3. 0.7 Hz peak
This was eventually found to be caused by magnetic stirrers (little magnets turning in beakers) from labs on the floor B1, above us. At some point (separate post below), they had added more stirrers in the room immediately above the MEG and the artefact was much worse. After investigating shielding options, they eventually agreed to remove the stirrers from that room.
Older observations and investigations:
Almost always present since late 2020, even nights and weekends. This is a sharp peak with frequency almost always 0.7 Hz, though it has also been observed at 0.8 and 0.9 Hz. A similar peak was observed in some older noise recordings (before the system shutdown Mar to Aug 2020), but not the majority. The peak frequency and amplitude does not vary with the recording sampling rate (not aliased).
It’s confirmed to be coming from outside the shielded room (larger with door open) and is not due to anything that can be turned off while collecting (most devices and computers in the lab).
Raw spectrum example
3rd gradient (CTF noise reduction)
SSP analysis of empty room recording in Brainstorm, showing the topography of two components that seem to fully capture the artefact. The component time courses further show the constancy of the artefact and a pattern of beats lasting roughly 50 seconds, so a narrow frequency peak indeed.
