J. Oberling, F. Clara

Summary: We think the 9.5Hz peak is coming from the new flow meters installed as part of the 2021 PSL laser upgrade.

Today we continued hunting for the 9.5Hz comb triplet that Ansel had tracked to the PSL back in May.  Robert subsequently used a temporary magnetometer and found a strong 9.5Hz feature in the neoLASE LVEA Control Box.  To try to isolate the source of this 9.5Hz peak we shut down the PSL in a controlled manner and watched the peak on the magnetometer.  The order of the shutdown:

• PSL Stabilization systems (PMC/ISS/FSS)
• Laser
• Chiller

There was no change in the peak observed during the PSL shutdown, so we next turned off the neoLASE Control Box; the peak clearly disappeared with the Control Box off.  Our next step was to check if the peak was being fed to the Control Box via something plugged into it.  We turned the Control Box back on and the peak returned (as expected), then we unplugged the cable that feeds the flow/temperature sensor signals to the Control Box (via the PSL Sensor Interface Chassis (SIC), which feeds the Control Box with several monitor signals (PDs, power meters, flow meters) and the system shutters).  Lo and behold, the peak was much smaller, sometimes disappearing entirely (see attached pictures; first is with the PSL off but the flow meters plugged in, second is with the flow meters unplugged).  We next took the top off of the SIC and placed the magnetometer near where the flow meter signals both come into and exit the SIC, and did not see the peak.  Fil then placed the magnetometer on the flow sensor connectors themselves (the flow sensor cables route to a pig tail on the back of the SIC, which goes from an individual M12 connector for each flow meter into a DB25 connector to plug into the SIC) and the 9.5Hz peak was clearly visible again. If we unplugged the main flow meter cable from the back of the Control Box the peak again disappeard; this makes sense as the Control Box supplies the 24 VDC the flow meters need, and unplugging the cable removes power from the flow meters.  We then put the top back on the SIC and placed the magnetometer back in its spot directly underneath the Control Box and plugged the flow meters back in; the peak again returned.  Fil disconnected the individual M12 connectors for the flow meters and the peak again disappeared.  We then plugged each individual flow meter in to observe how the peak behaved (I did not take pictures of this, unfortunately).  With 1 flow meter plugged in the peak was visible but much smaller; with 2 plugged in the peak was even smaller (???); with three plugged in the peak was back to its usual magnitude.  With this odd behavior we suspect a grounding issue somewhere.  To finish things off we unplugged the other items feeding signals into the Control Box (shutters, PDs/power meters, NPRO, Diode Box A, Diode Box B), one at a time, to observe how/if the peak changed; we observed no change in the peak during this process, the only thing that changed the peak with the Control Box turned on was unplugging the flow meter cable.  We think this fairly definitively tags the flow meters as the source for the 9.5Hz peak seen in Robert's magnetometer, and likely the source of the comb triplet seen around this frequency.

We are now looking through documentation to see how we can mitigate this.  One option is to tie the grounds of all 3 flow meters together and see if this changes the peak; this requires running long cables from each flow meter to tie them all at the same spot.  We also plan on using the spare Control Box to check into how the internal grounding of the Beckhoff terminal is done and see if there's any improvements we can make there.  Another option is to isolate power to the flow meters; currently they are powered by the Control Box, but we could power them separtely and see if that changes the peak.  Investigation continues, more to come.

The PSL is now back ON, and the PMC and FSS RefCav are locked; I'm letting the system warm up for a bit before locking the ISS.  This closes LHO WP 12031.

Edit to add: When unplugging the flow meters individually via their M12 connectors, we did try each of the 3 flow meters by themselves; the peak was the same regardless of which individual flow meter was plugged in.  Things got weird when we had 2 plugged in at the same time (the peak was smaller than with any of the individual flow meters plugged in, hence why we suspect a grounding issue somewhere), and the peak returned to its starting magnitude when all 3 were plugged in at the same time.