(Saito, Miyakawa)
Abstract:
We measured the PZT efficiency of the current IR laser using 15.0MHz sidebands. The result is 1.54 MHz/V.
Detail:
Laser PZT efficiency was measured many times in KAGRA, but the discrepancies were quite large.
- 1.48MHz/V: 2017/6/6 by Nakano, using 51.75MHz sidebands with FSS, https://klog.icrr.u-tokyo.ac.jp/osl/index.php?r=2853
- 3MHz/V: 2020/2/18 by Aso, using efficiency of AOM of ALS-X: https://klog.icrr.u-tokyo.ac.jp/osl/?r=13012
- 5.56MHz/V: 2021/5/8 by Yamamoto, using OLTF of IMC, https://klog.icrr.u-tokyo.ac.jp/osl/?r=16678
- 4.50MHz/V: 2024/4/11 by Tanaka, using OLTF of IMC, https://klog.icrr.u-tokyo.ac.jp/osl/?r=29194
- 3.018 +/- 0.037MHz/V, 2024/6/14 by Yamamoto, using 15MHz sideband with PMC : https://klog.icrr.u-tokyo.ac.jp/osl/?r=29898
1.48MHz/V in No.1 sounds reasonable as a nominal NPRO, but we are not sure that the measured laser was the same as No. 2-4. The use of the OLTF in No. 3 and 4 resulted in a large efficiency. Using the AOM of No.2 seemed to have a big error. No.5 is basically the same way as No.1, but there is a possibility that they measured the higher-order resonance instead of the sideband peak, as reported in the log.
This time, the method is similar to No.5. We applied a 0.5 Hz, ±10 V triangular-wave injection to the laser PZT using K1:IMC-SERVO_OFS_SLOWOUT_CALI_EXC. An important thing was that we confirmed the measured peaks were really caused by the 15.0MHz sideband. See Figs 1 and 2.
- In Fig. 1, we thought first that the peaks at the cursors were the sideband, but at least one of them was fake. At this stage, the modulation index was already increased (we increased LO amplitude from 8.5dBm to 19dBm). The modulation index can be roughly estimated to be ~0.01 order.
- In Fig. 2, you can see sidebands clearly with a more increased modulation index (we increased the LO amplitude from 8.5dBm to 19dBm, and a +20 dB RF amplifier). Even second-order sidebands can be seen. The modulation index can be roughly estimated to be ~0.1 order.
Also, you can see many excess peaks, which are likely higher-order modes at the 1% level of carrier height. No.5 had measured sidebands at the1% level of carrier peaks, so there was a possibility that they measured higher-order mode peaks inside the 15MHz sideband and resulted in higher PZT efficiency. Of course, the laser changed with neoLASE, but a typical NPRO should have 1MHz/V efficiency, so 3MHz/V or more sounds too high.
Anyway, we measured 10 frequency spacings between the carrier and one side of the 15MHz sideband, for both forward and backward scans, as an example in Fig.3.
- Forward (10 times average): 1.71+/-0.148 MHz/V
- Backward (10 times average): 1.36+/-0.326 MHz/V
I put the errors, but they are probably meaningless since the systematic error between directions is too large. This difference may come from the response of back and forth, but it is not clear whether it comes from the optical response or the intrinsic PZT response, or another reason. Anyway, it is probably reasonable to take the average of them since we are using such a PZT with an anti-symmetric response, so the result is 1.54 MHz/V. This value is close to the No.1 and sounds reasonable as NPRO.
