The performance of ncal during low-noise operation has previously been analyzed for rotor frequencies below about 15 Hz (56683). However, ncal can spin up to 30 Hz or so and I wanted to understand the noise implications of that. It seems that once the rotor frequency exceeds about 20 Hz, broadband excess noise appears in DARM. The attached plot shows ncal spinning at 28.6 Hz from 19:49:00 to 20:16:00 Z on 2023–02–01. A comparison of DARM at a quiet time is also shown, along with the BSC9 X accelerometer spectrum. During some of this time Jenne and I were also at EX adjusting the test mass camera focus, but this level of broadband excess noise appears independent of that.
Before unclamping the assembly from the BSC (67342), we (Dhruva and I) spun NCAL at a few different frequencies at the following times on 2023-02-09:
- 30.5 Hz: 19:40:00 – 20:10:00 Z
- 18.5 Hz: 20:17:00 – 20:34:00 Z
- 16.2 Hz: 20:38:00 – 20:47:00 Z
Note that the NCAL-induced BSC acceleration does not change much for the three rotor frequencies shown. The noise in DARM is not coherent with the acceleration channel. Robert suggested looking for an NCAL-induced signal in the magnetometers and EFM but we could not find one.
To slightly correct the above alog, the excess noise in DARM kicks in when NCAL spins above roughly 15 Hz. This is substantiated by the pdf attachment of a series of short spectra while the NCAL takes a few minutes to spin up to 30 Hz. To exactly nail down where this extra noise kicks in, it would be advantageous to take a few minutes of NCAL data at a discrete set of rotor frequencies.
Please note that the NCal has not been carefully balanced and aligned since it was disassembled and rebuilt at the observatory in 2019. This is why the NCal team limited our injections to rotation rates below 15 Hz. There could be misalignments in the bearing axes which will increase the torque the motor needs to apply to maintain the rotation rate. This may cause noise (both vibrations and EMI) as well as increase the wear on the bearings and motor. As such I would strongly suggest limiting the amount of time the NCal is spun above 15 Hz.
The noise seen here is not indicative of what we'd expect during a normal NCal calibration injection.
Additionally, the NCal should not be moved unless necessary. The positioning at the mm-scale needs to be maintained to utilize our previous surveying.