DataDisplay config. 

Here is a list of dataDisplay usual configs:

• CCL (CCL piezo and 20 MHz demodulation of the subtraction of HD photodiodes)
• HD_offset (HF photodiodes DC and audio signals in time and 1D)
• MZ (MZ DC and DAC)
• OPO (OPO DC and error signal, DAC on OPO piezo and sum of HD photodiodes)
• OPO_HD (OPO DC and DAC, HD photodiodes DC signals)
• SHG1 (SHG1 DC, DAC and error signal)
• SHG2 (SHG2 DC, DAC and error signal)
• SHGs (SHG1 and SHG2 DC and DAC signals)
• temps (OPO and SHG2 temperatures)

Hard and soft communication. 

How to communicate with the OPO translation stage? 

1. Connect the USB cable of the OPO translation stage to the pc-phasics (the pc should change at one moment).  
2. Open the Newport AG-UC2-UC8 applet. /!\ Be careful not to disconnect the usb cable until the applet is closed, this could break the controller /!\  
3. Click on « Discover to see the COM ports and then launch applet on the COM used by the stage (COM6 for now on pc-phasics, you can check that on the « Gestionnaire de périphérique« ) 
4. For this stage you can only do incremental motion. Write a number on the case and click on the left or right button depending on the direction you want to go. 
5. To disconnect the stage, close the applet and then (and only then) disconnect the USB cable.
   

How to communicate with the lens translation stages?

1. Connect the USB cable of the desired translation stage to the pc-phasics(the pc should change at one moment). 
2. Connect the power cable of the corresponding controller.  
3. Open the SMC100 Utility. /!\ Be careful not to disconnect power until the utility is closed, this could break the controller /!\ 
4. Click on « Discover to see the COM ports and then launch applet on the COM used by the stage 
5. For these stages you can chose an absolute position between 0 and 25 mm. 
6. To disconnect the stage, close the applet and then (and only then), disconnect the power cable of the controller and then disconnect the usb cable.

How to open the PyRPL GUI?

1. Open a Jupyter Notebook (this could take few seconds)  
2. In the opened internet window, click on « Documents » -> « Python Scripts » ->  « Pyrpl launcher.ipynb » (the loading of the page could take time after a long shutdown) 
3. Click on the second cell of the opended notebook (with « michael_pfd.yml ») and execute it. 
4. The GUI opens in a new window (this takes few seconds)
   

How to connect on TeamViewer? 

1. Open TeamViewer on the computer on which the screen you want to display is (computer 1) and on the computer you want to use the display the screen of the other computer (computer 2). 
2. On computer 2 write the computer ID found on computer 1 TeamViewer and click on “Connect” 
3. Enter the password and you should see computer 1 screen displayed on computer 2.
   

How to open a dataDisplay default config? 

 On a terminal on any calva linux computer, write: 

1. “SHG1” to open the SHG1 dataDisplay config (with DC, error and ramp signals)
   
2. “SHG2” to open the SHG2 dataDisplay config (with DC, error and ramp signals)
3. “SHGs ” to open the SHGs dataDisplay config (with DC and ramp signals for SHG1 and SHG2)
4. “MZ” to open the MZ dataDisplay config (with DC and ramp signals)
   
5. “OPO” to open the OPO dataDisplay config (with DC, error and ramp signals of the OPO and sum of the two homodyne photodiodes DC signal)
6. “OPO_HD” to open the OPO_HD dataDisplay config (with DC and ramp signals of the OPO and DC signals of the two homodyne photodiodes)
7. “HD_offset” to open the HD_offset dataDisplay config (with DC and Audio signals from the two homodyne photodiodes : time and 1D plots)
   
8. “CCL” to open the CCl dataDisplay config (with error and raamp signals)
9. “temps” to open the temps dataDisplay config (with SHG2 and OPO temperature signals)
    

Procedures. 

How to start sending data to the DAC? 

