1. 08/24/2004 Stefan Ballmer, MIT / LIGO Hanford G040430-00-I 1 Hanford 4km Recent improvements and currently limiting noise sources Stefan Ballmer Massachusetts Institute of Technology LIGO Hanford Observatory

2. 08/24/2004 Stefan Ballmer, MIT / LIGO Hanford G040430-00-I 2 Improvements since S3

3. 08/24/2004 Stefan Ballmer, MIT / LIGO Hanford G040430-00-I 3 Improvements below 100Hz Eliminated noise sources: Auxiliary loops coupling  Coupling correction  High loop gain in auxiliary loops Non-linearity in mixer due to 2-omega  Additional notches in antisymmetric port photodiodes

4. 08/24/2004 Stefan Ballmer, MIT / LIGO Hanford G040430-00-I 4 Auxiliary loops coupling correction Controlling beam splitter and recycling mirror But error signal is shot noise limited above 50Hz  We are adding noise Trick: feed back to end mirrors too!  Coupling reduced by 40 Laser antisymmetric signal symmetric signal pickoff signal transmitted power photodiodes LyLy LxLx lxlx lyly

5. 08/24/2004 Stefan Ballmer, MIT / LIGO Hanford G040430-00-I 5 Auxiliary loops high bandwidth This factor of 40 allowed us to  Back off cut-off filters  Increase bandwidth of those loops The higher gain reduced the loop’s error signal and thus reduced the up conversion

6. 08/24/2004 Stefan Ballmer, MIT / LIGO Hanford G040430-00-I 6 2-omega at antisymmetric port At AS port: Light dominated by sidebands  Beat at 49MHz is dominant RF signal  Anticipated in RF photodiode design, but notch was not deep enough  Non-linear up- conversion in mixer Lower Sideband Upper Sideband Residual Carrier 2x f mod =49MHz

7. 08/24/2004 Stefan Ballmer, MIT / LIGO Hanford G040430-00-I 7 Photodiode fix Added additional 20dB notch in RF amplifier stage  huge effect on differential arm error signal (=GW signal)

8. 08/24/2004 Stefan Ballmer, MIT / LIGO Hanford G040430-00-I 8 Limiting noise below 100Hz Now And how to fix them Actuation electronics  Coil driver modification (passive filter at output) in preparation Auxiliary loops (MICH_CTRL / PRC_CTRL)  Sqrt(2) improvement easy with additional diode  Other approaches possible (use reflected signal) SRD thermal noise prediction was conservative

9. 08/24/2004 Stefan Ballmer, MIT / LIGO Hanford G040430-00-I 9 Limiting noise, whole band Now