c1psl, c1susaux, and c1auxey today
MC autolocker got stuck (judging by wall StripTool traces, it has been this way for ~7 hours) because c1psl was unresponsive so I power cycled it. Now MC is locked.
I swapped the inputs to the ZHL-3A at the PSL table - so now the X beat RF signals from the beat mouth are going through what was previously the Y arm ALS electronics. From Attachment #1, you can see that the Y arm beat is now noisier than the X. The ~5kHz peak has also vanished.
So I will pursue this strategy of switching to try and isolate where the problem lies...
Somebody had forgotten to turn the HEPA variac on the PSL table down. It was set at 70. I set it at 20, and there is already a huge difference in the ALS spectra
Rendered the SOS assembly (D960001) with correct materials and all and it looks very nice. Will extend this to the building cad later.
[rana, kevin, udit, gautam]
quick notes of some discussions we had today:
RXA: 0805 size SMD thin film resistors have been ordered from Mouser, to be shipped on Monday. **note that these thin film resistors are black; i.e. it is NOT true that all black SMD resistors are thick film**
We use D990694 in various places. Today, Rana alerted me to an important consideration to be kept in mind when we use this board, which I found quite interesting. I still don't understand the problem at the BJT level, but I think one can appreciate the problem without going to the transistor design of the LT1125. I'm attaching an annotated schematic of the whitening section in question. If the following assumptions are valid, then I think my picture is valid.
Then, as one can see in the attached schematic, when we set the gain of any input to <24dB, we must ensure that the input voltage is less than approximately 2V. Otherwise, by asking too much of the first stage op-amp in the quad IC LT1125, we may be messign around with all the 4 op amps in the quad! Even the 0dB setting is not immune to this problem, as it uses one of the 4 op amps.
Now that I think about this a bit more - this problem shouldn't be significant for the usual LSC degrees of freedom when in lock, as the huge DC gain of the loop should squish large DC values of the error signals, and so there shouldn't be any danger of overloading the LT1125. But I don't know if we are being hurt by this effect when flashing through resonances, when the PDH horn-to-horn voltage can be quite high (which is in principle a good thing?). I don't know if there is any "hysterisis" effect where the overloaded quad IC has some relaxation time before it returns to normal operation, and if we are being limited in our ability to catch lock because if this effect.
The concerns remain valid for th ALS demodulated error signals though, for which the signals will remain large throughout.
this is the note from Hartmut Grote on this topic from 2004
After some research: -the- reason for the reduced +/- 20,000 swing in raw values is a default setting for having support for legacy devices enabled when using the acromag proprietary i2o peer-to-peer protocol. So this is doubly unnecessary because a) we don't have any legacy devices at all and b) we're using pure modbus/TCP and no i2o. To change the setting I have to connect via the USB configuration utility. In addition, I want to understand the averaging feature of the acromag units better, which is also configured via USB, and lets one set a fixed amount of samples to be averaged before updating the read-register value. The documentation says that the 8 channels are multiplexed into a single ADC, and that new input data is available after 10 ms for each channel, suggesting a sampling rate of 100 Hz per channel and that the multiplexing happens faster, but is not super-clear about this, so I want to test it in the cryo lab first before unleashing it onto c1auxex2.
Furthermore, the standard timing options for updating epics records are 10s, 5s, 2s, 1s, 0.5s, 0,2s, and 0.1s. On the previous c1auxex, the monitoring channels were set to 0.1s, but that clashes with the 16 Hz global EPICS rate, resulting in partial double-sampling. One can manually provide the option 0.0625s for 16Hz update rate. I will test this and how it deals with the averaging in the cryolab too.
After discussing with Koji, we looked at the aLIGO incarnation of this board. Interestingly, it too has a similar topology of 4 switchable gain stages with gains of 24, 12, 6 and 3dB. The main differences are that they use single Op27 ICs instead of the quad LT1125s, and also, they use a different combination of feedback resistors to realize the various gains.
We considered upping the feedback resistance (R15, R143) on the 24dB gain stage of our boards from (1k, 66.5ohms) to (3k, 200ohms) as on the aLIGO boards - but this doesn't really help? Because KCL demands that the same current flow in R15 and R143, and so the output Vsat of the op amp and its max current driving capabilities in combination determine if the inverting input can follow the non inverting input?
As Hartmut points out in his note, he was able to access the full range of ADC voltages when the gain was set to 3dB, despite the fact that the LT1125 was still getting internally saturated. Operating with minimum 24dB whitening gain doesn't really solve the problem either because the problem just gets shifted to the next gain stage in the chain, and we still have saturation. I also don't have a feeling for how much differential voltage these LT1125s can sustain before they are damaged - I guess the planned THD check will reveal if they are okay or not.
It seems to me like the only way to truly fix this problem of one stage saturating and screwing up the others is to use single Op27s (or equivalent) in place of the quad LT1125s. The aLIGO design also has a series resistance to the non-inverting input - this can help prevent current overdraw from the previous stage (due to a lowered input impedance of the OpAmp - but I wonder how low this can go?).
I have acquired 5 pieces of the Teledyne AP1053 from Koji - these are now at the 40m. I will determine an appropriate location for storage of these and update. We are also looking to acquire 5 more of these. The combination of high power output (26dBm), low gain (10dB), and low noise figure (1.5dB) are quite uncommon in an amplifier and so these should be used only when such properties are required simultaneously.