On the VPM webpage (https://pc-calva16.lal.in2p3.fr:8090/main.html), “DAC driving” section: 

1. Start the process “Initialisation” 
2. Start the process “DAC” 

If you want to send actuation to the coils: 

1. Start the process “Initialisation” 
2. Start the process “Coils” 
3. Start the process “DAC_LC” 

Then stop the process “Initialisation” 

How to control SHG1 crystal temperature? 

1. Start the commercial controller (ON/OFF on the rear panel) 
2. Select the temperature (displayed on the left) using the arrows. Ideally not too far (max 10 deg C) from the actual temperature (displayed on the right).  
3. Then use the “enable” button to send the actuation order (you should see “ON” displayed) 
   

You can change the temperature set event in “enable” mode.  

The last known value for the SHG1 crystal temperature is 52.8 deg C. This value can change by few degrees after few weeks of shutdown. So after long shutdown, check the optimum temperature by placing the powermeter on the green beam after the SHG1 dichroic plate, SHG1 cavity locked. Try to find the max value (depending on the SHG1 cavity, alignment this could vary between 100 and 200 mW) 

How to control SHG2/OPO crystal temperature? 

On the VPM webpage (https://pc-calva16.lal.in2p3.fr:8090/main.html), “Locking” section: 

1. Start the TempControl_SHG2/TempControl_OPO process 
2. Using the “T_set” button, fix the amplitude at 0 (so frequency is not important) and the offset at the chosen temperature in deg C, RampTime of 30s is fine
3. Using the “Scan/Lock Peltier” button, pass in lock mode (+1) 
4. Using the “Ramp” button, send voltage to the Peltier unit to heat the crystal and cross the temperature set, chosing the offset (0.9 V is a correct value to heat to more than 55 deg C and then is good to control around 52 deg; if you want to control around 30deg used 0.5 V ). The amplitude at 0.01 V, with 0.002 Hz frequency is OK, 0 could be either OK. 

 You can see the crystal temperature measure directly on VPM next to the GPS time (SHG_T1/OPO_T1) or using the temps dataDisplay config.

 If you want to change the temperature value while locked, adjust the temperature set offset and chose the RampTime depending on the amount of deg C you want to change (2 deg C <-> 30 s) not to lose the lock. 

 The last known value for the SHG2 crystal temperature is 53.1 deg C. This value can change by few degrees after few weeks of shutdown. So after long shutdown, check the optimum temperature by placing the powermeter on the green beam after the SHG2 dichroic plate, SHG2 cavity locked. Try to find the max value (depending on the SHG2 cavity, alignment this could vary between 20 and 40 mW). Another method to find the optimum temperature is to lock the SHG2 cavity and find the temperature minimizing the SHG2_DC power (as more infrared is converted into green).

 For the OPO crystal temperature, the best value to obtain nonlinear gain is 30 +/-1 deg C and if you don’t want nonlinear gain, 50 deg C is a good temperature. 

 How to lock the SHG1 cavity?

Starting point : on the VPM webpage (https://pc-calva16.lal.in2p3.fr:8090/main.html), “Locking” section, the “SHGMain” process is started 

1. Open the SHG1 dataDisplay config. 
2. Check that the SHG_DC signal is correct (~7-8 V for SHG1 and low high order modes) and errors signals look like PDH. 
3. Using the “Scan/Lock” button of the “SHGMain”  process, pass in lock mode (+1)  
4. If the lock is not so stable you can adjust the gain changing the filterGain parameter in the “SHG1_Filters.cfg” config files and then clicking on « Reload Filters » to send the new gain.
   

 How to lock the SHG2 cavity?

Starting point : on the VPM webpage (https://pc-calva16.lal.in2p3.fr:8090/main.html), “Locking” section, the “SHG2_Fast_Noise” and ”SHG2_Fast” process are started 

1. Open the SHG2 dataDisplay config. 
2. Check that the SHG2_DC signal is correct (~4-5 V for SHG2 and low high order modes) and errors signals look like PDH. 
3. Using the “Lock” button of the ”SHG2_Fast” process, pass in lock mode
4. If the lock is not so stable you can adjust the gain changing the filterGain parameter in the « SHG2_Filters_Fast.cfg » config files and then clicking on « Reload Filters » to send the new gain.

