I brought a new AD602AR from 40m and replaced the U5A AD602 which from the previous post seemed like the culprit, but it wasn't.

I'll think of some new way of figuring out the point of the problem. It would be nice if someone can help me with this. All the history of the issue is on this thread starting at CTN:2451.

Edit Thu Oct 31 10:26:50 2019

Issue fixed. See CTN:2469.

It turns out I did not have a full understanding of the problem and it was not really a problem. The blanking (pin 4) on U5A AD602 doesn't shut down the channel, it just reduced the gain by 100. So if the gain in previous stage is large enough, the lock can still be acquired. And that's what was happening.

Ideally, we need to keep the AD602 on all the time and lock by scanning the offset with low gain. The loop will catch the lock (the exact same thing I thought was a problem) and once that has happened, we can just increase the loop gain the set value.

Presently, the gain behind the U5A AD602 is 101, which is kind of high. I just need to check if the above-mentioned locking method would work robustly without wrongly getting locked to any higher-order modes with the gain slider set to some threshold value within -10 dB to 30 dB. If that can't be done, I might have to reduce the gain in the first stage. For now, the cavities are locked and beatnote is traveling towards set point.

First step in FSS Diagnostics was to see RF output from the RFPDs in FSS when they are locked. I ran some extensive measurements to cover all the information about this signal. The RF out is sent to the FSS box through ZFDC-20-5-S+ 19.5 dB directional coupler. The coupled output's spectrum is measured at different frequency ranges using both AG4395A and SR785. The measurement configuration files are included with data for metadata of the measurement. The signal is also analyzed in time series with measurement upto 1 GSa/s with TDS 3034C and one measurement at 5 GSa/s with TDS 3052B. Both measurements were done manually setting minimum possible voltage resolution and using DC coupling with 50 Ohm impedance. All data is attached raw here for now. More interpretation and analysis to come soon.

Data

Trying gain sliding like 40m

Rana told me that in 40m, the PMCs are autolocked by reducing the gain of the loop and changing the ramp until the lock is acquired. Then the gain is increased back to operation point. I tried this method with our South PMC as the usual method being used of 'changing Blanking state' wasn't working anymore. However, even with the gain set all the way to -10 dB, the loop was not locking exactly at the center of the TEM00 mode. And was unable to skip higher-order modes. There is a header H1 which changes the input stage gain. Removing this header pin, reducing the input stage gain by a factor of 100. Even after doing this, I was unable to robustly acquire the lock by this method. Besides, this reduced gain was the case earlier (CTN/2427) and it was too low as the VCA U5A AD602 had to be kept at maximum 30dB gain. So I did not want to reduce this first stage gain.

New method: Switch off input while changing Ramp

Somewhat similar to our FSS loops, I find it much cleaner to just not close the loop until we have reached near the lock point. This could be done fairly easily with the existing code. I just had to change the loopStateEnable variable from Engage (which changes the Blanking pin on U5A AD602) to input switch (FP1TEST for South and FP2TEST for North). So now, when finding a lock point, the input is changed to terminated inputs and the loop is closed when lock point is found. This works very nicely, just like the FSS autolocks.

This has finally fixed any problems with PMC autolocks.

I raised the North TTFSS box by 6 inches to make way for working on South box and to reduce the congestion of connectors in front of the two boxes. I have also clamped the boxes to a fixed position now, so they can't move. This would ensure the cables are not hitting the end of the platform and face any severe strain.

The next step towards improving lab cable hygiene and layout is to replace all RF cables with RG-405 Belden-N 1671J cables. However, the effects of this change would be less significant then fixing the sick FSS. So I'll first focus on that.

South Common Gain: 24 dB! , Fast Gain: 14 dB

North Common Gain: 10 dB, Fast Gain: 10 dB

Measurement method

• All measurements were taken withAG4395A for high frequency and SR785 for low frequency and the corresponding measurement parameter files are attached.
• For SR785, the SRmeasureWideSP.py script was used.
• HO 41800A Active Probes were used with AG4395A.
• For SR785, I simply used clips as it already has 1 MOhm input impedance. However, I have noticed oscillations in the spectrum.
• Note, South FSS box had higher gains, Common Gain: 24 dB, Fast Gain: 14 dB, while the Northside had Common Gain: 10 dB and Fast Gain 10 dB.
• TP1,5 and 4 are on common path on D040105 (North, South).
• TP17,14,15,16,18 and 19 are on PZT Path on D040105 (North, South).
• TP 11, 12 and 13 are on EOM Path on D040105 (North, South).

Data

I've borrowed a Marconi VCO, a mini-circuits LPF and mixer from the CTN lab for use at the WB EEshop.

AG Mon Dec 30 10:24:28 2019 : Has been returned to CTN now.

Method

I've wrote this script, nonlinTF.py which controls a Marconi 2023A and SR785 together. Marconi is used to providing a carrier frequency which is mixed with the Source Out signal from SR785 before feeding into the TEST2 input port on D040105 of TTFSS boxes. Then OUT1 port on D040105 of TTFSS box is used to read back at channel two of SR785 (channel one being fed with a copy of the Source Out signal). So SR785 is effectively measuring any downconversion in the loop (due to some nonlinearity) from micing of CF-IF, CF and CF+IF probe signals injected into the loop. The effectively closed-loop transfer function between TEST2 and OUT1 should be G/(1+G), so this injected signal should not suffer any suppression, nor should it affect the locks. The locks were maintained without any problems during the whole measurement. The CF frequency was stepped by 100 kHz from 100kHz to 10 MHz and then by 1 Mhz upto 100 MHz.

