Data from deployments of the FairfieldNodal three‐component nodes were used to analyze a persistently observed noise signal. The noise signal is most prominent in the 20‐ to 40‐Hz range but has been observed anywhere in the 10‐ to 100‐Hz range. Interestingly, the signal is affected by air temperature and moves to higher frequencies in colder temperatures. Nodes that were deployed in seismic vaults directly on flat concrete slabs do not show the noise signal, and nodes that were buried in the ground or covered in snow show a significant decrease in the noise signal. This suggests that whatever is causing this signal may be mitigated by better coupling to the ground. Spectral analysis of hydrothermal tremor in the Upper Geyser Basin, Yellowstone, suggests this noise signal can interfere with the true ground vibration and can impede the ability to accurately characterize these signals. It is our recommendation to always bury the nodes if it is possible to reduce this noise signal that can interfere with natural signals of interest in a similar frequency band. In addition, tests to better estimate the best gain setting were done, and results show that above 12 dB, the waveforms of teleseismic events on the three‐component nodes are very similar, suggesting that there is no advantage to using a gain setting higher than 18 dB for recording teleseismic events. If background noise is of interest in addition to teleseismic events, we see no adverse effects on the waveforms of teleseismic events using the max gain setting of 36 dB.