How to lock the MZ interferometer? How to change the pump power? 

Starting point : on the VPM webpage (https://pc-calva16.lal.in2p3.fr:8090/main.html), “Locking” section, the “MZ” process is started  

1. Open the MZ dataDisplay config and look at the MZ_DC signal. 
2. If you want to change pump power, without locking MZ (with a dump inside the MZ): 
   1. Turn the lambda/2 waveplate between the MZ and the polarizing cube (named “pump power”) to adjust the voltage on the MZ photodiode.  
   2. Note the calibration: pump power just before the OPO (mW) = 40.66 * MZ_DC (V) 
3. If you want to lock the MZ: 
   1. Check that you see a sinusoidal error signal on the MZ_DC channel, if not, check that there is no dump inside the MZ, if still not, this is probably a misalignment of the MZ… 
   2. Turn the lambda/2 waveplate between the MZ and the polarizing cube (named “pump power”) to adjust the voltage on the MZ photodiode so that the voltage value you want to lock on is on the sinusoid (better far to the maxima and minima) 
   3. Using the “MZ_Set_Value” button, select the voltage you want to lock one 
   4. Using the “MZ_DC_sw” button, select the lock thresholds min Thresh01 and max Thresh10 (usually Set_value
      +/-0.1 and power below 0.5V Set_Value+/-0.01) 
   5. Click on the “Apply” button to send the above values to the process. 
   6. Using the “Scan/Lock” button, pass in lock mode (+1) 
   7. If the lock is not so stable you can adjust the gain changing the filterGain parameter in the « MZ_Filters.cfg » config files and then clicking on « Reload Filters » to send the new gain

How to lock the OPO cavity? 

On the VPM webpage (https://pc-calva16.lal.in2p3.fr:8090/main.html), “Locking” section, “OPO” process started 

1. Open the OPO dataDisplay config 
2. Check that the OPO_DC signal is correct (low high order modes) and error signal OPO_err_sig looks like a PDH (maybe deformed if high power) centered on 0 (if not centered on 0 see point 6.) 
3. Zoom on the OPO_err_sig signal in a rising slope of OPO_PZT_dac and check the OPO_DC value in mW for which the OPO_err_sig signal enter the linear region. This is the Thresh01 value to put using the “OPO_DC_sw” button (no conversion needed). Note also the Thresh10 to stop locking, you can choose is close to the out of resonance power. 
4. Using the “Scan/Lock” button, pass in lock mode (+1) 
5. If the lock is not so stable you can adjust the gain changing the filterGain parameter in the « OPO_Filters.cfg » config files and then clicking on « Reload Filters » to send the new gain
6. If you are at high power (with huge Kerr effect), you may have to add an offset to the error signal using the “ErrSig_Offset” button. Sometimes you have to place the “0” of the error signal at the “flat” part of the error signal and adjust the locking threshold using “OPO_DC_sw” button to start locking in the “new linear” region.

How to lock the OPO at co-resonance? 

1. Send only the pump and seed beam into Ferrarix 
2. Open the OPO dataDisplay config 
3. Look at the pump beam in reflection of the OPO (OPO_DC on dataDisplay) and the seed beam in transmission of the OPO cavity (Thorlabs photodiode OPO_seed or, if Thorlabs removed, HD_PD3 or HD_PD4) 
4. Start the communication with the OPO translation stage (see How to communicate with the OPO translation stage) 
5. If the IR seed beam and the green pump beam resonances are far from each other, act on the crystal position by steps of 50 or 100.  
6. When you are close to co-resonance, adjust the OPO scanning ramp around the seed resonance (amplitude 0.5V and offset carefully chosen). Check that you still see fundamental mode peaks on both seed and pump. 
7. Lock the OPO cavity (see How to lock the OPO cavity?) 
8. Adjust finely the crystal position to maximize the seed power (and thus optimize the co-resonance) by steps of 5
   

Note: During the fine tuning of the crystal position, you may lose the seed resonance (after moving, power goes to 0), just unlock the OPO, check where the co-resonance is in the PZT ramp and adjust the PZT ramp offset to get back the seed/pump co-resonance in the PZT range and relock.