Mixer ZX05-1LHW (level +13 dBm) was used for the mixing and IF peak voltage was set to 30 mV. The configuration of the measurement for the transfer function is present in the configuration file in the folder.

Results?

• I had to upload plots in png as they were aggregated result of a lot of data, pdf would have been very heavy.
• North Side looks good with only minor bumps in 60 Hz harmonics. I'm not even sure if this just came because of SR785.
• The South side however looks very dramatic. A lot of stuff happening when carrier frequency was near 50 MHz.
• But I'm not sure if this measurement even makes any sense. South side is better in performance then North side, this wasn't expected.
• Also, the measurements are long-time measurements, so a lot of things could have changed between North and South (they were taken on different days too).
• I'll take more verifying measurements near the suspected frequencies tomorrow.
• I'm also thinking of taking spectrums instead of swept sine transfer functions next time.

Data

Method

• RFPD of the FSS loops were tested through there Test input ports (Updated schematics: South SN010, North SN009)
• I first took transfer function from the Test input port to the RF out port. Ideally, the transfer function of RFPD should be 100kOhm times this transfer function.
• Moku is used to generate two RF signals with a different frequency equivalent to RFPD's resonant frequency.
• These two outputs are combined with ZFSC-2-1W-S+ two-way splitter/combiner.
• I switched around the voltage amplitude of these frequencies and took a spectrum from AG4395A near difference frequency.
• The configuration file for spectrum measurement is present with data.
• This measurement was made through twoToneTest script present in the data folder.

Inferences:

• In first plot of each measurement set, I simply plotted the measured spectrum at the different signal voltage levels.
• In second plots, I have done some back calculations, but I'm not sure if this is a good of estimating non-linearity.
• I have used following formula to calculate ratio of total power in the difference frequency to the GM of powers of probe signals.
                                                               
• This ratio I have plot in dB in the second plot.
• It seems that the spectrum level generated is not dependent on how strong the probe signals are.
• This can be seen as along the diagonal, instead of remaining the same ratio, the ratio decreases by 20 dB in each step which is an increase in the denominator above.
• Lastly, I just connected the two ends of cable between which DUT goes and repeated the measurement.
• The non-linearity of the measurement setup itself is safely 50 dB below the measured values.

But what now?

• I need more time to make sense of these measurements. So that will come tomorrow.

Data

I calculated these values by integrating in the 8 MHz neighborhood around the marked harmonic peak, the power spectral density using the frequency at the point as the lower edge of the bin. Slew rate is calculated by multiplying the rms voltage level with the frequency and the fraction is calculated against the datasheet value for Max 4107.

59.221

2.1319

0.4264

9.248

0.6659

0.1332

1.135

0.1226

0.0245

0.39

0.0562

0.0112

0.621

0.1117

0.0223

23.205

0.8586

0.1717

3.248

0.2403

0.0481

0.495

0.055

0.011

0.301

0.0445

0.0089

0.512

0.0947

0.0189

These calculations at least show that MAX 4107 should be much far away from reaching its slew rate limit in both RFPDs.

In line with industrial practices, I did two tone third order intermodulation test today on the FSS RFPDs. This test was inspired by procedure described in this technical note by MiniCircuits and this paper at IEEE.

Method

• Similar to previous measurement, two RF tones are generated using Moku.
• However, now, they are separated by an audio frequency, 10 kHz in our case. This is marked as DF in the plots.
• These tones are centered around the resonant frequency of the RFPD marked as CF in the plots.
• The two signals are combined together through ZFSC-2-1W+.pdf, but this time all ports of the combiner have an attenuator on it.
• 3 dB attenuators at the input port and 6 dB at the output port. This was mentione din MiniCircuits as a step to improve impedance matching near the combiner.
• Ideally, two low pass filters also should be put at the input port, but I couldn't find any with high enough cut-off frequency.
• The combined source signal was getting attenuated by 8.75 dB. So I just drove the input signals higher by this amount.
• The Source signal is fed to the Test IN of the RFPD and RF out is read through AG4395A.
• The power in each harmonic is calculated from the spectrum and plotted together against source power.
• Since all RF signals are the near-resonant peak of the photodiode, there would be little to no difference in TF seen by them.

Inference:

• I think because of the 100 kOhm resistor at the Test In port of RFPDs makes the source signal too feeble.
• I'm just measuring instrument noise upto source power of -10 dB. I can't go above 0 dB in current setup atleast.
• At 0 dB source power, which is equivalent to 3 uApk photocurrent from the photodiode, the third-order harmonics are 60 dB lower.
• One way of reporting this is, Spurious Free Dynamic Range (SFDR) at 36(37) MHz is 60 dBc at 0 dBm source power of SN009 (SN010).
• Another way this figure is reported in industry is by extrapolating at which point the third-order harmonic power becomes the same as fundamental.
• But I do not have enough distortion showing data points to extrapolate to calculate the intercept.
• Last plot is when I connected the cables together which go to DUT. We see SFDR of 60 dBc here itself.
• The last two data points are when input range of AG4395A are very high and hence has increased noise floor.
• So probably in my measurements also, I'm just seeing distortion due to measurement apparatus itself.
• But this at least sets an upper bound of distortion to SFDR of 60 dBc at the resonant frequencies (within the errors that I, a graduate student, can make).

Datasheet for MAX4107

• The datasheeet says that between 30-40 MHz, the SFDR is approximately 50 dBc (this is actually lower than what we measured).
• Also, the third-order intercept is mentioned at 15 dBm source power.

