Differences between revisions 8 and 9
Revision 8 as of 2020-03-06 01:31:32
Size: 15607
Comment:
Revision 9 as of 2020-03-06 01:50:26
Size: 10455
Comment:
Deletions are marked like this. Additions are marked like this.
Line 62: Line 62:
  * We need to consider about the grounding or floating. How about the optical table, chambers, and so on.
Line 64: Line 63:
  * For making the GND, we already have some hole at the center area. But due to the mine water, we couldn't judge whether we can use as a GND or not.
Line 66: Line 64:
  * We need to take into account the long GND from Mozumi entrance. One GND come from the Mozumi entrance.
Line 68: Line 65:
  * The instrument for measuring the ground resistance is not ready. After arriving to Kamioka, we want to check the GND.
Line 70: Line 66:
   * There may be instrument in the KEK room. Kimura-san will come to Kamioka next week. At this time, Kimura-san and Yoshimura-san will have a discussion(or if possible measurement)
Line 72: Line 67:
   * Is there the value measured by Kamioka mine company? Maybe Yes. But we don't have value. It is better to ask to Kamioka mine company.
Line 75: Line 69:
  * We used the Mozumi GND in KAGRA now.
Line 77: Line 70:
   * Mozumi entrance -> Yend -> Center area -> Xend
Line 79: Line 71:
  * The guideline is to make the GND for the Low noise
Line 81: Line 72:
  * It is better to prepare multiple GND for analog circuit,
Line 83: Line 73:
  * This is a information from SK(Shiozawa-san). They use SK tank as a GND for circuit. In case of KAGRA, we can have a choice to use the chambers as GND. But, there is large vacuum pumps, the voltage of the chamber is noisy. There is choice to use water, but it is better to prepare earth GND to each area.
Line 85: Line 74:
  * From the experience by Aso-san, the OMC chamber seems noisy. Chamber may be an antenna for higher frequency. Or large power instruments may make the large magnetic field and make coupling. As a result, there is potential difference
Line 87: Line 75:
  * From the experience by Tomaru-san, during the noise investigation by PCal, he tried to use beam duct as GND, but the result is not so good.
Line 89: Line 76:
  * For disconnecting the power instruments and chamber, we should use the insulator(ceramic). There is insulator at the cooler.
Line 91: Line 77:
  * Doing many things is important for search the answer for GND problem.
Line 93: Line 78:
  * Question : Will we prepare the GND of the same potential difference in each area(XYend and center area?). Answer : Maybe no we need local GND.
Line 95: Line 79:
  * In SK, they used AVR(Automatic Voltage Regulator?). Distinguish the low noise instrument from high voltage instruments. They prepared the clean current for the circuits
Line 97: Line 80:
  * From the requirement from the Kamiizumi-san, Each GND for each rack is ideal. It is very good to prepare GND close to the racks.
Line 99: Line 81:
  * We use 18V DC GND as a GND of the racks and circuits.
Line 101: Line 82:
   * Test GND by ALS racks ans IOO rack.
Line 103: Line 83:
  * 18V DC GND connected to the GND of the circuits. circuit GND connected to the chassis. Chassis GND (weakly) connected to the rack with the ear of the circuit. The reason of the weak is that there is painting. 18V DC GND is not connect to the rack completely. If you found the loose screw, the GND problem may not solve.
Line 105: Line 84:
  * There is the effect by the situation of the GND connection in each rack.
Line 107: Line 85:
  * Cabling issue : The surface of the Dsub cable may be connected to circuit GND(default). Of-caurse only one side was connected to GND. Avoiding the ground loop. If we used the gender changer, the GND connection may affect.
Line 109: Line 86:
   * Correct connection of the Dsub cable(surface GND, 5pin GND), the GND of the circuit and cable may not cause the problem.
Line 111: Line 87:
  * If we have a cabling diagram, we can set the correct GND connection by changing the jumper pins inside the circuits.
Line 113: Line 88:
  * Question by Kimura-san, did we use independent GND for analog part and digital part? Answer : No. If we use the same GND for the IO chassis, oscillator(used in the LVDT), the circuit became noisy.
Line 115: Line 89:
  * Is there GND pin in the SCSI cable? Not clear...
Line 117: Line 90:
  * We may manage the connection of the GND by the AA and AI chassis. They have so many jumper pins. We can choose the connection between the digital part, power supply part, cable part and chassis part.
Line 119: Line 91:
  * From the experience of the iLIGO, we use the differential cables, basically. That may cause the not so bad GND problem.
Line 121: Line 92:
  * We use the differential cables with correct way, make GND for circuits and racks, and avoid the ground loop.
Line 123: Line 93:
  * Can we eliminate the BNC cable?
Line 125: Line 94:
  * We will use the independent GND for AC instrument(oscillator, power for PEM, and so on).
Line 127: Line 95:
  * We will make a manual for correct GND connection. We can judge the correct GND connection by checking the interferometer signal. The interferometer may answer correct or not.
Line 129: Line 96:
  * We will reduce the unused and unconnected cables.
Line 131: Line 97:
  * The GND of the PSL optical table is connected with circuit GND. We should float all optical tables?
Line 133: Line 98:
  * We should understand the structure of the DSub cable. 5pin, 8pin and connector.
Line 135: Line 99:
  * We will make a rule and check lists
Line 137: Line 100:
   * The configuration of the jumper pins in the circuit
Line 139: Line 101:
   * Cabling diagram
Line 141: Line 102:
  * We may use the Altium by manage the cabling diagram
Line 143: Line 103:
  * Check the cable diagram
Line 145: Line 104:
   * There are mysterious cables(where are the connected? What purpose?)
Line 147: Line 105:
   * Rule of cable tags
Line 149: Line 106:
    * XX to YY in the tag
Line 151: Line 107:
    * Fake tags (previous-used tags)
  * Type-B optical table was connected to the large structures.
  * ADC for ground monitor,
  * Is there any electronic problem?
  * Ground for chamber

