1. Stefan Hild 1GWADW, Elba, May 2006 Experience with Signal- Recycling in GEO 600 Stefan Hild, AEI Hannover for the GEO-team

2. Stefan Hild 2GWADW, Elba, May 2006 Motivation GEO600 is the 1st large scale GW detector using the advanced technology of Signal-Recycling: During commissioning of Dual-Recycling many (new) problems came up. Some problems are GEO specific, many are generally connected to the Signal-Recycling technique. We learned to cope with many of these new issues. Some of our experience is applicable to future detectors which may use Dual-Recycling.

3. Stefan Hild 3GWADW, Elba, May 2006 Signal-Recycling in short Two main parameters: Bandwidth (of the SR resonance) Tuning (Fourier frequency of the SR resonance) narrowband broadband tuned detuned An additional recycling mirror (MSR) at the dark port allows: enhancing the GW signal shaping the detector response MSR

4. Stefan Hild 4GWADW, Elba, May 2006 Shaping shot noise

5. Stefan Hild 5GWADW, Elba, May 2006 Bandwidth of Signal-Recycling The bandwidth of the Signal-Recycling resonance is determined by the reflectivity of MSR. Shot noise for GEO600 with a light power of 10 kW @ beam splitter

6. Stefan Hild 6GWADW, Elba, May 2006 Etalon  SR-tuning by microscopic position of Etalon  SR-bandwidth by temperature „Jukebox“  Use several conventional mirrors  Time-consuming  Long detctor downtimes Michelson Interferometer  SR-tuning by common-mode  SR-bandwidth by differential-mode Cavity  SR-tuning by common-mode  SR-bandwidth by differential-mode Concepts for a variable bandwidth MSR MSR detector heater

7. Stefan Hild 7GWADW, Elba, May 2006 Demonstration of thermally tunable SR at Garching-Prototype (Keita Kawabe et al, 2003) Transmittance of MSR [%] SR gain

8. Stefan Hild 8GWADW, Elba, May 2006 Tuning of Signal-Recycling The tuning of the Signal-Recycling resonance is determined by the microscopic position of MSR. Shot noise for GEO600 with a light power of 10 kW @ beam splitter

9. Stefan Hild 9GWADW, Elba, May 2006 Locking and tuning

10. Stefan Hild 10GWADW, Elba, May 2006 Lock acquisition in GEO600 Can‘t use the SR sideband (SR 1f) signal for initial lock: strong dependence on varoius parameters (alignment, dark fringe offset) small capture range Actual procedure: 1. Locking to SR 2f at a detuning of 2.2 kHz 2. Switching to MI 2f 3. Switching to the SR sideband signal 4. Tuning the detector in small steps to its operation point.

11. Stefan Hild 11GWADW, Elba, May 2006 Sideband picture (RF sidebands in SR cavity) Some sidebands see resonances during downtuning. carrier Comb of equidistant SR resonances SR sidebands MI sidebands Frequency of the light Frequency of the tuning

12. Stefan Hild 12GWADW, Elba, May 2006 Downtuning / Optical transfer function Downtuning: About 70 steps of each 25 Hz (every 400ms) 6 Parameters need to be adjusted: SR frequency SR gain SR phase MI gain MI phase MI autoalignment gain With this method we are able to tune SR to frequencies as low as 250 Hz.

13. Stefan Hild 13GWADW, Elba, May 2006 Jumping to the lower SR sideband and to tuned Signal-Recycling For various reasons we are not able to tune further down to the tuned case and then to the lower SR sideband We can jump to the other sideband (only 2.8 nm for MSR) and to the tuned case (only 1.4 nm) Nominal operating point, SR tuning = 350 Hz, Upper sideband SR tuning = 350 Hz, lower sideband tuned SR

14. Stefan Hild 14GWADW, Elba, May 2006 Kicking MSR Kicking MSR in a controlled way: Fast enough that all other loops can‘t recognize. 4 ms of acceleration and 4 ms of deceleration. Works fine: Jumping to tuned and to the lower SR sideband

15. Stefan Hild 15GWADW, Elba, May 2006 Sideband picture for tuned SR Tuned SR = symmetric sidebands

16. Stefan Hild 16GWADW, Elba, May 2006 Sensitivity on different locking points Sensitivity is identical for the two different locking points.

17. Stefan Hild 17GWADW, Elba, May 2006 SR in operation

18. Stefan Hild 18GWADW, Elba, May 2006 Mode healing Power-Recycled Michelson Dual-Recycled Michelson Future detectors will operate at much higher light levels. Even with thermal compensation beam distortion might be a big problem. Two options: Mode healing and/or output mode cleaner Using Signal recycling provides an increase of intracavity power of about 80%.

19. Stefan Hild 19GWADW, Elba, May 2006 Calibration of a GW detector with SR When using Signal-Recycling (and RF readout) the GW-signal is spread over both quadratures. (The distribution is frequency dependent) You need to carefully choose the demodulation phase. You need to calibrate two signals.

20. Stefan Hild 20GWADW, Elba, May 2006 Combining the two output quadratures You can optimally combine the two calibrated signals to an h(t)-channel. Advantages: h(t) has best GW content at all frequencies data analysists only need to handle a single signal

21. Stefan Hild 21GWADW, Elba, May 2006 Detuned SR complicates various noise couplings and TFs Frequency noise coupling to h(t) Laser amplitude noise coupling to h(t) In a detuned detector TF may become complicated due to interaction and different resonance conditions of various sidebands.

22. Stefan Hild 22GWADW, Elba, May 2006 Sideband picture for detuned SR detuned SR = asymmetric sidebands

23. Stefan Hild 23GWADW, Elba, May 2006 Summary Implementing SR is more than installing an additional mirror. (=> completely different detector) GEO demonstrated reliable operation of SR in a large scale GW detector Demonstration of detuned and tuned SR Advantages of SR Problems connected to SR Shaping the detector response Modehealing More complex system (less intuitive understanding) Complex noise couplings GW signal in both output quadratures

24. Stefan Hild 24GWADW, Elba, May 2006 E n d