jeffrey.bartlett@LIGO.ORG - posted 11:02, Friday 06 July 2018 - last comment - 09:24, Tuesday 10 July 2018(42787)
Move Chiller Bypass from PSL Enclosure to Chiller Room
Peter K., Robert s., Jeff B.
This morning Peter and Jeff moved the chiller bypass from inside the PSL enclosure to the Chiller Room. We fully closed off the bypass valve on the PSL manifold (inside the PSL enclosure) and opened up the bypass between the Crystal Chiller supply and return lines. Next we adjusted the bypass valve in the Chiller Room and the pressure regulators on the Front End, Shutter, and 70W amp circuits to bring them back to their proper flow rates.
After making the changes, Robert ran a noise check. The noise is OK at lower frequencies, and has improved at higher frequencies. There is still some work needed to lower the noise in the 400 to 900 hertz bands.
| Before Changes | |
|---|---|
| P1 - Manifold In | 4.6 Bar |
| P2 - Manifold Out | 3.5 Bar |
| Crystal Chiller Flow Rate | 16.6 L/m |
| Front End Flow | 1.7 L/m |
| Front End Pressure | 64 psi |
| Shutter Flow | 1.5 L/m |
| Shutter Pressure | 63 psi |
| 70W Amp Flow | 2.1 L/m |
| 70W Amp Pressure | 55 psi |
| After Changes | |
|---|---|
| P1 - Manifold In | 4.8 Bar |
| P2 - Manifold Out | 3.8 Bar |
| Crystal Chiller Flow | 21.8 L/m |
| Front End Flow | 1.6 L/m |
| Front End Pressure | 68 psi |
| Shutter Flow | 1.4 L/m |
| Shutter Pressure | 65 psi |
| 70W Amp Flow | 2.0 L/m |
| 70W Amp Pressure | 60 psi |
12 hour plot of various PSL cooling system temperatures, pressures, and flows; from before and after the change
The figure shows that today's move of the water bypass from inside to outside the PSL reduced the motion of the PSL periscope above 1000 Hz by several. This is the improvement we expected because the water flow noise under the table was coupling acoustically to the table/periscope (https://alog.ligo-wa.caltech.edu/aLOG/index.php?callRep=42087). The red trace from before today's work was already lower (relative to a month ago) in the 100-200 Hz region from recent work straightening the 70W amplifier and dump cooling circuits. The vibration level during O2 is shown in green and the vibration level with the water off is shown in black. The worst of the circuits identified in the above log, the front end cooling circuit, has not been straightened yet, and we hope to get further reduction in the several hundred Hz region when this is done.
Here is a comparison of the periscope accelerometer X-motion, between H1 and L1. The H1 data corresponds to the BLUE curve in Robert's plot. Sorry for the ADC units, but LDV doesn't have a calibration factor capability. The calibration is 6.1 um/sec^2 per ADC count. As Robert notes, there is still some noise reduction work to do.