[Gerardo, Koji]
As reported before [LHO ALOG 39359], we confirmed the disqualification of the installed OFI in terms of the optical isolation. The OFI was pulled out and place in the optics lab. [LHO ALOG 39366].
Yesterday, the optical performance of the OFI was tested without any adjustment and then after the performance optimization. The reasonable performance was recovered. It's hard to believe that we can keep this 43dB isolation forever under environmental fluctuations.
Next steps: Today, some of the measurement will be repeated for confirmation of the stability. In addition, a dry run of the in-situ isolation measurement will be done in the optics lab to pin down the reference numbers and accuracy/precision.
Here the detailed testing procedure can be found in E1700350.
| LHO OFI 2017/11/13 Before the adjustment |
LHO OFI 2017/11/13 After the adjustment |
|
| Transmission From the input port to the output port |
0.973 +/- 0.002 (2.7% loss) |
0.967 +/- 0.002 (3.3% loss) |
| Back-scatter isolation S-pol From the input port to the squeezer port |
1960 +/- 20 ppm | 1300 +/- 10 ppm |
| TFP AR reflection From the input port to the squeezer port |
N/A | 890 +/- 10 ppm |
| Isolation From the output port to the input port |
1800 +/- 200 ppm ( = 27.4 +/- 0.5 dB isolation) |
52 +/- 13 ppm ( = 43 +/- 1 dB isolation) |
| Back-scattering From the input port to the input port |
2.3 +/- 0.4 ppm | 3.0 +/- 0.5 ppm |
| Squeezer transmission From the squeezer port to the input port |
0.974 +/- 0.002 (2.6% loss) |
0.978 +/- 0.002 (2.2% loss) |
The transmission was checked with a Thorlabs power meter with an integrated sphere. The numbers were 0.972+/-0.004 and 0.971+/-0.006. But the systematics is large (+/-1%) everytime the number is measured.
Other OFI Performance to compare:
LLO OFI Performance summary [LLO ALOG 25788]
Previous LHO OFI measurement in the opt lab [LHO ALOG 39276]
What has been done:
- A noticeable thing was that the half-wave plate (HWP) was not tightly fastened in the holder. It is a notorious holder (Attached Pic) that the HWP can still jiggle in the holder even the screw is fastened with a normal torque. This was the issue the initial OFI of the LLO (actually... this unit). This time, the screw was fastened much tighter with an Allen key.
- Other alignment was basically fine. The angle of the thin film polarizer (TFP) was a bit tweaked to minimize the backscatter iso, but it seems that this is the minimum. This may mean that the isolation was optimized by sacrificing the polarization purity of the transmitted beam through the faraday rotator and half wave plate.
Can you please note the temperature in the optics lab? Temperature changes of a few degrees will limit us to <40dB of isolation.
I don't have a thermometer right now. We will measure it.
Degradation of the isolation due to the temperature dependence of Verdet effect for TGG
When the isolation was optimized (< -40 dB), the backscatter isolation (S-pol of the transmission from FR+HWP) was 1300 ppm. Assume this was caused by the deviation of the rotation angle at the faraday rotator from the nominal 45deg.
Backward beam
Faraday: phi_0 = 45 + delta [deg]
HWP: phi_H = 45 - delta [deg]
Total: phi_0 + phi_H = 90 [deg]
Forward beam
Total: phi_0 - phi_H = 2 delta
==> S-pol power: sin[2 delta/180*pi]^2 == 1300ppm
==> delta ~ 1deg
Temp dependence of the faraday rotation
FR had the rotation of 45deg + delta at T=298. According to JOSA B 9 (1992) 1912, the Verdet effect of TGG crystal has the temperature dependence of the follwoing form: phi = a (6751/T - 6.968) [deg]. a is a constant. T is the temperature in K. In our case, phi = phi_0 at T=298. This gives us a=2.935. ==> dphi/dT @T=298 = -0.223 [deg/K]
This will cause additional dphi = dphi/dT x dT rotation on the backward beam
Temp dependence of the isolation
So, the isolation is given by the following expression:
sin[dphi/180*pi]^2 = sin[-0.223 dT/180*pi]^2
This will hit 40dB (1e-4) at dT = 2.6 K and 30dB (1e-3) at dT= 8.1 K.
Temp dependence of the backscatter-isolation (and transmission)
This 2.6K deviation cause the backscatter isolation changing {+830ppm,-630ppm}.
For the 8.1K deviation, it changes {+3300ppm, -1300ppm}. The sign depends on the sign of delta.
To be checked:
- Temp dependence of the retardation of the half wave plate
- Temp dependence of the other optical parameters
Attached are the temperature trends from inside the optics lab (lab1 channel) and the vacuum prep lab (lab2 channel), data taken from the dust monitors.