How to do the PLL? 

1. Open the PyRPL GUI (see how to open the PyRPL GUI). 
2. Send the pump and MCL beams into the Ferrarix tank. 
3. Check on the spectrumanalyzer that you see the green beatnote on in1 (else see You don’t see the green beatnote on the RedPitaya). 
4. On the lockbox toolbox, click on “Go to this stage” on stage 0. If the green beat note is below -50 dB, you need to increase the pump power, see « How to lock the MZ interferometer? How to change the pump power?« 
5. Adjust the temperature offset on stage 0 (by small 0.2e-2 steps, you need to click on ‘Go to this stage’ to validate the new offset) to have the main green beatnote just below 40 MHz (but not too close to 40MHz, let’s say between 32 and 38 MHz). 
6. Click on “Go to this stage” on stage 1. The green main beatnote should be locked on 40 MHz. Then click on “Go to this stage” on stage 2.
   

Note: If it locks on another value, maybe the temperature offset in stage 0 is too low and you are on the mirror side of the spectrum (if the offset is below –0.07, value that can change after stop/start of laser heads). If you can’t see green beatnote below 40 MHz with offset higher than –0.05, go back to the low offset you found and try to change the gain factor sign in stage 1. If now it locks at 40 MHz, don’t forget to change also the gain factor sign in stage 2. Else, good luck! 

How to adjust the homodyne photodiodes electronic offset?

On the VPM webpage (https://pc-calva16.lal.in2p3.fr:8090/main.html), “Locking” section, “SqzPhaseCCL” process started 

1. Open the HD_offset dataDisplay config.
2. If the channels are not centered on 0, using the Offsets buttons of the « SqzPhaseCCL » process, write down a 0 offset.
3. Then Stop/Start the dataDisplay window and look at the offset value measured with the 1D distribution
4. Using the Offsets buttons of the « SqzPhaseCCL » process, write down the correct offset to have 1D distribution centered on 0.

Note that the offset could change in a few hour basis.

How to measure the fringe visibility? 

On the VPM webpage (https://pc-calva16.lal.in2p3.fr:8090/main.html), “Locking” section, “SqzPhaseCCL” process started 

1. Heat the OPO crystal to 50 deg to ensure no nonlinear gain (see How to control the OPO temperature?) 
2. Lock the OPO cavity at co-resonance (see How to lock the OPO cavity at co-resonance?) 
3. Note the power on the homodyne photodiodes (HD_PD3/PD4_DC) with only the seed beam on them. 
4. Put only the LO beam on the homodyne photodiodes and add a neutral density in front of its fiber input. Adjust its orientation so that the LO power on the homodyne photodiodes is almost the same than the one of the seed. 
5. Send both LO and seed beams to the homodyne photodiodes and scan the CCL piezo at 1Hz using the “Ramp” button of the “SqzPhaseCCL” process. You should see fringes. If not check the beam superposition with the card. 
6. The visibility is (Pmax-Pmin)/(Pmax+Pmin)
   

How to lock the squeezing ellipse?

On the VPM webpage (https://pc-calva16.lal.in2p3.fr:8090/main.html), “Locking” section, “SqzPhaseCCL” process started 

1. Lock the OPO at co-resonance with nonlinear gain (OPO crystal at 30 deg) 
2. Lock the PLL. 
3. Open the CCL dataDisplay config. 
4. Send the MCL and LO beam inside the Ferrarix tank 
5. Scan the CCL piezo at 1Hz using the “Ramp” button. You should see fringes on the CCL_RF_20MHz_I channel. 
6. Using the “Scan/Lock” button, pass in lock mode (+1) 
7. If the lock is not so stable you can adjust the gain changing the filterGain parameter in the « CCL_Filters.cfg » config files (or the poles and zeros) and then clicking on « Reload Filters » to send the new gain
8. You can change the squeezing phase using the CCL_RF_20MHz_phi button (in rad)
   

How to do a squeezing measurement?