*Steve informs me that these amps have been stored in the RF cabinet E6 along the east arm.
Steve's note: Teledyne rf amp product selection guide
Teledyne rf low noise amp guide
I did some work on the PSL table today. Main motivations were to get a pickoff for the BeatMouth PSL beam before any RF modulations are imposed on it, and to improve the mode-matching into the fiber. Currently, we use the IR light reflected by the post doubling oven harmonic separator. This has the PMC modulation sideband on it, and also some green leakage.
So I picked off ~8.5mW of PSL light from the first PBS (pre Faraday rotator), out of the ~40 mW available here, using a BS-80-1064-S. I dumped the 80% reflected light into the large beam dump that was previously being used to dump this PBS reflection. Initially, I used a R=10% BS for S-pol that I found on the SP table, but Koji tipped me off on the fact that these produce multiple reflected beams, so I changed strategy to use the R=80% BS instead.
The transmitted 20% is routed to the West edge of the PSL table via 2 1" Y1-1037-45S optics, towards the rough vicinity of the fiber coupler. For now it is just dumped, tomorrow I will work on the mode matching. We may want to cut the power further - ideally, we want ~2.5mW of power in the fiber - this is then divided by 4 inside the beat mouth before reaching the beat PD, and with other losses, I expect ~500mW of PSL power and comparable AUX light, we will have a strong >0dBm beat.
Attachment #1 is a picture of my modifications. For this work, I
I plan to do some characterization of this problem. The plan is to use THD as a metric for whether we are having hidden saturations. Pg 9 of the LT1125 datasheet tells us what fraction of THD to expect. I will use one of the several unused DAC channels available at the LSC rack to drive a 100Hz sine wave into one of the inputs of the whitening chassis, and measure the THD up to a reasonable harmonic number (will probably be set by the ADC noise) for (i) various whitening gain settings and (ii) various input signal amplitudes.
The motivation is to attempt to quantify the problem better:
Then we can decide what, if anything, to do about this issue.
On Friday, while Udit and I were doing some characterization of the EX+PSL IR beat at the LSC rack, I noticed that there were sidebands around the main beat peak at 20dBm lower level. These were offset from the main peak by ~200kHz - I didn't do a careful characterization but because of the symmetric nature of these sidebands and the fact that they appeared with the same offset from the main peak for various values of the central beat frequency, I hypothesize that these are from the modulation sidebands we use for PDH locking the EX laser to the arm cavity. So we can estimate the modulation depth from the relative powers of the main beat peak and the ~200kHz offset sidebands.
Since the IR light is used for the beat and we directly couple it to the fiber to make the beat, there is no green or IR cavity pole involved here. 20dBm in power means . And so the modulation depth, . I will do a more careful meaurement of this, but this method of measuring the modulation depth can give us a precise estimate - for what it's worth, this number is in the same ballpark as the measurement I quote in elog12105.
What is the implication of having these sidebands on our ALS noise? I need to think about this, effectively the phase noise of the SR function generators we use to do the phase modulation of the EX laser is getting imprinted on the ALS noise? Is this hurting us in any frequency range that matters?
I was looking into the physics of polarization maintaining fibers, and then I was trying to remember whether the fibers we use are actually polarization maintaining. Looking up the photos I put in the elog of the fibers when I cleaned them some months ago, at least the short length of fiber attached to the PD doesn't show any stress elements that I did see in the Thorlabs fibers. I'm pretty sure the fiber beam splitters also don't have any stress elements (see Attached photo). So at least ~1m of fiber length before the PD sensing element is probably not PM - just something to keep in mind when thinking about mode overlap and how much beat we actually get.
I replaced the two remaining D-Sub M/M cables that still had gender-changers with M/F cables today, completing the mechanical and wiring work on the ETMX rack. All backplane adapter boards were secured to a cross-strut of the crate using zip ties. This was necessary because the adapter boards don't fit the crate with their panels attached ( the ETMX eurocrate is the only one with slightly different dimensions from all the others), and the we can't mount them to the strut using the panels. This won't be an issue on any of the other crates.
In other news:
I disabled the legacy support in the three Acromag ADC units and set the input averaging to 10 samples via the USB configuration utility. The documentation is unfortunately a little sparse about what this actually means. The manual states that "fresh input data is available to the network every 10ms", so the sampling rate is for sure faster than 100Hz. Since the IOC updates its channels every .1 seconds I assume that an averaging value of 10 to reduce the input noise is safe. The maximum value the configuration tool permits is 200. I tried this using the CryoLab DAQ and set all input channels to 200 and used StripTool to look at the time series of a slow oscillation (.1Hz) with a large amplitude (16Vpp) and looked for missed data points, indicating too long wait times for channels updates. There was no such qualitative difference between 1 sample, 10 samples, and 200 samples, so even pushing the averaging value to max seemed okay. I went with the conservative value of 10 for the ETMX DAQ, but we can likely increase this if noise on the slow inputs becomes an issue.
The input scaling of the ADC channels has been corrected. I changed the conversion method in the EPICS records from manual using the ASLO and AOFF fields to using engineering units via EGUF and EGUL. This required a little attention. The Acromags scale the dynamic input range of +/- 10V to +/- 30,000 raw value, but the EPICS IOC interprets the data type as ranging from -32767 to +32768, so the EGUF and EGUL fields must be set to -10.923 and +10.923 to achieve proper scaling. I also changed the SCAN field on all ADC channels to 0.1 seconds. This has been done for all ADC and DAC channel records.