Minutes of the Kick-off the KAGRA GND meeting

  • 2020/03/05(Thu.) 10:00-11:00(JST)
  • Kickoff meeting meeting (Japanese, English)
  • Web meeting with Zoom2 (Zoom2)

  • Attend : Yokozawa, Washimi, Oshino, Fabian, Shimode, Kamiizumi, Tomura, Uchiyama, Aso, Miyoki, Miyakawa, Yoshimura, Furuta, Takahashi(ICRR), Yamada, Tomaru, Kimura, Ushiba
  • Minutes of the discussion
    • We need to consider about the grounding or floating. How about the optical table, chambers, and so on.
    • For making the GND, we already have some hole at the center area. But due to the mine water, we couldn't judge whether we can use as a GND or not.
    • We need to take into account the long GND from Mozumi entrance. One GND come from the Mozumi entrance.
    • The instrument for measuring the ground resistance is not ready. After arriving to Kamioka, we want to check the GND.
      • There may be instrument in the KEK room. Kimura-san will come to Kamioka next week. At this time, Kimura-san and Yoshimura-san will have a discussion(or if possible measurement)
      • Is there the value measured by Kamioka mine company? Maybe Yes. But we don't have value. It is better to ask to Kamioka mine company.
    • We used the Mozumi GND in KAGRA now.
      • Mozumi entrance -> Yend -> Center area -> Xend

    • The guideline is to make the GND for the Low noise
    • It is better to prepare multiple GND for analog circuit,
    • This is a information from SK(Shiozawa-san). They use SK tank as a GND for circuit. In case of KAGRA, we can have a choice to use the chambers as GND. But, there is large vacuum pumps, the voltage of the chamber is noisy. There is choice to use water, but it is better to prepare earth GND to each area.
    • From the experience by Aso-san, the OMC chamber seems noisy. Chamber may be an antenna for higher frequency. Or large power instruments may make the large magnetic field and make coupling. As a result, there is potential difference
    • From the experience by Tomaru-san, during the noise investigation by PCal, he tried to use beam duct as GND, but the result is not so good.
    • For disconnecting the power instruments and chamber, we should use the insulator(ceramic). There is insulator at the cooler.
    • Doing many things is important for search the answer for GND problem.
    • Question : Will we prepare the GND of the same potential difference in each area(XYend and center area?). Answer : Maybe no we need local GND.
    • In SK, they used AVR(Automatic Voltage Regulator?). Distinguish the low noise instrument from high voltage instruments. They prepared the clean current for the circuits
    • From the requirement from the Kamiizumi-san, Each GND for each rack is ideal. It is very good to prepare GND close to the racks.
    • We use 18V DC GND as a GND of the racks and circuits.
      • Test GND by ALS racks ans IOO rack.
    • 18V DC GND connected to the GND of the circuits. circuit GND connected to the chassis. Chassis GND (weakly) connected to the rack with the ear of the circuit. The reason of the weak is that there is painting. 18V DC GND is not connect to the rack completely. If you found the loose screw, the GND problem may not solve.
    • There is the effect by the situation of the GND connection in each rack.
    • Cabling issue : The surface of the Dsub cable may be connected to circuit GND(default). Of-caurse only one side was connected to GND. Avoiding the ground loop. If we used the gender changer, the GND connection may affect.
      • Correct connection of the Dsub cable(surface GND, 5pin GND), the GND of the circuit and cable may not cause the problem.
    • If we have a cabling diagram, we can set the correct GND connection by changing the jumper pins inside the circuits.
    • Question by Kimura-san, did we use independent GND for analog part and digital part? Answer : No. If we use the same GND for the IO chassis, oscillator(used in the LVDT), the circuit became noisy.
    • Is there GND pin in the SCSI cable? Not clear...
    • We may manage the connection of the GND by the AA and AI chassis. They have so many jumper pins. We can choose the connection between the digital part, power supply part, cable part and chassis part.
    • From the experience of the iLIGO, we use the differential cables, basically. That may cause the not so bad GND problem.
    • We use the differential cables with correct way, make GND for circuits and racks, and avoid the ground loop.
    • Can we eliminate the BNC cable?
    • We will use the independent GND for AC instrument(oscillator, power for PEM, and so on).
    • We will make a manual for correct GND connection. We can judge the correct GND connection by checking the interferometer signal. The interferometer may answer correct or not.
    • We will reduce the unused and unconnected cables.
    • The GND of the PSL optical table is connected with circuit GND. We should float all optical tables?
    • We should understand the structure of the DSub cable. 5pin, 8pin and connector.
    • We will make a rule and check lists
      • The configuration of the jumper pins in the circuit
      • Cabling diagram
    • We may use the Altium by manage the cabling diagram
    • Check the cable diagram
      • There are mysterious cables(where are the connected? What purpose?)
      • Rule of cable tags
        • XX to YY in the tag
        • Fake tags (previous-used tags)
    • Type-B optical table was connected to the large structures.
    • ADC for ground monitor,
    • Is there any electronic problem?
    • Ground for chamber
  • 議論メモ(日本語)
    • groundingとfroatingを分ける、各定盤はどのようにするとか、チェンバーはどうするのかとか、いろいろな場面が考えられる。
    • 地面に穴があるが、水とかで結局GNDにできないかもしれない。。
    • 茂住のこうぐちからもって来ることも検討
    • 接地抵抗測定用の装置の手配がまだなので、到着後に測定をしてみる。
      • KEKのものがあるかもしれないので、来週木村さんと相談する
      • 神岡鉱業さんが設置抵抗の数値を持っている?
      • AC用、Low voltage用のものはあるはずだが、手元の資料がない。参考になると思う。
    • 現状は茂住から持ってきている
      • 茂住口->Yend->CS->Xend

