Izumi, Kokeyama, Nakano, Enomoto
See also 4946 and comments to this log.
= Summary =
We repeated measurements several times and got numbers somewhat close to the designed asymmetry length (3.3298 m), but the result does not match with each other and they seem to be not reproducible.
List of 3 measurement results:
3.4(5) m
4.5(7) m
3.9(6) m
= Detail =
We performed the measurement 3 times. We measured transfer function from IMC-SERVO_BTEST2_CON_IN to CAL-CS_PROC_C00_DISPLACEMENT, with excitaion signal injected to VIS-MCE_TM_TEST_L.
-- 1st measurement (last night)
I put the results of TF measurement below.
Note that this measurement was done just after last night's optical gain measurement by Yamamoto-san (4957), so the optical gain at that time can be estimated to be 7.44e10 cnt/m,
though CAL-CS_PROC_C00_DISPLACEMENT is calculated with 1.07e11 cnt/m; this difference should be taken into account.
And the fluctuation of the optical gain was roughly ~ 15 %.
| Freq [Hz] | Mag. [dB] | Phase [deg] | Coh. | EXC [cnts] |
| 100 | -26.18 | -3.7 | 0.993 | 300 |
| 190 | -26.18 | -7.2 | 0.998 | 600 |
| 310 | -26.86 | -15.2 | 0.997 | 2000 |
| 80 | -25.83 | -0.3 | 0.977 | 200 |
| 80 | -27.09 | -0.7 | 0.987 | 200 |
| 100 | -26.70 | -4.0 | 0.990 | 300 |
| 310 | -26.09 | -15.0 | 0.985 | 2000 |
| 190 | -26.77 | -9.3 | 0.985 | 600 |
Simple average of the above list gives us the magnitude of TF of 0.048 nm/cnt.
Considering cnt to Hz conversion factor is 5.71e-4 V/cnt * 10.02 MHz/V, and magnitude of TF should be multiplied by optical gain correction factor of 1.07e11 / 7.44e10,
we will have lasym = (δlMICH / δf) * flaser = 3.4(5) m --> consistent with the designed value
-- 2nd measurement (this morning)
I put the results of TF measurement below.
Note that this measurement was done just after this morning's optical gain measurement by Yamamoto-san (4957), so the optical gain at that time can be estimated to be 8.98e10 cnt/m,
though CAL-CS_PROC_C00_DISPLACEMENT is calculated with 1.07e11 cnt/m; this difference should be taken into account.
And the fluctuation of the optical gain was roughly ~ 15 %.
| Freq [Hz] | Mag. [dB] | Phase [deg] | Coh. | EXC [cnts] |
| 100 | -22.42 | -2.9 | 0.996 | 300 |
| 190 | -22.34 | -7.0 | 0.998 | 600 |
| 310 | -23.29 | -13.6 | 0.997 | 2000 |
| 80 | -23.13 | 1.6 | 0.972 | 200 |
| 80 | -21.78 | 3.2 | 0.959 | 200 |
| 310 | -22.43 | -12.4 | 0.997 | 2000 |
| 190 | -22.51 | -6.2 | 0.998 | 600 |
| 100 | -21.48 | -1.9 | 0.999 | 300 |
Simple average of the above list gives us the magnitude of TF of 0.076 nm/cnt.
Considering cnt to Hz conversion factor is 5.71e-4 V/cnt * 10.02 MHz/V, and magnitude of TF should be multiplied by optical gain correction factor of 1.07e11 / 8.98e10,
we will have lasym = (δlMICH / δf) * flaser = 4.5(7) m --> INCONSISTENT with the designed value
-- 3rd measurement (this noon)
We did the basically same measurement to above two but measured TF with swept sine method.
Text file containing the results is attached.
Averaged magnitude of the TF was 0.066 nm/cnt.
Most likely estimation of the optical gain at that time is 8.98e10 cnt/m.
Considering cnt to Hz conversion factor is 5.71e-4 V/cnt * 10.02 MHz/V, and magnitude of TF should be multiplied by optical gain correction factor of 1.07e11 / 8.98e10,
we will have lasym = (δlMICH / δf) * flaser = 3.9(6) m --> consistent with the designed value
= Notes =
- there were high fluctuation in the optical gain, so this was source of large uncertainty.
- all the three results were systematically larger than the designed value.
- there might be (non-linear) spurious coupling from laser frequency to MICH error signal, since there were sidebands to the main peak of excitation whose phase seem random, and so it is very difficult to get high coherence if we adopt frequency bin size larger than 0.01 Hz.
I personally doubt the following path: frequency shift at IMC --> AM after IMC --> AM comes out in MICH signal by coupling with misalignment; the amount of AM is proportional to DC power and DC power is determined by misalignment.