1. Heat the OPO crystal at 30 deg (or other temperature if found better) 
2. Lock the MZ at 1.3 V (almost 55 mW of pump on the OPO) 
3. Without any beam entering Ferrarix, do the dark noise spectrum on HD_sub_FFT channel. 
4. Send only the LO inside Ferrarix, do the LO spectrum on HD_sub_FFT channel (that should be shot noise limited at least at high frequency) 
5. Lock the OPO at co-resonance using seed beam 
6. Lock the PLL 
7. Lock the CCL phase at fixed phase. 
8. Do the spectrum on HD_sub_FFT channel.  
9. Change the CCL phase and do a new spectrum on HD_sub_FFT channel. The shot noise level should be different for CCL phase = 0, 0.78 and 1.57.

How to lock the local control?

The local control strategy is not fully operational. It has been tested with adjusted gain for M3 but not for M2. Moreover, the Damping_Angles and Damping filters works properly while the LC filter does not improve the lock.

On the VPM webpage (https://pc-calva16.lal.in2p3.fr:8090/main.html), the  procedure uses “DAC Driving” section, “Coils” and « DAC_LC » processes and on « Locking » section, « LC » process.

There could still be bugs on it. And there will be things to change. Thus there is no written procedure yet.

How to find the diagonalization matrix for the local control?

On the VPM webpage (https://pc-calva16.lal.in2p3.fr:8090/main.html),  “DAC Driving” section, “Coils” and « DAC_LC » processes started and on « Locking » section, « LC » and « Diagonalisation » processes started.

The procedure for M2 and M3 is the same, just replacing M2 by M3 in the names. Maybe there could be some adjustement on this procedure.

1. On a dataDisplay look at the M2_TX, M2_TY and M2_Z FFT on 20s and a averaging of 5.
2. Using the « M2_G2 » button of the « Diagonalisation » process, put a gain bias of 0.002 on G2. This will send a sinusoidal signal at 9 Hz on coil 2 of mirror M2.
3. Then using the « M2_Gi » buttons with i from 3 to 5, one by one, put gain bias close to 0.002 on the coil i you want to determined and adjust this bias value and sign so that you almost remove the movement in TX and have a pure TY movement. Write down the optimum Gi value
4. Then, geometrically, to have pure Ty:
   • G2 and G3 should be of same sign
   • G4 and G5 should be of same sign
   • G2 and G4 should be of opposite sign
   • G3 and G5 should be of opposite sign
5. If it is not the case you have to carrefully choose the sign of the gain in the « Coil » proces config, « ## Driving », « # Gain of the coils », numerical value of the « M2_Ji_err » sum channels. For instance for M3, we measured sign(G2) = – sign (G3) = – sign(G4) and sign(G5) = – sign(G3) = – sign(G4). Thus G3 and G5 are cable in the opposite way than G2 and G4 and we put a minus sign on the M3_J3_err and M3_J5_err sum factor.
6. Finally to determine the gain value (factor in the « M2_Ji_err » sum channels) use the « diagonalisation.py » code.

Debugging 

The pump power (at max power with the MZ lambda/2, MZ blocked) is low (below 1.5 V on the MZ photodiode) or the SHGMain lock is unstable/take more time to relock when delocked (same for auxiliary green and SHG2)