    • Low noise用の特別なGNDを作成するのがbetter
    • アナログ回路専用、複数個あるとさらに良い
    • SK(塩澤さん)では、でかいタンクそのものがGNDとなっている。KAGRAで大きいチェンバーにつけてしまうと、ポンプなどが設置されているとかの関係で、逆にノイズを大きくしてしまう。水も考えたが、各エリアに準備をした方が良い。
    • OMCチェンバの周りはnoisy、高周波はアンテナとなってしまう?大きなパワーが磁場とカップリングを起こして電位差が生じたりする。
    • PCalの実験でノイズレベルを下げる際に、tubeに落としたりもしたが、そんなに良い結果を得られなかった。
    • cryocoolerとチェンバーが繋がらないように絶縁体(セラミックなど)を入れることもある。coolerには入っている。
    • とにかく、やってみないとわからない。
    • XYendとCSで全く同じ電位にする必要があるか?localで統一すれば良い。
    • SKではAVRを使用している。そして、他の電源ハイボルなどから切り離している。とにかく綺麗な電流を回路用に使用している。
    • 理想的には、ラック単位でGNDが欲しい。現実的には、ある程度固まっているラックで一つとか
    • 現在は、計算機室からのDC電源の中点をGNDとしている。
      • ALS2ラックとIOOあたりを落としてみたりしている。
    • 中点が回路基板内のGNDとつながっている。それがシャーシに落ちている。そして、耳からラックにつながるが、塗装があるので、ラックと中点とが完全につながっているわけではない。ネジが緩んでいたり。
    • ラックとどこかにつながっているかでその影響を受けてしまっている。
    • Dsubのケーブルに落とすようにしている。片側だけ落としている。loopにならないように。gendar changerなどで変な繋げ方をしていると、それは定かではない。
      • Dsub 5pinを正当な使い方をしていれば、正しくGND配線ができるはずである。
    • どのように繋ぐような設計になっているかがわかれば、jumper pinでちゃんと設定できる。
    • ラック内でアナログ部とデジタル部のGNDを区別しているか?していない。IO chassisや発信機、がらっくにある。それが回路とつながったりしてしまうと汚される。
    • SCSIケーブルにGNDをつながってしまうようなものがあるのか?はっきりとはわからない。
    • アナログ部とデジタル部のGNDの管理はAA, AI chassisでできるだろうと思われている。
    • differentialで信号処理をしていることがだいぶ功を奏している。
    • differentialを徹底する、GNDを作成する、GND loopを作らないを徹底する。
    • BNDを排除できないか?
    • AC駆動している、発信機などのものをどうするのか?できれば繋ぎたくないので、別GNDにすることを考えなくてはいけない。
    • 正しいGNDかどうかは、干渉計の信号を見て判断するべき。そのためのマニュアル管理を作る。
    • 18Vの電源の先が最終的にどこにつながっているかわからないものもたくさんある。
    • PSLの定盤が回路のGNDとつながっている。
    • 5番ピンや8番ピンを徹底すれば、なんとかなる場合がある。
    • ルールを作成する、チェックリストを作成する、
      • 回路のジャンパーの確認
      • 配線図
    • アルティウム?などの配線図を管理するソフトウェアをする。
    • 配線を現場で見てわかるように記述する
      • 謎ケーブルが多い
      • ケーブルタグのテンプレート
        • XX to YY と書くべき
        • 昔使っていた嘘のタグなど

KAGRA/Subgroups/PEM/GNDmeeting/200305/Minute (last edited 2020-03-06 02:02:28 by TakaakiYokozawa)