Data

cool - how bout use this new no-how to estimate the excess in FSS?

I did some theoretical calculations using the datasheet value of second harmonic SFDR from MAX4107 and the transfer function I measured from Test IN ports of RFPDs (using 100 kOhm series resistance).

Calculation:

• Attached notebook has all the calculations. In all places, "Omega" refers to the modulation frequency.
• I first calculated power at 0, 1-Omega and 2-Omega from the resonance frequency in the light incident on the cavity using Bessel functions.
• I used reflection function from the cavity (using transmittance of 5 ppm, cavity length of 1.45" and assuming lossless cavity):
  
• Using this, I calculated reflected power at  0, 1-Omega and 2-Omega from the resonance frequency.
• Then I used -40 dBc SFDR for second harmonic generation mentioned in MAX4107 datasheet when used with a gain of 10. I used following formula to calculate this:
  
• Finally, I also calculated ratio of 1-Omega signal (actual beat between Sideband and Carrier) and the 1-Omega Distortion signal (generated) due to the non-linearity of MAX4107.

Inferences:

• These are purely theoretical calculations but I can put in the second harmonic SFDR by measurement tomorrow.
• The suppression of 2nd harmonics by the notch in the RFPD circuits is very bad. It only suppresses the second harmonic by 25-30 dB.
• This is not enough as this 2-Omega signal would mix a lot with 1-Omega and would also go forward into the TTFSS box after downconversion.
• The ratio of Signal at 1-Omega to the distortion at 1-Omega is about just 8 at 100 Hz. This doesn't reach 100 even till 1 kHz.
• And below 10 Hz, the distortion dominates the 1-Omega signal from RF out.
• This looks like a big limitation to our FSS loop. The notch filters in the RFPDs need to be doing much better job then they are doing.

Edited Wed Nov 20 14:43:07 2019: Corrected an error in code.

I took time-series data at TP1 on NFSS. This is just after the elliptical filter which is after the demodulation on board.

As also seen in the spectrum measurement at this testpoint in CTN:2474, there is a lot of power at around 435 kHz.  But this is not noise!

As seen on the oscilloscope, this is a near-perfect sinusoid. So this must be either of the following:

• A mixed down frequency due to non-linearity of MAX4107 or the Mixer JMS1-H .
• This could be an oscillation in the closed-loop of FSS at the unity gain frequency. But last I checked (CTN:2443), the UGF for the same COM Gain (10dB) and FAST Gain (10 dB) was 369 kHz.

This measurement was taken with a 500 MHz 10x Probe with a 300MHz TDS 3034C oscilloscope at 0.5 GSa/s sampling rate.

Interestingly, there is no such oscillation or peak on the South side. However, the south sides COM Gain is 24 dB and Fast Gain is 14 dB. So it could be because it is just suppressing this non-linear effect properly or just has a very high UGF.

Data

I quickly took a high-frequency Open Loop Gain measurement of NFSS loop at 10 dB COM Gain and 10 dB FAST gain, using the same measurement method as in CTN:2443. The UGF has not changed much but there is a dip at 435 kHz. This was there before too, I was just not paying enough attention to this part of OLTF before. So, we can say with some confidence that the 435 kHz signal seen in the oscilloscope in CTN:2482 at TP1 is actually due to some non-linear effect most probably and does not get suppressed at all. The phase margin near UGF looks about 135 degrees so there is no solid reason to believe this could be due to loop oscillation.

So I got to think of what combination of RF frequencies might be mixing down to create this oscillation and where. This oscillation is also visible in Plot 6 and 7 of NFSS_RFPD_Output_Oscilloscope.pdf of the measurements done in CTN:2470.

Data

I doubt it.

When feedback loops be noisy and busted,

the swept sine TFs cannot be trusted

The new circuit design has been built and is shown below. I tested the transfer functions and the PZT path looks good by eye but the EOM path looks like there is a solder point that is bad. I will show the graphs when I have them done.

This has happened few times now that acromag channel for the can heater driver stopped updating according to the PID script and the can gets heated to a very high temperature. This pushes the temperature out of the ranges of the current AD590 temperature sensor board. I have changed the range of channel 2 (this was being used for out-of-loop) to ensure we can still see some meaningful temperature value when such incidents happen. I have replaced R18 from 100k to 27k. The updated table is:

Weird phenomenon?

• I'm not sure this problem occurs though. Right now the out-of-loop temperature sensor shows that the temperature of the can is 55 Degree Celsius.
• This is also not cooling down fast. Last time it took days for it to cool down to set value.
• The alignment of the cavities change vertically when the can temperature is significantly different from the set value of 34.38 Degree Celcius.
• This hinders them from locking properly unless I tune the alignment back. But when the can will cool down finally, they will be misaligned again.
• Also, I just refuse to believe it is actually that hot, but I check the voltage after the transimpedance amplifier in temperature sensor boards and two independent AD590s are reporting this.
• And anyways, why is the acromag output channel getting frozen anyways.
• That being said, this is an irreproducible but non-harmful problem yet, so lower priority than the FSS saga going on.

I took a spectrum of PMC error signal when the FSS loop is not closed. This should provide a rough estimate of the free running laser noise. We had earlier seen a peak at 435 kHz in the Northside, hence I wanted to take this data with some references. First of all, this peak is very similar in the description of relaxation-oscillation peaks of these NPRO lasers mentioned on page 52 of this manual. The "Noise Eater (NE)" is supposed to suppress this peak significantly. However, in the spectrum of the PMC error signal, there is no difference when noise eater was ON or OFF.