I compiled the burt binaries on c1auxex2 which took a little fiddling with dependencies and paths but nothing too major. The complete local epics folder (/opt/epics/) which contains the base epics binaries, modbus and burt for 32-bit linux has been copied to the shared drive at /opt/rtapps/epics-3.15.5. They belong to the most recent stable release. This was so we can now automatically call burt after the IOC initialization on c1auxex2 to restore the backed-up channel values.
I also copied the database definition and modbus instruction files to /cvs/cds/caltech/target/c1auxex2, from where they are now being read upon IOC initialization. This is an excerpt of the service file:
#ExecStart=/usr/bin/procServ -f -L /home/controls/modbusIOC/modbusIOC.log -p /run/modbusioc.pid 8008 /opt/epics/modules/modbus/bin/linux-x86/modbusApp /cvs/cds/caltech/target/c1auxex2/ETMXaux2.cmd <-- Contains logging to file, see note 1)
ExecStart=/usr/bin/procServ -f -p /run/modbusioc.pid 8008 /opt/epics/modules/modbus/bin/linux-x86/modbusApp /cvs/cds/caltech/target/c1auxex2/ETMXaux2.cmd <-- Initializes the EPICS IOC with Modbus support
ExecStop=/bin/kill -9 ` cat /run/modbusioc.pid` <-- Kills the detached process by its process ID
ExecStartPost=/bin/bash -c "/opt/epics/extensions/bin/linux-x86/burtwb -f /opt/rtcds/caltech/c1/burt/autoburt/latest/c1auxex.snap" <-- Restores general channel values
ExecStartPost=/bin/bash -c "/opt/epics/extensions/bin/linux-x86/burtwb -f /opt/rtcds/caltech/c1/medm/MISC/ifoalign/burt/ETMX.snap" <-- Restores PIT and YAW values from align MEDM screen
ExecStartPost=/bin/bash -c ". /home/controls/modbusIOC/ETMXaux2.sh" <-- Enables writing to PIT and YAW DAC channels, see note 2)
Note 1) I removed the logging to file for now because I noticed that if there are Acromag communication issues the logfile tends to grow in size VERY fast. In the cryo lab is had gotten to over 70GB just over the winter break. I don't think it's absolutely necessary to have it, and if diagnostics are needed we can easily uncomment it temporarily.
Note 2) I modified the static EPICS records of the four OSEM bias adjust channels so they won't start updating as soon as the IOC starts up (and before the channel defaults are restored by burt). This was done by setting the OMSL (output mode select) field from "closed_loop" to "supervisory". Sample record:
field(DESC,"Bias Adjust for ETMX UL Coil Output")
field(OUT, "@asynMask(C1AUXEX_XT1541A_DAC, 0, -16)MODBUS_DATA")
field(OMSL,"supervisory") <-- Used to be "closed_loop"
field(DOL, "C1:SUS-ETMX_ULBiasSet PP")
Now, on reboort/IOC re-initialization the physical DAC channels are performing a one-time readback of the last stored value in the Acromag's register, then idle until the last StartPost statement executes the script ETMXaux.sh, which changes their OMSL field back to "closed_loop". This causes them to start updating their output from the calc records defined in their DOL field (which have by then recovered their default values curtesy of burt). The result is a smooth transition from idling to the controlled state with no sudden or large offset changes.
The result is a smooth transition from idling to the controlled state with no sudden or large offset changes.
While checking how smooth the transition is we still noticed significant motion of ETMX by looking at the locked green laser and OpLevs. We found that this motion was not caused by interruption of the slow offset adjust, but rather the Watchdog being re-initialized to its OFF state, which cuts the fast channels OFF. On other optics this is observed too, but not as severe. The cause is a rather large offset on the LR coil coming from the fast DAQ, which was reported as 50mV by the slow readback channel (while other readback values are <10mV). It is present even when turning the output of the CDS model OFF, but vanishes when the watchdog is triggered. This helped us trace it to an offset of the DAC output itself: it is present at the output of the AI board but vanishes when the DAC is disconnected. The actual offset is ~40mV, as opposed to other channels on the same board, which ahve offsets in the range 3-7mV.
While we can compensate for this offset in software - it made us wonder if the DAC channel is somehow busted and if that's what causing the 'wandering' of ETMX that we have been observing recently. There are two free DAC channels on the AI chassis that has the side coil and the green temperature control signals. We could re-route the LR signal through a different DAC channel to fix this.
gautam: 40mV offset at the AI board output gets multiplied by 3 in the dewhitening board, so there is a 120mV DC offset going to the coil (measured at dewhite board output with DMM). The offset itself isn't hurting us, but the fact that it is several times larger than other channels led us to wonder if it could be drifting around as well. From my SOS pitch balancing forays, in my head I have the number 30mrad as being the full range of the OSEM actuation - so if the offset swings by 120mV, that's ~150urad of motion, which is quite large, and is of the order of magnitude I'm used to seeing ETMX move around by.
M4 local earthquake at 10:10 UTC There is no sign of damage.
