KAGRA Logbook

AEL (General)

dan.chen - 16:36 Wednesday 21 May 2025 (33854)

Pre-Installation Test Report of High Power Coil Driver for ETMX

With: Shingo Hido, Hiroshi Takaba, Takahumi Ushiba, Masakazu Aoumi

Summary

We conducted a transfer function and noise measurement of the High Power Coil Driver (HPCD) unit that is planned to be newly installed at ETMX. While most channels behaved as expected, Channel 3 showed an unexpected transfer function response. We concluded that this unit requires repair before installation.

Background

The current coil driver for the test mass (TM) of ETMX is a High Power Coil Driver (HPCD) with a beacon input port.
To reduce actuation noise during observation, we plan to use an HPCD with an additional output resistor and the beacon input port, if possible.
The configuration measured on May 19 was the currently installed HPCD + beacon input.
Today, we tested another HPCD unit (with beacon input) that will be modified by adding an output resistor and made switchable.
This measurement was a pre-installation test of that unit.

Measurement Target

HPCD with beacon input port (S/N: S2315284)
Design document: S2315284 Design Doc

Measurement Setup

Equipment: MokuLab Blue with a special DC power supply
Input voltage: 0.3 V

Measurements

1. Transfer Function: Beacon Input to Coil Out

The measured transfer function looks reasonable.

2. Transfer Function: "Test In" to "Coil Out"

Nothing was connected to "BIO 2", so no de-whitening filter was active.
Channel 3 showed unusual behavior — a pole(?) was observed around several hundred Hz. (Attached figure)

3. Transfer Function: Test In to Coil Out with BIO Setting Change

The BIO setting was changed from “2” (no filter) to “4” (three-stage filter) to observe differences. Ch 1, 2, and 4 look fine.

4. Noise Measurement

The output was amplified by a factor of 100 using an SR560.
The "Test In" inputs were terminated with 50 Ω. But the beacon input port was open.
Background noise was measured with the input of the SR560 terminated, which was lower than the main measurement data by several tens.
The BIO setting was “1”, meaning no de-whitening filter.
The noise floor, accounting for the SR560 gain, is estimated to be 500 nV/√Hz. Attached one is one of the measured raw data showning about 50 uV/√Hz. (We may need to check whether the electrical GND connection was suitable.)

Results and Notes

The circuit of Channel 2 needs to be fixed before installation at the X-end.
The noise floor, accounting for the SR560 gain, is estimated to be 500 nV/√Hz.
(Measurement data could not be extracted from the iPad at this time...)

Images attached to this report

Comments to this report:

dan.chen - 7:08 Thursday 22 May 2025 (33856)

Pictures

dan.chen - 5:30 Friday 23 May 2025 (33875)

Date: May 22, 2025

Summary

Following the pre-installation test of the HPCD (S2315284) for ETMX, we conducted additional investigation to identify the cause of the irregular behavior seen in Channel 3. A flipped 24 kHz input cable was found and fixed, and a malfunction of the relay switch in the final stage of Channel 3 was identified. However, these issues alone do not fully explain the previously observed abnormal transfer function. As a result, we plan to migrate the 24 kHz injection circuit to a new HPCD unit (S2305285), with the filter configuration updated to 3–30 Hz. The migration will begin next week. Before the migration, we will verify the operation of the new circuit.

Details

One of the 24 kHz input cables was found to be flipped and has been corrected.
The final-stage relay switch in Channel 3 was confirmed to be malfunctioning.
However, the irregular transfer function observed previously cannot be fully explained by the switch issue alone.
Therefore, we have decided to migrate the 24 kHz injection circuit to another HPCD unit, S2305285.
During this migration, the internal filter configuration will also be updated to 3–30 Hz.
The migration work is scheduled to begin next week.
Prior to that, a functional check of the new HPCD circuit (S2305285) will be conducted.

dan.chen - 6:42 Friday 23 May 2025 (33876)

With Shingo Hido

> we will verify the operation of the new circuit.
We measured the TF of the new circuit (S2305285) with changing the relay switchs.
These results look fine.

The file names are showing the measured condition including channel number and filter configurations.
Filter configuration: for instance, "110" means the first and the second filters were ON, but the third one = the last one was OFF, counting from the input side of the circuit.

Images attached to this comment

33876_1747949939_S2315285_diff_ch1_000_20250523_055053_Screenshot.png

33876_1747949940_S2315285_diff_ch1_100_20250523_055307_Screenshot.png

33876_1747949940_S2315285_diff_ch1_110_20250523_055540_Screenshot.png

33876_1747949940_S2315285_diff_ch1_111_20250523_055852_Screenshot.png

33876_1747949940_S2315285_diff_ch2_100_20250523_060628_Screenshot.png

33876_1747949941_S2315285_diff_ch2_110_20250523_060743_Screenshot.png

33876_1747949941_S2315285_diff_ch2_000_20250523_060317_Screenshot.png

33876_1747949941_S2315285_diff_ch2_111_20250523_060851_Screenshot.png

33876_1747949941_S2315285_diff_ch3_000_20250523_061208_Screenshot.png

33876_1747949941_S2315285_diff_ch3_100_20250523_061323_Screenshot.png

33876_1747949942_S2315285_diff_ch3_110_20250523_061441_Screenshot.png

33876_1747949942_S2315285_diff_ch3_111_20250523_061556_Screenshot.png

33876_1747949942_S2315285_diff_ch4_000_20250523_061940_Screenshot.png

33876_1747949942_S2315285_diff_ch4_100_20250523_062102_Screenshot.png

33876_1747949942_S2315285_diff_ch4_110_20250523_062207_Screenshot.png

33876_1747949943_S2315285_diff_ch4_111_20250523_062317_Screenshot.png

dan.chen - 15:35 Friday 23 May 2025 (33881)

Aomi-san replaced the capacitors in the circuit to modify it into a 3–30 Hz de-whitening filter. Following this modification, we measured the transfer function of the updated circuit.

The filter section consists of three stages, where filter 1 is closest to the input side of the circuit and filter 3 is closest to the output side. The ON/OFF states of these filters were encoded in the file names as a 3-bit binary string (filter1–filter3), and measurements were conducted for the following configurations:

000 (all filters OFF)
100 (filter 1 ON only)
110 (filters 1 and 2 ON)
111 (all filters ON)
010 (filter 2 ON only)
001 (filter 3 ON only)

The measured transfer functions for all configurations were consistent with expectations, confirming that the filter stages function correctly after the modification.

Images attached to this comment