I took a spectrum of Southside as well, just to see if I could see action of Noise eater there. For south laser, the noise eater suppressed noise only till 100 kHz or so and probably this side also has a similar relaxation-oscillation peak problem but is shadowed by a large feature at 30 kHz. Not, the absolute value of the spectrum between North and south are vastly different due to different amount of light, different transimpedances od the PDs and different gain values in the feedback circuit.

However, the noise eater is supposed to reduce relative intensity noise only. And the error signals of PMC should really be telling us noise in the frequency of the laser. So maybe I'm connecting two dots in different Hilbert spaces. But Rana suggested that a busted Noise Eater could be the reason for the 435 kHz peak, I just do not understand how RIN would cause frequency noise so badly. I thought photothermal transfer functions from RIN to frequency noise were very small.

when the NPRO crystal is oscillatin

          no tellin what is relaxin

I tested the circuit and got the attached results. The PZT transfer function looks reasonable. some adjustments to the values of the resistors in the circuit should have the match up. The phase is still out of alignment for both. The transfer function of the EOM looks far from the model. I suspect that this is due to a bad solder point or poor choice of components rather than the circuit's design.

I'm trying to think hard with my small brain how the distortion would affect the PDH functioning and inject noise in the frequency of the laser. I have a line of reasoning which starts with a question.:

• The two RF sidebands of laser that fall on the cavity, upon reflection, do they destructively interfere? Because the calculation I did in CTN:2481 suggests that the power at second harmonic (2-Omega frequency) is so high because of the sidebands beating with each other.
• In case they do not destructively interfere, the generated second-order harmonic (2-Omega) will have laser amplitude noise on it. This when it mixes with the (1-Omega) signal which actually carries the cavity length noise at the inputs of MAX4107, the distorted 1-Omega created would have laser amplitude noise on it.
• This might be the reason why FSS loops get overwhelmed with seeing a lot of laser amplitude noise which ideally it is not supposed to see. And it tries to correct this noise in phase quadrature making the situation worse.
• Since Andrew left, I increased the laser power reaching the cavities to go above the shot-noise limit of the photodiodes. Maybe, this increased light level increased this effect to a point where we are witnessing the problem.

Of course, all this depends on the RF sidebands interfering constructively upon reflection. I remember (I don't know from where) that it is the opposite. Either there is a fault in my calculations or this is indeed what is happening. I need to understand this properly to go further. Need help.

First, off I forgot a resistor on the previous diagram so here is the fixed version. Also, Anchal said that the EOM noise is probably caused by instrument noise at the low end. The phase response should be flat on both and on the PZT path the non-flat response is caused by some high pass filter caused by AC coupling or something of the sort.

I am still trying to figure out how to get a flat response on the phase. Anchal did some math that looks at the possibility of it and I am going through it.

I put laser settings on both North and South Cavities back to default. From this point onwards, all settings about the lasers would be known and kept track of. The red values are the settings that were changed.

NPRO Laser Settings

Property	Display Symbol	North	South	Units	Notes
Laser Model	-	M126N-1064-700, SN 5519, Dec 2006	126N-1064-500, SN 280, Nov 1997	-
Diode Temperature	DT	22.3	28.7		Informational only.
Diode TEC Voltage	DTEC	0.8	0.7	V	Informational only. +ve -> cooling, -ve -> heating.
Measured Laser Crystal Temperature	LT	40.8	55.2		Informational only. Calibration dependent.
Laser TEC Voltage	LTEC	0.0	-0.5	V	Informational only. +ve -> cooling, -ve -> heating. Manual says typically should be 0.0V.
Target Laser Crystal Temperature	T	40.087 -> 40.0010	48.0010		Changed back to factory set value on North Side.
Laser Head Power Level	PWR	66  ->  624	92  ->  101	mW	Calibration dependent. CHanged the calibration to meet the power meter but even then, power meter says a maximum 500 mW, so North Side is not entirely correct. On the south side, it was difficult to mount power meter perpendicular to the beam, so there might be some clipping loss in calibration.
Power Adjustment	ADJ	0	-2  ->  0	-	From -50 (off) to +10. Changed the diode current around set value.
Diode Current	DC	2.06	2.04	A	It can be changed to change power level. Reflects measured value.
Diode Power Monitor	DPM	0.00	0.00	V	Calibration Dependent.
Noise Easter	NE	ON	ON	-	-
Laser Diode Status	LD	ON	ON	-
Nominal Diode Current	-	All the way clockwise	All the way clockwise	-	It can be changed by turning the left potentiometer from the back of the laser head. Factory default is all the way clockwise. I have set both North and South Lasers to this point.

Note:

While turning the nominal diode current of south laser all the way clockwise, I found that the laser power peaks before the maximum diode current is reached. This diode current is about 1.9 A. This is unexpected. Any explanations on this would be helpful.

Testing the circuit using SRmeasure and AGmeasure, I got more accurate readings of the transfer functions. The PZT path TF looks good. The EOM path looks good up until 10⁵ Hz, then it doesn't drop like it is supposed to. My next task is to fix the high frequency EOM path. These graphs are a work in progress. I will include plots with the ideal TF at a later time.

please overlay all the plots and also include the ideal TFs - its hard to understand anything without the comparisons

I took beatnote spectrums in the current modulation index (was set to around 0.3 earlier). Then I took spectrum after attenuating the modulation signal power to both EOMs on North and South Path by 3 dB and 6 dB. This should reduce the modulation depth by  and 1/2. After every change, a small displacement happens in the thermal control of the cavities, so I had to wait for some time to let it settle. The gains of the FSS loops were kept constant to make sure only the modulation depth is the parameter that is changing. Gain values were 24 dB and 16 dB for South COM and FAST gains, and 11 dB and 10 dB for North COM and FAST gain.

