Charge measurement and mitigation for the main test masses of the GEO 600 gravitational wave observatory

Abstract

Spurious charging of the test masses in gravitational wave interferometers is a well-known problem. Typically, concern arises due to the possibility of increased thermal noise due to a lowering of the quality factor of modes of the test-mass suspension, or due to the potential for increased displacement noise arising from charge migration on the surface of the test masses. Recent experience gained at the GEO 600 gravitational wave detector has highlighted an additional problem. GEO 600 uses electrostatic actuators to control the longitudinal position of the main test masses. The presence of charge on the test masses is shown to strongly affect the performance of the electrostatic actuators. This paper reports on a measurement scheme whereby the charge state of the GEO 600 test masses can be measured using the electrostatic actuators. The resulting measurements are expressed in terms of an effective bias voltage on the electrostatic actuators. We also describe attempts to remove the charge from the test masses and we show that the use of UV illumination was the most successful. Using UV illumination we were able to discharge and re-charge the test masses.PACS numbers: 95.55.Ym, 04.80.Nn

Sensitivity of the Advanced LIGO detectors at the beginning of gravitational wave astronomy

“…7). This mismatch in actuation strength is caused by electrostatic charge [66], which is distributed on the test masses in a non-uniform manner and has a slow time dependence.…”

Interaction between Stray Electrostatic Fields and a Charged Free-Falling Test Mass

“…Limiting stray forces on a test mass (TM) is crucial for precise experimental gravitation, from gravitational wave (GW) observation [1][2][3] to tests of the equivalence principle [4], short range gravity [5,6], and relativistic gyroscope precession [7,8]. In all these experiments, electrostatic force noise is cited as a precision-limiting effect.…”

“…Here, N = 4s 2 ∆a for the 4 X/Y faces andl 2 is a mean square armlength, withl 2 = 2 s s/2 0 x 2 dx = s 2 12 for ∆a ≪ s 2 . This is approximate in the case of only a At left is a LISA prototype TM (cube sidelength s = 46 mm), with the outlines of the sensor X electrode footprints, with surface area Ax = 529 mm 2 and a on-center semi-separation Rx = 10.75 mm and (dashed) guard ring surfaces.…”

Characterization of electrostatic actuators for suspended mirror control with modulated bias

“…The effects of this term were already observed on a similar set-up [9], and some techniques for its mitigation were already developed [10].…”

“…All the measurements were repeated in DC bias (f M = 0) and alternate bias (f M = 100 Hz) conditions. In particular, for each set of measurements in DC bias, also the sign of the voltage applied to the HV amplifier was changed to enhance the effects due to the stray charges on the mass, following the suggestions given in [10]. This change can be formally taken into account by changing the sign of G in equation (7).…”