1. Pass the SHGMain process in scan mode (Scan/Lock = -1) 
2. Look on dataDisplay at SHG_DC TIME channel. If the max peak is below 5 V, and order 1 is high, you will have to realign the main laser on the SHG1 cavity. 
3. For that, unplug the white BNC cable on the photodiode in transmission of SHG1, and plug the oscilloscope BNC cable on it. 
4. Check that the second channel on the oscilloscope is plugged on the HV/100 channel of the HV amplifier driving SHG1. 
5. Check that the trigger on the oscilloscope is done on the driving ramp of SHG1 
6. Check that the « measure » on the scope is on the photodiode channel and show the max value 
7. Scan the SHG1 cavity at 10 HZ using the SHGMain process (Ramp button -> frequency = 10) 
8. Use the last IR steering mirror (which is a dichroic plate transmitting green) before the SHG1 cavity to recover 8 V on the fundamental peak (often you just have to sligthly turn the x-axis screw – the bottom screw moving the beam horizontally) 
9. Replug the white BNC cable on the photodiode to see the spectrum on dataDisplay SHG_DC channel. 
10. Scan the cavity at 1 Hz 
11. Relock the SHG1 cavity (SHGMain process => Scan/Lock = 1)

If the problem is on the auxiliary green beam production, the procedure is the same, remplacing: 

• SHG1 by SHG2, 
• SHGMain process by SHG2process 
• maximum value of the peak of the order of 4 V on the oscilloscope 
  

The SHG1/SHG2/OPO cavity lock is unstable, the DC spectrum when scanning is good, but the error signal is strange 

1. Look at the SHG_sample/SHG2_sample/OPO_sample FFTs and clock_SHG_sample/clock_SHG2_sample/clock_OPO_sample (these signals are the direct FFTs of the RF signals received on the demodulation board) 
2. The noise floor should be of the order of 10^-7, if it comes up to 10^-4 – 10^-5, there should be a problem on the demodulation channel. You have to reconfig the corresponding demodulation mezzanine. For that: 
   1. On the VPM webpage (https://pc-calva16.lal.in2p3.fr:8090/main.html), “Data Acquisition” section, open the DaqBox user interface (nut on the left of the DaqBox process) 
   2. Click on the reconfigure button (circular arrow) of the mezzanine that have problem.  To know on which mezzanine the channel you find to have problem is, open the DaqBox configuration file and search for the name of the channel 
3. If the noise floor is correct, check that you can see the modulation frequency used (12.4 MHz for SHG1 and OPO and 50 MHz for SHG2) is the two sample channels of your cavity. If it not visible on one of the channels, check the cabling. If it is not visible on both channels, check the generator.
   

The pump spectrum reflected by the OPO cavity (OPO_DC channel on dataDisplay) is not good (with many high order mode)

You could have to realign the pump beam on the OPO cavity (never move the OPO cavity mirrors) 

1. Pass the OPO process in scan mode (Scan/Lock = -1)  
2. If available, place a Thorlabs photodiode in reflection of the OPO cavity to have a signal on an oscilloscope (it is easier with the scope than with dataDisplay, mainly due to 2s delay on dataDisplay between action and signal observed), otherwise, use the OPO_DC channel of dataDisplay and do small movement with 2s waiting time between each movement. 
3. If using the scope, send the actuation ramp from the DAC channel to a second channel of the scope to place a trigger on it and scan the OPO at 10 Hz (OPO_process, ramp button -> frequency = 10) on its full range (ramp button -> offset = 0.2, amplitude = 2.8). 
4. Use the two pump steering mirrors (place the red knobs on their screws) before the OPO (the second one is after MCL/pump recombination) to recover a proper spectrum. You may have to do tilt/tip in x and y direction. 
5. If you have to much trouble and see no more spectrum, you can add a camera in transmission of the OPO and use the end of the standard alignment procedure.  
6. When the spectrum is OK, rescan the cavity at 0.5 Hz and lock it.
   