Inference:

• There is almost no difference in noise upto1 kHz.
• There is a new bump I'm seeing between 100 and 200 Hz though which was there on Dec 12th but not on Dec 13th measurements.
• This bump might have been reduced due to a decrease in laser power or something might have changed between the two days. I'll check with increasing power what happens.
• Above 1 kHz, we see that as modulation depth is deceased, we see a reduction in noise. The same goes for power.
• Overall, this does show that the effect of these changes is more or less as expected and the effect of non-linearity, if any is not much.
• It is still valuable to quantify properly how much the non-linearity might be affecting the final beatnote spectrum though. Working on it.

Data

I found the latest available photos for the Cavity Reflection RFPD circuits.

There were minor changes made after taking these photos which are logged at: https://nodus.ligo.caltech.edu:30889/ATFWiki/doku.php?id=main:experiments:psl:rfpd

I'll update these above photos whenever the next time I get a chance. They would be present at:

SN009: https://git.ligo.org/cit-ctnlab/ctn_electronics/blob/master/RFPD/Photos/SN009.jpg

SN010: https://nodus.ligo.caltech.edu:30889/ATFWiki/doku.php?id=main:experiments:psl:rfpd

I rechecked the 20dB coupled RF output of SN009 (RFPD on north Cavity reflection) and SN010 (RFPD on south Cavity refleciton). The following are the mean values over 10s taken at DC coupling with 50 Ohm input impedance, +/- 40 mV input range with two different oscilloscopes.

Note that these are 20dB coupled values, so the actual offset reaching the FSS board is about 70 mV. The PD input is AC coupled through a transformer so it shouldn't be reaching further than there but ideally MAX4107 at the RFPD is supposed to have a maximum input offset voltage of 3 mV which at gain 10  should look like 30 mV. So what we are seeing (7.0 mV) is more than twice the rated maximum input offset. I'm not sure if this means our MAX4017 is busted or something is wrong in the loop. Help needed in understanding this result.

In talks with Anchal, we have adapted our model and the transfer function fits rather well. These measurements were taken with the AGmeasure and SRmeasure script on the larger instruments in the CTN lab (I can't remember the exact make of them). Two were used for different frequency bands: one up to 100 kHz and the other one above. In order to have continuous measurements, the high-frequency measurement was shifted to be continuous with the lower frequency measurement. This was not a significant shift and does not impact the overall interpretation. It is merely a correction for different calibrations in the Instruments used to measure the transfer functions.

The Magnitude of the transfer functions matches well and Anchal agrees that they also look good. So the circuit can be used in its current form. It isn't currently in a box but that is an easy fix. I am still making full documentation on it and that should be done over break. I am also looking to make it as a single PCB to eliminate any points of failure that come from attaching multiple PCBs together. It is not entirely necessary because as is seen in the graphs, the circuit works well but designing a PCB is a good task for over break because it does not really require any in lab work and will be highly beneficial down the road. I am still talking to Anchal about the circuit and its implementation.

As suggested few group meetings ago, I took time series data at the input and output of boost stage opamp U7 (at TP15 and TP16) using TDS 3034C. I know trusting oscilloscope for synchronous measurement of two channels is not a great idea, but this is a good zeroth order approximation for this approach. 500 MHz 10MOhm impedance probes were used for the measurement of the signals. The data is taken at two different acquisition rates and I used basic math to calculate the expected ideal output. Following formula was used:

Ofcourse the offset point above wouldn't work. Also, I have not included the notch filter that is also present at this stage.

The results of this analysis are attached. I also, calculated the required slew rate of the AD847 at this position. At the 100 MSa/s sampling rate, we saw that maximum required slew rate for ideal signal was around 110 V/us, much less than rated 300 V/us. I don't see a bad change in shape of the waveform. At 1 GSa/s sampling rate, we see that  required slew rate reached about 220 V/us at indeed oscillations at high frequencies are limited even though this is below the rated value.

Looking from data taken at CTN:2474 present in this directory, these fast oscillations that we are seeing on the oscilloscope are around the modulation frequency 36 MHz.  These leftover downconverted 2-Omega from teh RFPD might be saturating slew rate limtis of many opamps in the TTFSS box. I think I should look back into Andrew's implimentation of the elliptical filter right after the mixer in the RF board. He did say that we need to make the filter lossy to ensure the reflected signal at 36 MHz gets absorbed at a 50 Ohm resistor. For this we need to add a resistor to gnd at node rfbn2. Seeking permission to make modifications to board.

Data

If we are going to keep the same laser with busted Noise Eater, should be go in the direction of implementing the ISS?

In Oct, 2018, I along with Johannes developed ISS boards and photodiode transimpedance amplifier boards. These were characterized by Johannes at Cryo_Lab:2180 and Cryo_Lab:2181. However, Johannes was using AOM while we have EOAM here. There are slight differences which should be addressable given we have an offset port also in the ISS board. But the expected open loop gain needs to be worked out along with right choice of transimpedance and cavity pole neutralizer stage in ISS.

Johannes gave me his M2ISS photodiode mounts before leaving, so we have them too. This looks like a 1-week project from start to finish. So with the thumb rule of "nothing goes as expected" and me being a grad student, this should take 2 weeks. Is it worth it? Should I start working on this? Or can I just get a laser with working NE?