....here is an other one.........M5.8 Ferndale, CA at 16:40 UTC
We started the actual heating test today and it seems to be working so far. Hoping to heat it to about 40C. We also set up another temperature sensor to measure the lab temperature and connected it to J7, bottom.
Gautam and I set up the insulated seismometer can in the lab today. I had previously wired up the two heaters I placed onto the sides of the can in parallel to get a total resistance of 12.5 ohms and then I wrapped the whole can in 3 layers of insulation (k-factor 0.25). We placed it on a large sheet of insulation as to not crush the wires leading out the bottom of the can. I stuck on one of my AD590 sensors to the inside of the can onto the copper lining using duct tape, though this is only a temporary solution. In the future, it would be nice to have some sort of thermal clamp to secure the sensor to the can. To provide power to the heater circuit board and the temperature sensor board, we got a powerstrip and plugged in two power supplies and a function generator into it. The heater circuit (attachment 3) is powered by one of the power supplies and the function generator, while the temperature sensor (attachment 5) is stuck to the side of the can and is powered by the second power supply. The heater circuit's MOSFET (IRF640, attachment 4) is placed on a metal block and sandwiched between two more to make sure it doesn't move around. The temperature sensor is connected by a long BNC cable to the channels in attachment 6.
gautam: we plugged the BNC output of Kira's temperature sensor circuit to J7 on the AA input chassis in 1X2 - this corresponds to ADC1 input 12 in c1ioo. I then made a "PEM" namespace block inside the c1als model, and placed a single CDS filter module inside it (this can be used for calibration purposes). The filter module is named "C1:PEM-SEIS_EX_TEMP", and has the usual CDSfilt channels available. I DQ'ed the output of the filter module (@256 Hz, probably too high, but I'm holding off on a recompile for now). Recompilation and model restart of c1als went smoothly.
2 bench power supplies are being used for this test, we can think of a more permanent solution later.
**25 Jan noon: Added another filter module, "C1:PEM-SEIS_EX_TEMP", to which Kira is hooking up a second temperature sensor, which will serve as a monitor of the "Ambient" lab temperature. Added DQ channel for the output of this filter module, fixed sampling to 32Hz. Compile and restart went smooth.
I was looking at this a little more closely. As I understand it, the purpose of the audio differential IF amplifier is:
Attachment #1 shows, the changes to the TF of this stage as a result of changing R19->50ohm, R17->500ohm. For the ALS application, we expect the beat signal to be in the range 20-100MHz, so the 2f frequency component of the mixer output will be between 40-200MHz, where the proposed change preserves >50dB attenuation. The Q of the ~500kHz resonance because of the series LCR at the input is increased as a result of reducing R17, so we have slightly more gain there.
At the meeting yesterday, Koji suggested incorporating some whitening in the preamp itself, but I don't see a non-hacky way to use the existing PCB footprint and just replace components to get whitening at audio frequencies. I'm going to try and measure the spectrum of the I and Q demodulated outputs with the actual beat signal to see if the lack of whitening is going to limit the ALS noise in some frequency band of interest.
Does this look okay?
The demod circuit board is configured to have gain of x100 post demod (conversion loss of the mixer is ~-8dB). This works well for the PDH cavity locking type of demod scheme, where the loop squishes the error signal in lock, so most of the time, the RF signal is tiny, and so a gain of x100 is good. For ALS, the application needs are rather different. So we lowered the gain of the "Audio IF amplifier" stage of the circuit from x100 to x10, by effecting the resistor swaps 10ohms->50ohms, 1kohm->500ohms (more details about this later).
After almost 3 hours the temperature rose by about 3.5C. Seems a bit slow, but we can drive it more if necssary. The heating curve itself is exponentiial, which is a good sign.
The final temperature reached in about 4.5 hours is 30.5C, while the starting temperature is about 24C. I can't seem to screenshot the data for some reason.
Also, I will calibrate the lab temperature sensor to Celcius in the near future so that we would have a working sensor inside the lab.
I tried to couple the PSL pickoff into the fiber today for several hours, but got nowhere really, achieved a maximum coupling efficiency of ~10%. TBC tomorrow... Work done yesterday and today:
I think part of the problem was that the rejected beam from the PBS was not really very Gaussian - looking at the spot on the beam profiler, I saw at least 3 local maxima in the intensity profile. So I'm now switching strategies to use a leakage beam from one of the PMC input steering optics- this isn't ideal as it already has the PMC modulation sideband on it, and this field won't be attenuated by the PMC transmission - but at least we can use a pre-doubler pickoff. This beam looks beautifully Gaussian with the beam profiler. Pics to follow shortly...
I tried to couple the PSL pickoff into the fiber today for several hours, but got nowhere really, achieved a maximum coupling efficiency of ~10%. TBC tomorrow... Work done yesterday and today
pd80b rga scan at 175 day. IFO pressure 7.3e-6 Torr-it
Condition: vacuum normal, annuloses not pumped. Rga turned on yesterday.
The rga was not on since last poweroutage Jan 2, 2018 It is warming up and outgassing Atm2
I've often gotten confused by the labeling on the SUS MEDM screens about the coil "Vmon" fields - they're labelled as "30 Hz HPF", and indeed this is one of the many readbacks available on the coil driver board. But the actual EPICS channel that is being displayed in this field is from the "EPICS VMON" monitor point on the coil driver board. It has a gain of 1/2, so the actual voltage going to the coil is twice the channel value. Today, I fixed the SUS master screen to avoid this confusion - new labeling is shown in Attachment #1.
