Looking at the entry studying the 50Hz FF, I have noticed that on March 14 bruco is able to partially explain the bump at 125Hz. In the list of channels coherent with h(t) around 125Hz, SWEB_B8_DC is the largest contributor (although still with a small coherence of 15%).

Figure 1 shows the bruco projection for SWEB_B8_DC with a noise projected bump around 125Hz, a factor ~3 below the actual bump in h(t).

Figure 2 compares in purple March 14 when the bump is large to March 13 when it is small (or non existent). Clearly when the bump is visible in h(t) it also becomes visible on B8, and the two become slightly coherent.

Figure 3. shows that these bumps are not visible on SNEB_B7_DC and that the level of sensitivity of B7_DC is the same. There is also no coherence between B7_DC and h(t) in 100-200Hz band.