You don’t see the green beatnote on the RedPitaya

• If you don’t see the IR beatnote: 

1. On the PyRPL GUI, “lockbox” toolbox, click on “Go to this stage” in stage 0. You should see peaks moving and then stabilizing.  
2. If not, check that the IR photodiode is correctly aligned with both IR beams superposed on it (use an oscilloscope) 

• If you don’t see the green beatnote: 

1. Check that the IR main beatnote is below 30 MHz (the green beatnote should be 2 times the frequency of the IR beatnote) 
2. Check that both pump and MCL beams are sent to the Ferrarix tank. 
3. Look on dataDisplay at the SERV12_PD1_DC channel (the green PLL photodiode) and check that you can see both beam on the photodiode, if not, check that there is nothing inside Ferrarix blocking the beams. If not, you will have to do realignment: 
   1.  First add a mirror after the pump/MCL recombination on the PLL side to check the superposition in near field, far field. 
   2.  If you see that the beams are not superposed in near field and far field, check whether the pump beam is well aligned on the OPO (if not realign it and recheck the beatnote) 
   3.  If the beams seem superposed in near field and far field, the pump is well aligned on the OPO, you will have to move the MCL beam. For that, first check that the seed and LO beams are well balanced on the homodyne photodiodes (if not, see The LO/seed beam is not well balanced on the homodyne photodiodes). Then look both at the PLL spectrum and the HD_PD3 and HD_PD4 DC signals while scanning the OPO cavity (you should see the MCL spectrum in reflection of the OPO cavity on both photodiodes with the same amount of power and a clear fundamental mode deeper than noise and other modes).
      

The LO beam is not well balanced on the homodyne photodiodes

1. On the VPM webpage (https://pc-calva16.lal.in2p3.fr:8090/main.html?), “Data Acquisition” section, open the DaqBox user interface (nut on the left of the DaqBox process), check that both photodiodes are rearmed (SERV18_PD0 and SERV18_PD1) and SERV18_PD0 has its shutter open. 
2. Realign the LO beam on the photodiodes using its mirrors (you have to maximize power on both homodyne photodiodes).

The seed beam is not well balanced on the homodyne photodiodes

1. On the VPM webpage (https://pc-calva16.lal.in2p3.fr:8090/main.html?), “Data Acquisition” section, open the DaqBox user interface (nut on the left of the DaqBox process), check that both photodiodes are rearmed (SERV18_PD0 and SERV18_PD1) and SERV18_PD0 has its shutter open. 
2. Do a fringe visibility measurement. 
3. Optimize the fringe visibility by acting on the seed mirrors just before LO/seed recombination.

The MCL beam is not well balanced on the homodyne photodiodes

1. On the VPM webpage (https://pc-calva16.lal.in2p3.fr:8090/main.html?), “Data Acquisition” section, open the DaqBox user interface (nut on the left of the DaqBox process), check that both photodiodes are rearmed (SERV18_PD0 and SERV18_PD1) and SERV18_PD0 has its shutter open. 
2. Check that the seed beam is well balanced on the homodyne photodiodes. If not see « The seed beam is not well balanced on the honmodyne photodiodes« . If yes, you may have to slightly realign the MCL looking at the PLL signal (ensuring to keep the beatnote) and the MCL signal reflected by the OPO cavity (like HD_add channel on dataDisplay, ensuring a correct scan with little high order mode)
   

I click on reload config on another process than the one I wanted

For any process except temperature ones, this should not be a problem. If it is TempControl_SHG2 or TempControl_OPO, just ensure to lock back the temperature to the desired value.

The camera always displays « not connected » after few seconds

1. If this is only « the camera is in live mode but no image », just click 2 times on the « live » button. 
2. Block the usb cable under the computer to ensure no movement at the connection on the computer side 
3. If possible try to use no cable extension 
4. Disconnect and reconnect on the camera and computer side (and extension if there is one) 
5. Change computer/camera!
   

There are strange bi-peaks on the whole SHG2 spectrum

This may be due to the PLL no more locked and stopped at a temperature offset close to a mode jump. On the PyRPL GUI change the auxiliary laser temperature offset. 

The in-vacuum photodiodes don’t have full voltage and/or strange FFT

On the VPM webpage (https://pc-calva16.lal.in2p3.fr:8090/main.html?), “Data Acquisition” section, open the DaqBox user interface (nut on the left of the DaqBox process), check that the photodiodes are rearmed and SERV18_PD0 has its shutter open.