I moved the epics IOC server process for the single Acromag ADC that monitors the PSL signals from megatron to c1auxex2.
First, I disabled the legacy support on all channels as explained in elog 13565. Then I copied the files npro_config.cmd and NPRO.db from /opt/rtcds/caltech/c1/scripts/Acromag to /cvs/cds/caltech/target/c1psl2/ following the pattern of the old Motorola machines and the new c1auxex2. I had to make some edits for correct paths and expanded the epics records to the standard we're using for ETMX.
I then added a service to systemd on c1auxex2 that runs the epics IOC for the Acromag PSL channels: /etc/systemd/system/modbusPSL.service. No more tmux on megatron.
Running two IOCs on a signle machine at the same time did not produce any errors and seems fine so far.
Attachment #1 shows the current situation of the PSL table IR pickoff. It isn't the greatest photo but it's hard to get a good one of this setup. Now there is no need to open the Green PSL shutter for there to be an IR beat note.
All this lead me to conclude that I have reached at least some sort of local maximum. The AR coating of the lens has ~0.5% reflection at 8 degrees AOI according to spec, and EricG mentioned today that the fiber itself probably has ~4% reflection at the interface due to there not being any special AR coating. There is also the fact that the mode of the collimator isn't exactly Gaussian. Anyways I think this is a big improvement from what was the situation before, and I am moving on to debugging the ALS electronics.
There is 3.65mW of power coupled into the fiber - our fiber coupled PDs have a damage threshold of 2mW, and this 3.65mW does get split by 4 before reaching the PDs, but good to keep this number in mind. For a quick measurement of the PMC and X end PDH modulation depth measurements, I used an ND=0.5 filter in the beam path.
Annual crane inspection with load tests is scheduled for Monday, Feb 5, 2018 from 8 to 11:30am
Konecranes rescheduled this appointment to: Monday, Feb 12, 2018
I used the Beat Mouth to make a quick measurement of the PMC and EX modulation depths. They are, respectively, 60mrad and 90mrad. See Attachments #1 and #2 for spectra from the beat photodiode outputs, monitored using the Agilent analyzer, 16 averages, IF bandwidth set to resolve peaks offset from the main beat frequency peak by 33.5MHz for the PMC and by ~230kHz for the EX green PDH.
For this work, I had to re-align the IFO so as to lock the arms to IR. c1susaux was unresponsive and had to be power-cycled. As mentioned in the earlier elog, to avoid saturating the Fiber Coupled beat PDs, I placed a ND=0.5 filter in the fiber collimator path, such that the coupled power was ~1mW, which is well inside the safe regime.
For the EX modulation depth, I could have gotten multiple estimates of the modulation depth using the higher order products that are visible in the spectrum, but I didn't.
We installed some rails to mount the 2U chassis containing ~100m of delay line cabling, and the 1U chassis containing the FET demodulators for the ALS signals in the LSC rack. This has made it MUCH easier for a single person to work there and remove/reinstall these chassis. The delay line box has 100m of cable inside it, and so was rather heavy (~8kg) - previously, it was being supported only by a pair of brackets on the front, so the new arrangement is much more robust. Steve is looking into acquiring plastic spacers of the appropriate width, so that we can secure the units to the rack using usual rack mount screws (but the material of the newly installed rails and the screw heads holding them in place necessitate this plastic spacer).
Delay line box has been re-installed, demodulator chassis has been removed by me for characterization. Steve will put up photos once the units are re-installed.
For this work, I had to disconnect a bunch of cabling, but only those connected to ALS. All cables were labelled, and I will re-connect them once I am done with the demod chassis.
Anyways I think this is a big improvement from what was the situation before, and I am moving on to debugging the ALS electronics.
The plan is to lower the gain of the IF amplifier stage on the FET demodulator board from 100 to 10. As per Attachment #1, this will make the overall gain from RF beatnote from the Beat Mouth to the signal input to the D990694 whitening board +19dB, assuming "typical" values for the conversion loss of the mixer, and the various other passive components on the FET demod board. I've used numbers I measured a couple of weeks ago for the delay line loss and the cabling loss from the PSL table to the LSC rack. This in turn will set a limit on how much RF beat power we can handle, from the Beat Mouth. According to this power budget, if we have -5dBm of beat, we will have an input to the whitening board of ~6Vpp, which is about half its full range. The trouble is, I don't know what the transimpedance gain of the Fiber Beat PDs are. The datasheet suggests a "maximum gain" of 5e4 V/W, which presumably takes into account the InGaAs responsivity and the actual transimpedance gain. However, according to the last power budget I did inside the Beat Mouth, I had -8dBm of beat for a combined 400uW of PSL+EX light, which definitely does not add up. I've emailed the company to ask about the spec, haven't gotten anything useful yet...
The problem is further complicated by the fact that the fiber inside the Beat Mouth is NOT polarization maintaining, and so the actual relative polarizations of the arm IR light and the PSL IR light is unpredictable, and also uncontrolled. I suppose we could simply place a HWP before the fiber collimator at either end, and rotate the polarization until we get a desired amount of beat, but this still does not solve the problem of the polarization being uncontrolled.
