One hypothesis for the source of excess vibration noise on the PSL table from the crystal chiller circuit (here) is that the ~2 mm aperture in the quick-connect hose fittings on the 5/8” ID hoses causes a turbulent jet of water that, especially if bubbles are present, modulates the flow of water. Figure 1, lower plot, shows that a quick-connect is about two orders of magnitude more noisy than a barbed barrel connector for the 10 - 15 l/m flow through the crystal circuit of the crystal chiller. Figure 2 has a photograph of the setup.

To see if the noise from the quick connect might be loud enough to explain the noise from the whole circuit, we clamped the quick connect setup used for the above measurement to the outside manifold of the crystal box (setup photo shown in Figure 2). The upper plot in Figure 1 compares the nominal noise to the noise from this setup. The peaks produced in this way are relatively large but don’t match the peak from nominal flow through the box. However, the momentum transfer to the table for our test is much different than in the actual setup.

I think the results are suggestive enough that it would be worth an experiment where we replace the two quick connects at the crystal box manifold with barbed connectors. To further improve the flow, we suggest that we use the capped threaded holes for this in order to make one less 90 degree turn. Figure 2 also has a photograph illustrating the proposed modification.

Robert S., Michael R., Rick S.