Also, there is no real documentation of why we stopped using the existing ISS in CTN? Last mention of ISS is at CTN:2132 and after that I came and was working on ISS boards which we never installed in CTN since or FSS, RFPDs and thermal controls were priority then.

After the last calibration, South Laser power at Laser head determined by its internal photodiode and show as setting PWR today (noticed today) dropped to 80 mW with none of the other settings changed. The diode current is still the same, so this could be one of the following two things:
1) The internal circuit to measure laser power at head went wrong.

2) Something is wrong with the laser crystal.

The laser power as seen by the reflection photodiode at South PMC captured a glitch on Dec 17th, 2019 8:30 am (might be 7:30 am in PST). The glitch shows that PMC went out of lock due to it but then returned to a laser power lower than before. The voltage level decreased from 2.72 V to 2.3 V after lock, which is equivalent to a drop of ~15.4%. The laser head power level according to PWR monitor dropped from 101 mW to 80 mW, which is equivalent to ~20.8%. While these don't match, PMC reflection is also not a true measure of power level. But I am sure this power has actually reduced and even if the laser head power meter is off a little bit, we have witnessed a big drop in power.

Restarting the laser (Put it on Standby and switch it back on) didn't change the power level.

Restarting the laser with turning the key off and on also didn't change the power level.

Even though 80 mW is enough for our experiment, this sudden decline in power shows there is something happening with the laser that we do not understand.

Data

Today I inserted two photodiodes, Thorlabs PDA10CS at (110,44) at dumped end of a PBS before PMC and at (64,28) at dumbed end of INput PBS of Faraday Isolator after PMC. These photodiodes were set at 20dB gain which according to the manufacturer gives us a bandwidth of 1.9 MHz.

Then, I just took Noise Spectrums using AG4395A with the parameter file attached in the data directory, from 0 to 1 MHz at 1 kHz ResBW and 50 averages. The output was divided by the DC level of the photodiode which was measured using TDS 3052B averaged over 10s. This measurement was done with FSS ON or OFF.

Inference

• Whenever FSS is ON, the loop is completed, so noise due to PMC and FSS is present in the measured RIN.
• But when FSS is OFF, before PMC, the noise should be just of free running laser even though PMC is locked downstream.
• The most important observation is that when FSS is OFF, before PMC, the RIN is pretty small with no peaks in 100s kHz, particularly the notorious 433 kHz peak.
• The dome from 1kHz to almost 100kHz peaks at 10kHz is most probably due to FSS as it goes away whenever FSS is OFF.
• The peak at 433 kHz persists even after switching off FSS and is present only after PMC. So this peak is clearly due to PMC and not due to laser as was thought before.
• Conclusion is, that laser is free of any large RIN and most of the RIN introduces is due to FSS or PMC. 
• I've left the PDs in there with termination. I'll change some PMC loop and FSS loop parameters tomorrow to see nature of their contribution to RIN.

North Laser is hereby acquitted.

Data

Could be that either

1. the FSS makes RIN through driving the PZT / making beam jitter
2. FSS drives EOM and makes polarization rotation if EOM is misaligned or beam is not linearly pol and aligned to EOM axis

I suggest measuring the RIN/PZT and RIN/EOM TFs, both before and after the PMC.

Is the PMC servo making the 430 kHz peak?

I took some TF measurements, but I'm not sure if I used the right method to do this. All the resutls look essentially the same, so maybe I'm just measuring instrument noise and nothing else. Regardless, I' posting the results. 

Measurement details:

For both measurements, I sent the source signal, first directly to PZT or EOM and second with a 50 Ohm termination in parallel. In case of EOM, I also did another measurement with low power to see if I can uncover any saturation effects my high powered source might be causing.

The output of two photodiodes, Thorlabs PDA10CS at (110,44) at dumped end of a PBS before PMC and at (64,28) at dumbed end of INput PBS of Faraday Isolator after PMC was measured and I took DC level of the outputs averaged over 10 s right after taking TF measurement. This was divided by the measured TF to get units of 1/V i.e. from PZT/EOM to RIN.

The parameter configuration files for the measurements are in the data directory.

What might have gone wrong:

I suspect that maybe AG4395A is unable to drive the capacitive load of PZT and EOM after a certain frequency. I need to find a better way to actuate the PZT and EOM in a known fashion, possibly through the FSS box or some buffer driver.

In the case of EOM, I'm also uncertain if the amplitude of actuation would be enough to do anything whatsover. Maybe the transfer function needs to be taken with high voltage driver.

Any comments on my measurement techniques are welcome as I'm surely not doing this right.

Rana's question about PMC:

I concluded that PMC is causing the 433 kHz peak because the only time I do not see it in CTN/:2502 is behind PMC when FSS is OFF. I couldn't think of a way to check if PMC Servo is causing it on its own. Can I do that without closing the loop somehow?

Data

TF Measurement Method

• I switched on Test1 switch on RF Board to cut off RFPD input to FSS.
• I switched on Test2 switch to connect the excitation port at the summing point.
• I switched on Ramp Engage for FSS to ensure paths to PZT and EOM are connected.
• Then, I disconnected and shorted EOM input when taking the measurement for PZT and vice versa.
• COM Gain was 6dB and FAST gain was 10 dB.
• TF was measured from Test2 EXC Port to the output of PDs. This was divided by 10s averaged DC value of PD output to get TF to RIN.
• Measurements were taken in turns for low and high frequencies with IFBW 10Hz and 1 kHz.
• Again, PDs are  Thorlabs PDA10CS placed at (110,44) at dumped end of a PBS before PMC and at (64,28) at dumped end of Input PBS of Faraday Isolator after PMC.