I am going to characterize the demod board using E1100114. I am unsure as to the conversion loss of the mixer - the datasheet suggested a number of 8dB, but T1000044 suggests that the conversion loss is actually only 4dB. I figure it's best to just measure it. Would also be good to verify that the overall transfer function and noise of the IF amplifier stage match my expectation from the LISO model.
Option #1: Rana ordered 50ohm and 500ohm SMD resistors of the 0805 package size, I asked Steve to get a few more values just in case we want to twiddle with the gain of this stage further (specifically, I asked for values such that we can set it to x5, x3 and x1). But changing the feedback resistors modifies the overall TF shape - see e.g. Attachment #2. Need to also look at how the noise performance varies.
Another possibility is to turn down the gain of the IF amplifier stage to x10, retire the ZHL-3A, and use a lower gain amplifier in its place. We do have the recently acquired Teledyne amplifiers, but we would have to package it in such a way that it can be integrated into the existing Fiber ALS signal chain. This would allow us to handle significantly larger RF beatnote powers, which I expect we will have if we improve the mode matching into the fibers (provided the aforementioned polarization drift possibility doesn't hurt us too much).
A third possibility is to attenuate the power coupled into the fibers to lower the RF beatnote amplitude. I don't like this option so much because placing an ND filter or a PBS+HWP combo in the beam path is likely to screw up the mode-matching into the fiber collimator, which I have already spent so many hours trying to improve, but if it must be done, it must be done.
The correct option is of course the one that gives us the lowest ALS noise. It is not clear to me which one that is at this point.
I effected the change to the Audio IF preamp stage on channels 3 and 4 (Xarm and Yarm respectively) using the resistors Steve ordered (the ones Rana ordered don't have any labeling on them, and I couldn't tell the 50ohm and 500ohm ones apart except by looking at the label on the ziplock bag they came in, so I decided against using them). I've started a DCC page to collect photos, characterization data, and marked up schematic etc for this part. Characterization is ongoing, more to follow soon. Note that for the photo-taking, I disconnected all the on-board SMA connectors so that the cabling wouldn't block components. I have since restored them for testing purposes, and was careful to use the torque-limited SMA tightening tool when restoring the connections.
In order to test various things like conversion loss etc, I figured it would be useful to have two RF signal sources, so I scavenged the Fluke RF generator that Johannes was using from under the PSL table. In the process, I accidentally bumped the PSL interlock on the southeast corner of the PSL table. I immediately turned the NPRO back on, and relocked PMC/IMC. Everything looks normal now. Acromag may even have caught my transgression.
Stuff is beginning to look clearer now that I've done some initial characterization of the demod boards. I will upload a more detailed report of the characterization on the DCC page, but important findings are:
The delay line has a loss of ~3dB. The power splitter has a loss of 3dB. So putting everything together, 17dBm at the input of the power splitter gives us just the right amount of RF power to have the LO input driven at 14dBm, and the IF output be ~5Vpp into a High-Z load, which is about half the ADC full range.
After taking the measurements, calibrating them (approximately), and filterting them, I created the following plot. The exponential fit is quite good, as the error is not more than 0.03 C. I used the python function curve_fit in order to get this, and it gave me the time constant as well, which came out to 0.357 hr. From my previous calculations here, I plugged in the values we have (m = 12.2 kg, c = 500 J/kg*k, d = 0.0762 m, k = 0.26 W/(m^2*K), A = 1 m^2), and got that
This is a bit off, but it's probably due to the parameters not being exactly what I supposed them to be, and heat losses through the bottom of the can.
I saw some interesting behaviour of the Audio IF amplifier stage on the demod board today, by accident. I was testing the board for I/Q orthogonality and gain balance, when I noticed a large gain imbalance between the I and Q channels for both Board #3 and #4, which are the ones we use for the IR ALS demodulation. This puzzled me for some time, but then I realized that I had only reduced the gain of this stage from x100 to x10 for the I channel, and not for the Q channel! The surprising thing though was that the output waveform still looked like a clean sinusoid on the o'scope, and there was no evidence of the voltage clipping that is characteristic of an op-amp being driven beyond its voltage rails. The conversion factor with a preamp gain on x10 was measured today to be 2V IF / 1V RF. But this means that for a preamp stage gain of x100, we expect 20V IF / 1V RF, which is well in the saturation regime of the AD829, since the Vcc is only +/-15V. I'm guessing the diodes D2 and D3 are for overvoltage protection, but given that the pre-amp gain is x100, the input signal at the inverting input of the AD829 is only 0.2V at DC, which isn't above the forward bias voltage for the switching diode BAV99. Perhaps there is some interaction between the pre-amp and the FET demodulator that I dont understand, or I am missing something about the differential to single-ended topology that would explain this behaviour.
I found it puzzling why the large preamp stage gain didn't hurt us with the green beat - even though the green optical beat signal was smaller than the current IR beat, a back-of-the-envelope calculation suggested that it would still have saturated the ADC with a x100 gain on the preamp. Perhaps this observation is part of the story, and there is also the unpredictable behaviour of the D990694 board for an input signal with large DC levels...
I did the following tests on this board today:
I didn't really measure the transfer function of the preamp stage after the modification because there wasn't a convenient test point and I couldn't find the high impedance FET probe for the Agilent - I wonder if somebody in WB has it? Anyways, all the tests suggested the board is operating as expected, and I now have calibrations for the back panel DSUB for LO/RF power levels, and also the conversion gain from RF to IF. I will put together a python notebook with all my measurements and upload it to the DCC page for this part. I need to double check expected noise levels from LISO to match up to the measurement.
