where Rob left it on Thursday and then maximized the transmission with MC2. Then we went back and maximized with MC1/3 keeping in mind the Faraday. We got a good transmission and the X-arm had a transmission of 0.8 without us touching its alignment.
Upon looking at the AP table portion of the MC_trans path, we decided that it was all pretty bad. The light travels around the edge of the AP table, then out the corner of the table toward the PSL table. A periscope brings it down to the level of the PSL table, and then it travels through a few optics to the MC_trans QPD.
The light was clipping on the way through the periscope, and so the MC_trans QPD was totally unreliable as a method of fine-tuning the alignment of the Mode Cleaner. Ideally we'd like to be able to maximize MC_trans, and say that that's a good MC alignment, but that doesn't work when the beam is clipped.
1. The first turning mirror on the AP table after the beam comes out of the vacuum was changed from a 1" optic to a 2" optic, because the spot size is ~4-6mm. We were careful to avoid clipping the OMCT beam, by using a nifty U200 mount (C-shaped instead of ring-shaped).
2. We placed a lens with a RoC of 1m (focal length for 1064nm is ~2m), a 2" optic, between the first two mirrors, to help keep the beam small-ish when it gets to the periscope, to help avoid clipping.
3. Rana adjusted the angle of the upper periscope mirror, because even when the beam was centered on the steering mirror directly in front of the periscope and the spot was centered on the first periscope mirror, the beam wouldn't hit the bottom periscope mirror.
4. We noticed that the bottom periscope mirror was mounted much too low. It was mounted as if the optics after it were 3" high, which is true for all of the input optics on the PSL table. However, for the MC_trans stuff, all the optics are 4". We moved the periscope up one hole, which made it the correct height.
5. We removed the skinny beam tube which guided/protected the beam coming off the periscope after a steering mirror since it (a) wasn't necessary and (b) was clipping the beam. We cannot use such skinny tubes anymore Steve.
6. There was a lens just before the 2nd steering mirror on the PSL table portion, which we removed since we had placed the other lens earlier in the path. 2 lenses made the beam too skinny at the QPD.
7. After this 2nd steering mirror, there had been a pickoff, to send a bit of beam at a crazy angle over to the RFAM mon, which we removed. This results in a much brighter beam at the MC_trans QPD, and at the camera. The QPDs readouts are now a factor of ~3.5 higher than they used to be. These (especially the camera) could use some ND-filtering action.
8. The steering optic directly in front of the MC_trans QPD is a beamsplitter, and instead of dumping the light which doesn't go to the MC_trans QPD, we used this to go over to the RFAM mon (instead of the pickoff which we had removed).
9. Koji fixed up the optics directly in front of the RFAM mon, accomodating the new position of the input light (now at a much more reasonable angle, and about 15cm farther back from the PD). Note the beam dump which is preventing the cables from the FSS board from entering the beam path. This included removing an ND filter wheel, so the RFAM mon values will all be higher now. Koji also has the beam going to the PD going at a slight angle, so that the beam isn't directly reflected on itself, so that it can be dumped.
10. We aligned the beam onto the MC_trans QPD using the first steering mirror on the PSL table.
11. We also removed the giant wall of beam dumps separating the squeezing section of the table from the rest of the table.
Alberto will elog things about the RFAM mon, including different values of the PD output, etc.
Still on the to-do list:
A. Replace the second steering mirror on the AP table after the MC_trans light leaves the vacuum with a 2" optic, since the lens we placed isn't tight enough to make the spot small there yet. Us a U200A mount if possible, because they are really nice mounts.
B. Put an ND filter in front of the MC_trans camera, because the image is too bright.
C. Normalize the MC_trans QPD - the horz and vert are pretty much direct voltage readouts, with no normalization. They should be divided by the DC value. This lack of normalization results in higher sensitivity to input pointing.
D. Long term, next time someone wants to optimize the MC_trans path, move all the optics, including the MC_trans QPD and the camera closer to the periscope on the PSL table. There's no reason for the beam to be traveling nearly the full width of the PSL table when we're not manuvering around squeezing stuff.
E. Never, ever purchase these horrible U100 or U200 mounts with the full ring and the little plastic clips. They are the "AC28" version. Bad, bad, bad.
Image 1: The new setup of the AP table, Mc_trans portion.
Image 2: New setup of the MC_trans part of the PSL table.
Some pictures of "magnet inspection" from Picasa.
The coating of some magnets are chipped...
On Friday we modified the POP22 set up: now the PD output goes to a bias tee. The DC output goes to the ADC board, while the RF output goes to an amplifier (Mini-circuits ZFL-1000LN+), to a band pass filter at 21.4 MHz and then to the ADC