Calculations:

• The noise spectrum in FSS loop was measured at OUT2 port on the common board.
• This was divided by 50/(453+50) voltage division caused by 50 Ohm input impedance of AG4395A. This gives spectrum at the output of the U3 adder stage in the common board.
• The TF measured above are divided by 392/1200 which is the gain from Test2 EXC to output of U3. So the TFs are no referenced from the output of the U3.
• The measured noise spectrum is multiplied by the TF from that point to RIN at After and Before PMC due to PZT or EOM.
• These are summed quadratically to get total estimated RIN ASD (before and after PMC) due to the noise from FSS.
• Finally, I also plotted the RIN ASD measured in CTN:2502 before and after PMC when FSS was ON.

Inferences:

• The measured RIN, both after and before PMC are much higher than the estimated noise due to FSS.
• In particular, the bump around 10kHz in measured RIN before and after PMC is not present in transfer functions from FSS to RIN or the noise spectrum itself in FSS.
• The notorious 433 kHz peak is present everywhere, but there's order's of the magnitude of difference between FSS contribution and seen RIN. So it is mostly the other way round, the RIN causing these peaks in FSS.
• Either that, or there is something wrong with the way I took measurements or did the above calculations.
• Also, I'm aware that the RIN is actually fed back into the FSS through PD and it is not an open-loop. But I did this simpler analysis first. Maybe that's what is wrong.

Data

Edited on Mon Dec 23 15:21:50 2019 .

For the convenience of others, I'm summarizing the open questions I asked on elog in December. Comments on the posts, advice or answers to my questions would be nice.

• CTN:2492 Why are output powers of these lasers much lower than rated power when diode current is maximum? And why do I see a maxima in South Laser Power when diode current is increased, shouldn't it just increase with diode current monotonically?
• CTN:2495 Comments on beatnote noise differences with different laser power and modulation depth. Can I infer some good choice of laser power and modulation index from this? Suggestions for more tests on this line?
• CTN:2496 Rana asked for the latest available photos of FSS RFPDS. We want to add an active notch in the final RF amplifier stage.
• CTN:2497 There is indeed actual DC offsets at the output of FSS RFPDs RF out. Please cross-verify my measurement method and does this mean the MAX4107s are busted. Should I replace them?
• CTN:2499 Time series measurement of the signal before and after the summing stage in FSS on Northside. There is significant leftover 1-Omega frequency even after the elliptical filter. I think I need to make the input end lossy to absorb back reflections from the elliptical filter. Need permission to modify the circuit and test if this helps.
• CTN:2501 There is a logged event of South Laser losing about 20% of its output power with the same diode current. Is this laser dying? What is happening?
• CTN:2505 As instructed, I have taken transfer functions through FSS from PZT path and EOm path to RIN before and after the PMC. I need validation on method used. My conclusion is the noise in FSS is due to RIN and not the other way round (FSS causing RIN).  I tried taking an OLTF of the PMC loop but the NPMC Loop unlocks as soon as I connect the source port of AG4395A to an excitation port in a summing stage. So I've been unable to verify the source of this RIN yet, but previous measurement (CTN:2502) suggests it is PMC.

HAPPY HOLIDAYS

Method:

• I injected excitation at FP1Test and readout at Mixer Out (FP3Test).
• All TF were taken in three stages encompassing the whole frequency range. The same configuration files were used.
• First, I took TF measurement when the loop was closed and the variable gain is 6 dB.
• Then I opened the loop by removing mixed down PD signal from FP2Test and shorting it. This gave direct transfer function from FP1Test to Mixer Out as in second plot.
• Using the two measurements, I calculated the open-loop transfer function for the NPMC Servo Loop.

Inference:

• My hope was to see some activity or unity gain around 433 kHz where we see the spike of noise.
• But the unity gain is much before around 6 kHz with phase margin of about 90 degrees.
• There is a small dip around 433 kHz but again it seems like a reaction to the problem rather than the source of the problem.
• The measurement, however, is pretty trashy. I could have used SR785 for the lower frequency region to make it better.
• I think we are back to square one. Any comments to infer more or suggest further tests are welcome.

Data and Analysis

I repeated this measurement to compare later after changes to RFPDs.

Data

Circuit changes (Updated Schematic)

I made the following changes to SN009:

• R15 -> 10k Ohms for increasing Test Input Power.
• R1   ->  226 Ohms
• R2   ->  25 Ohms (Two 50 Ohm resistors stacked in parallel)
• Added Ln, 2.2 uH fixed inductor.
• Added Cn 0.5-1.5 pF Trim Tunable Capactior.
• Ln and Cn are in series such that Ln + Cn is parallel to R1.
• Rewired the 5V line from one end of C3 to one end of C14 (U8-14) with yellow 20 AWG single core wire.

Updated Photo

Measurements

• I used the setup in 40m to take light transfer function of the modified SN009.
• Method described in CTN:2247 was used.

Inference

• It turns out, however, that the placement of "active" notch didn't help at all. Infact, the ratio between peak and notch went from 30 to 20 :(
• I needed to reduce the values of R1 and R2 to ensure the notch doesn't effect the peak transimpedance much.
• During the tuning, I found that as I turned the screw on the capacitor, the transimpedance will just cycle through phases of low and high impedance. I dont' think I ever saw a very clear notch because of the new changes.
• I also tweaked the passive notch slightly to bring it closer to 72 MHz.
• I'll think more about what has happened and will try another pair of inductor and capacitor.