I will now proceed to the next piece (#3?) of this puzzle, which is to understand how the D990694 which receives the signals from this unit reacts to the expected DC voltage level of ~4Vpp.
After discussion with Koji, I have also decided to look into putting together a daughter board for an alternative Audio IF preamp stage. The motivation is that for the ALS application, we expect a high DC signal level all the time (because the loop does not suppress the beat note amplitude). So we would like for the preamp stage to have the usual shape of some zero around 4Hz, a pole around 40Hz, and then the LowPass profile of the existing preamp stage (to cut out the 2f frequency product, but also to minimize the possibility of the fast AD829 going into some unpredictable regime where it oscillates). So, the desired features are:
While setting up for this measurement, I noticed something odd with the whitening switching for the ALS channels. For the usual LSC channels, the whitening is set up such that switching FM1 on the MEDM screen changes a BIO bit which then enables/disables the analog whitening stage. But this feature doesn't seem to be working for the ALS channels - I terminated all 4 channels at the LSC rack, and measured the spectrum of the IN1 signals with DTT in the two settings, such that I expect to see a difference in the spectra if the whitening is enabled or disabled - FM1 enabled (expected analog whitening to be engaged) and FM1 disabled (expected analog whitening to be bypassed). But I see no difference in the spectra. I confirmed that the BIO bit switching is happening at least on the software level (i.e. the bit indicator MEDM screens indicate state toggling when FM1 is ON/OFF). But I don't know if something is amiss in the signal chain, especially since we are using Hardware channels that were previously used for AS_165 and POP_55 signals.
Is the whitening shape such that we expect the terminated noise level to be below ADC Noise even when the whitening is engaged? I just checked the shape of the de-whitening filter, and it has -40dB gain above 150Hz, so the inverse shape should have +40dB gain.
I will now proceed to the next piece (#3?) of this puzzle, which is to understand how the D990694 which receives the signals from this unit reacts to the expected DC voltage level of ~4Vpp
gautam 2.15pm: This was a FALSE ALARM, with the inputs terminated, the electronics noise really is that low such that it is buried under ADC noise even with +40dB gain. I cranked up the flat whitening gain from 0dB to 45dB for the X channels (but left the Y channels at 0dB). Attachment #2 is the comparison. Looks like the switching works just fine.
I could not understand why 'netgpibdata' scripts are missing in "scripts/general" folder on pianosa... Where did they go???
Also, I found that the PROLOGIX GPIB-LAN controller for crocetta (192.168.113.108) is no longer working. I need to reconfigure it with "telnet"...
The netgpibdata scripts are now under git version control at /opt/rtcds/caltech/c1/scripts/general/labutils/netgpibdata. I think the idea was to make this directory a collection of useful utilities that we could then pull at various labs / at the sites.
Attached the program I used to create the plot
I've been trying to setup for the THD measuremetn at the LSC rack for a couple of days now, but am plagued by a problem summarized in Attachment #1: there are huge harmonics present in the channel when I hook up the input to the whitening board D990694 to the output of a spare DAC channel at the LSC rack. Attachment #2 summarizes my setup. I've done the following checks in trying to debug this problem, but am no closer to solving it:
Am I missing something obvious here? I think it is impossible to do a THD measurement with the spectrum in this condition...
crochetta was reconfigured to have 192.168.113.108. It was confirmed that it can be used with netgpibdata.py
Configuration: I have connected my mac with the unit using an Apple USB-Ethernet adapter. The adapter was configured to have a manual IP of 192.168.113.222/255.255.255.0. "netfinder.exe" was run to assign the IP addr to the unit. It seemed that NVRAM of the unit evaporated as it had the IP of 0.0.0.0. Once it was configrued, it could be run with netgpibdata as usual.
Did some quick additional checks to figure out what's going on here.
So either something is busted on this board (power regulating capacitor perhaps?), or we have some kind of ground loop between electronics in the same chassis (despite the D990694 being differential input receiving). Seems like further investigation is needed. Note that the D000316 just two boards over in the same Eurocrate chassis is responsible for driving our input steering mirror Tip-Tilt suspensions. I wonder if that board too is suffering from a similarly noisy ground?
I think I've narrowed down the source of this ground loop. It originates from the fact that the DAC from which the signals for this board are derived sits in an expansion chassis in 1Y3, whereas the LSC electronics are all in 1Y2.
Looking at Jamie's old elog from the time when this infrastructure was installed, there is a remark that the signal didn't look too noisy - so either this is a new problem, or the characterization back then wasn't done in detail. The main reason why I think this is non-ideal is because the tip-tilt steering mirrors sending the beam into the IFO is controlled by analogous infrastructure - I confirmed using the LEMO monitor points on the D000316 that routes signals to TT1 and TT2 that they look similarly noisy (see e.g. Attachment #1). So we are injecting some amount (about 10% of the DC level) of beam jitter into the IFO because of this noisy signal - seems non-ideal. If I understand correctly, there is no damping loops on these suspensions which would suppress this injection.
How should we go about eliminating this ground loop?
I decided to plot the temperatures measured over two days for the sensor inside the can and inside the lab just to see if there was any significant difference between the two, and obtained the following plot. This shows that there is a difference in measurements of a few 0.01 C. The insulated seismometer can didn't change temperature as much as the lab did, which is as expected. I'll work on properly calibrating the sensors sometime in the future so that we can use the sensor that's just in the lab as an accurate thermometer.
We discussed possible solutions to this ground loop problem. Here's what we came up with:
Why do we care about this so much anyways? Koji pointed out that the tip tilt suspensions do have passive eddy current damping, but that presumably isn't very effective at frequencies in the 10Hz-1kHz range, which is where I observed the noise injection.
Note that all our SOS suspensions are also possibly being plagued by this problem - the AI board that receives signals is D000186, but not revision D I think. But perhaps for the SOS optics this isn't really a problem, as the expansion chassis and the coil driver electronics may share a common power source?
gautam 1530 7 Feb: Judging by the footprint of the front panel connectors, I would say that the AI boards that receive signals from the DACs for our SOS suspended optics are of the Rev B variety, and so receive the DAC voltages single ended. Of course, the real test would be to look inside these boards. But they certainly look distinct from the black front panelled RevD variant linked above, which has differential inputs. Rev D uses OP27s, although rana mentioned that the LT1125 isn't the right choice and from what I remember, LT1125 is just Quad OP27...
After emailing the technical team at Menlo, I have uploaded the more detailed information they have given me on our wiki.
The trouble is, I don't know what the transimpedance gain of the Fiber Beat PDs are. The datasheet suggests a "maximum gain" of 5e4 V/W, which presumably takes into account the InGaAs responsivity and the actual transimpedance gain.
I subtracted out the lab temperature change during the period of cooling to see if it would have a significant effect on the time constant, but when I fit the new data, the time constant came out to 0.355 hr, which is not a significant change from the value of 0.357 that I got earlier.
Hornet cold cathode gauge analoge output [ DSub9 pin 3 and 7 ] are wired to go ETMX Acromag. It was reading 4.9V at 7.8e-6 Torr [ 3,110 V 8.35e-5A ] at the end of a 24ft BNC cable. Now it has to be hook up to an Acromag channel.
This will replace the not functioning C1: Vac-CC1_pressure
gautam: the motivation behind hooking this gauge up to our DAQ system is that non-vacuum-system-experts have a quick diagnostic to make sure everything is in order. This gauge is physically placed adjacent to V1, and so if something goes wrong with our vacuum pumps, we would see the effect here immediately. we did note that occassionally, the reading fluctuated by ~1V on the DMM used to check the voltage output at the end of the BNC cable, so we still need to run some long-term stability analysis once this channel is hooked up to the Acromag. For future reference, in order to make this gauge work, we need to check that
There was a power outage.
The IFO pressure is 12.8 mTorr-it and it is not pumped. V1 is still closed. TP1 is not running. The Rga is not powered.
The PSL output shutter is still closed. 2W Innolight turned on and manual beam block placed in its beampath.
3 AC units turned on at room temp 84F
IFO pumped down from 44 mTorr to 9.6e-6 Torr with Maglev backed with only TP3
Aux drypump was helping our std drypump during this 1 hour period. TP3 reached 32 C and slowed down 47K rpm
The peak foreline pressure at P2 was ~3 Torr
Hornet cold cathode gauge setting: research mode, air,
2830 HV 1e-4A at 9.6e-6 Torr,
[ 3110 HV 8e-5A at 7.4e-6 Torr one day later ]
Annuloses are at 2 Torr, not pumped
Valve configuration: vacuum normal, RGA is still off
PSL shutter is opened automatically. Manual block removed.
End IR lasers and doublers are turned on.
NOTE: Maglev " rotation X " on vacuum medm screen is not working! " C1:Vac-TP1_rot " channel was removed. Use " NORMAL X " for rotation monitoring.
*We removed this (i.e. rotation) field from the MEDM screen to avoid confusion.
Summary of my tests of the demod boards, post gain modification:
Everything looks within the typical performance specs outlined in E1100114, except that the measured noise levels don't quite line up with the LISO model predictions. The measurement was made with the scheme shown in Attachment #1. I didn't do a point-by-point debugging of this on the board. I have uploaded the data + notebook summarizing my characterization to the DCC page for this part. I recommend looking at the HTML version for the plots.
*I'd put up the wrong attachment, corrected it now...
I will put together a python notebook with all my measurements and upload it to the DCC page for this part. I need to double check expected noise levels from LISO to match up to the measurement.
gautam 9 Feb 2018 9pm: Adding a useful quote here from the LISO manual (pg28). I think if I add the Johnson noise from the output impedance of the mixer (assumed as 50ohms, I get better agreement between the measured and observed noises (although the variance between the 4 channels is still puzzling). The other possible explanation is small variations in the voltage noise at the various mixer output ports. Could we also be seeing the cyclostationary shot noise difference between the I and Q channels?
In any case, I am happy with this level of agreement, so I am going to stick this 1U chassis back in its rack with the primary aim of measuring a spectrum of the beatnote, so that I have some idea of what kind of whitening filter shape is useful for the ALS signals. May need to pull it out again for actually implementing the daughter board idea though... I have updated DCC page with LISO source, and also the updated python notebooks.
The ETMX Sorrenson power supply -15V was running at -13.9V