Data

I made a more advanced model of the circuit using Zero and compared them to the experimental results.

I also designed a board for the circuit. This is based on the Zero model from the previous post.

(These designs have not been given a LIGO DCC number yet so I haven't added one to the design)

Circuit changes (Updated Schematic)

I made the following changes to SN009:

• R1   ->  100 Ohms
• R2   ->  10 Ohms
• Added Ln, 220 nH fixed inductor. (Coilcraft 1206CS-221X_L_)
• Added Cn 4.5-50 pF Trim Tunable Capactior (GKY50086).
• Ln and Cn are in series such that Ln + Cn is parallel to R1.

Updated Photo

Measurements

• I used the setup in 40m to take light transfer function of the modified SN009.
• Method described in CTN:2247 was used.

Inference

• The ratio of resonance peak at 1-Omega (36 MHz) to 2-Omega (72 MHz) has been imporved from about 30 to about 200!
• Overall gain has been reduced but we can just increased the laser power to compensate for that.
• The problem from last time got resolved by chosing a combination of L and C such that C is significantly more than any stray capacitances.

Data

Circuit changes (Updated Schematic)

I made the following changes to SN009:

• R15 -> 10k Ohms.
• R1   ->  100 Ohms
• R2   ->  10 Ohms
• Added Ln, 220 nH fixed inductor. (Coilcraft 1206CS-221X_L_)
• Added Cn 4.5-50 pF Trim Tunable Capactior (GKY50086).
• Ln and Cn are in series such that Ln + Cn is parallel to R1.

Updated Photo

Measurements

• I used the setup in 40m to take light transfer function of the modified SN010.
• Method described in CTN:2247 was used.

Inference

• The ratio of resonance peak at 1-Omega (37 MHz) to 2-Omega (74 MHz) has been improved from about 20 to about 120.
• This is less improvement than what we saw on the SN009 case.
• Also, the measurement through test port and using lasers is not matching up. This could mean that my estimation of DC transimpedance is wrong.
• In the last plot, I tried multiplying the measurements of the result by a factor of 3.56 and it makes the transimpedance measurements match with test port measurements.
• This shows there is a factoring error in my estimation of dc transimpedance or maybe the whole thing is damped down when light falls on the photodiode instead of test signal.
• Also, there is a significant oscillation peak at around 200 MHz. Maybe that is the reason for the damped overall transimpedance. Working on this next.

Data

Circuit changes (Updated Schematic)

I made the following changes to SN009:

• R1   ->  200 Ohms
• R2   ->  10 Ohms
• Ln   -> 470 nH r. (Coilcraft 1206CS-471X_L_)
• Replaced U1, MAX4107.

Updated Photo

Measurements

• I used the setup in 40m to take light transfer function of the modified SN010.
• Method described in CTN:2247 was used.

Inference

• The ratio of resonance peak at 1-Omega (37 MHz) to 2-Omega (74 MHz) has been improved from about 20 to about 230.
• Also, the measurement through test port and using lasers is still not matching up. This could mean that my estimation of DC transimpedance is wrong.
• In the last plot, I tried multiplying the measurements of the result by a factor of 3.56 and it makes the transimpedance measurements match with test port measurements.
• This shows there is a factoring error in my estimation of dc transimpedance or maybe the whole thing is damped down when light falls on the photodiode instead of test signal.
• But the oscillation peak problem has been resolved. So the photodiode is good to be used. Minor tweaks might be required to see why test port estimation is different from laser measurement.

Data

Parallel relevant threads:

CTN:2516 : Notch improved on FSS RFPDs. Gain values increased as well.

CTN:2517: TTFSS OLTFs with Maximum Gain

Repeated these measurements after changes to the RFPDs (Notch improved).

Inferences:

• No significant shape change is observed in the Open loop Transfer Functions (OLTFs).
• Only minor absolute value changes have happened which just indicate a change in overall transimpedance at omega frequencies of the RFPDs.
• This indicates that if non-linearity was affecting anything, it at least wasn't the shape of the OLTFs or the suppression of laser noise.
• Also, these measurements endorse that no major harm has been done to the circuits in the process of modification (as should be the case).

Data

New Gain Values:

South Common Gain: 24 dB , Fast Gain: 18 dB

North Common Gain: 14 dB, Fast Gain: 14 dB

I incrased the above values as much as I could without getting oscillations in the loop.

RFPD Changes:

• The North FSS RFPD SN009 (CTN:2512) and South FSS RFPD SN010 (CTN:2514) have been improved in terms of ratio of their peak transimpedance to their notch transimpedance by implementing an active notch in the output RF amplifier.
• This was supposed to reduce non-linear effects (if any) because less of the 2-Omega frequency would go through the amplifier.
• It seems like it didn't help at all as the beatnote noise spectrum is at the same place as before.
• While the gain values of FSS have increased, the transimpedances of the RPFDs decreased after the changes,

Latest BN Spectrum: CTN_Latest_BN_Spec.pdf

Daily BN Spectrum: CTN_Daily_BN_Spec.pdf

Parallel relevant threads:

CTN:2514 : FSS Diagnostics - SN010 (South RFPD) Notch Improved

CTN:2517: TTFSS OLTFs with Maximum Gain

Repeated these measurements with maximum possible gain values in the FSS loops.

Data

Parallel relevant threads:

CTN:2514 : FSS Diagnostics - SN010 (South RFPD) Notch Improved

CTN:2516 : Notch improved on FSS RFPDs. Gain values increased as well.