Pulse cables for XFEL

Transcription

1 Pulse cables for XFEL

2 Why Pulse cable? Basic idea came from the TESLA LC About 600 modulators have to be installed This large number of units should NOT be installed into the tunnel. Reliability Maintainability Space would have been to small for the 800 GeV upgrade The modulators shall be installed in service halls at ground level The pulse transformer have to be installed inside the tunnel The energy has to be transported into the tunnel via these cables

3 Overview TESLA

4 View of TESLA tunnel Pulse cables

5 Tunnel Design 30 cm 80 cm 190 cm 50 cm 520 cm 125 cm 90 cm 210 cm 440 cm Figure 2. Main LINAC, Damping Ring & Electronic Station Pulse cables

6 View of the XFEL Tunnel

7 XFEL Technical data of the modulator for TESLA max. klystron gun voltage: 120 kv max. klystron gun current: 140 A Primary Voltage 10 kv Primary Current 1680 A high voltage pulse length: 1.7 ms pulse repetition frequency: 10 Hz max. pulse power: 16,8 MW average power: app. 300 kw Number of Modulators in TESLA: 38

8 Requirements No significant delay for the pulses No distortion of the waveform No electro smog ejected into the tunnel Low losses Radiation hardness Good fire resistance due to the large amount of burning material Very high reliability Good price

9 Questions to be solved How to the terminations and splices look like? How to transport the cables? How to install the cables?

10 Pulse cable for XFEL Construction drawing

11 Photo of pulse cable

12 Termination

13 Terminations

14 General comments The cable is called a pulse cable due to the fact that the modulator produces pulses However when looking at the time of 1,7 ms and the rise times of more than 100 µs for the current, the cable is working in the DC region.

15 No significant delay for the pulses this would decrease efficiency Simulation showed that a decrease in Z 0 lead to a better pulse behavior. Too achieve this two possibilities exist. : Increase the diameter of the inner conductor The construction is not easy Splices in the tunnel since only short pieces can be manufactured Put cable in parallel Good solution since standard manufacturing process

16 Low losses The losses are mainly determined by the resistive part of the copper There is no significant damping of the semiconductor layers as seen in high frequency/short pulse applications Semi conducting layers can be used Again standard manufacturing for high voltage cables can be used Skin effects have to be taken into account By paralleling the effect of the skin effect reduced since

17 No reflections on the cables The pulse traveling along the cables will see the pulse transformer at one end. With the stray inductance the cable is not adapted and the pulse would be reflected. Therefore an adaptation network consisting of resistor and a capacitor has to be introduced for a broad band adaptation. For the four cables a value of 6.6 Ohm and 2.5µF give good results

18 RC-Adaptation-Network

19 No electro smog into the tunnel For the cable a triaxial design was chosen. The fields will stay within the cable between the inner and middle layer Additionally an outer shield consisting of a enclosing foil was introduced Measurements for EMC are still to be performed

20 Radiation hardness By the choice of the material there are no problems to be expected. XLPE (cross linked Poly ethylene) has a radiation hardness of 5 * 10 7 rad Additionally the cables will be beneath concrete

21 Good fire resistance Due to the large number of cables and the amount of XLPE a large fire load is introduced into the tunnel To counteract to this the construction was carefully chosen. The cable is specified according to IEC Cat. A The outer shield is built as a closed foil building a barrier for the fire and the burning material Two layers of FRNC (Flame retardant non corrosive) materials are introduced. This was proven by a fire propagation test

22 Help on construction, calculation M. Filtz, Institut für theoretische Elektrotechnik, TU Berlin Dr. Pfeiffer, Draka Multimedia Cable Dr. Fricke, Siemens W. Buchwald, Kaiser Kabel, später Nexans Ansoft, Supply of modells for cable simulation

23 With these requirements the cable was ordered The cable was purchased in a European wide bid for a tender Several companies who are able to manufacture the cable. The cable can be produced with standard production procedures. The order was given to HVT, Germany Dielectric Sciences, US Essex, UK SAGEM, France

24 Let s come to the existing cable

25 Tests with the cable in factory Routine tests as for any other high voltage cable Type test for fire propagation test IEC category A Smoke density test

26 Preparation for fire propagation test

27 Cable being fired duration 40 min

28 Cable after fire propagation test self extinguished after 20 min

29 Cable after fire propagation test Allowed propagation: 2.5 m Reached propagation: 0.8 m

30 Cable after smoke density test

31 Destruction during smoke density test

32 Transport of the cable drum with 3000 m

33 Installed pulse cable

34 tricky parts of installation

35 Electrical Test First tests with modulator at low voltage (max. 5kV) The commissioning of the test was extremely successful Tests on test load to check the short circuit behaviour

36 Simulation model for modulator, cable and klystron incl. arcs 200m CTRL := S1 100m R2 C := 1.4m V0 := 8k CH := XYVaristor.VAL R3 40u L2 # Varistor D3 C1 D1 80u A AM1 R1 160 A + VM1 V LTRA1 = J R := 119u L := 50n G := 0 C := 1n LEN := 1.5k A AM2 Ca 2.5u + V Ra 6.6 VM2 TFR1P 2 W 1 LM := 0.8 LS1 := 0.194m LS2 := 0.1u S3 CTRL := S S2 Dk CTRL := S + VM4 V Dklys1 Rkurz 100m N Kl ystron N U_brenn KLYSTRON AM3 XY XYVaristor FILE := resistor_ssh XYVaristor.mdx L1 330u C_Bounc 2m 700 TH1 CTRL := zuenden D2 t_1>0.1

37 Current in modulator and transformer 1.2k AM1.I [A] A M2.I [A ] 1k 0.8k 0.6k 0.4k 0.2k 0-0.2k 0 0.5m 1m 1.5m 2m 2.5m 3m 3.5m t [s ]

38 Simulated klystron voltage at low voltage 0.1Meg VM4.V [V] KLYSTRON17.KLYSTR_ 80k 60k 40k 20k 0-10k 0 0.5m 1m 1.5m 2m 2.5m 3m 3.5m t [s ]

39 Measurement U Klystron I Klystron I Modulator U Modulator

40 Test load Resistive value of 8.9 Ohms Inductance of 200 µf Ignitron to produce a short circuit at any moment during the pulse Crowbar test with a wire to prove the allowed amount of Joule deposited

41 Test lead for 10 kv, Crowbar test

42 Tests with test load

43 Tests with Test load Short circuit

44 Goal reached 128 kv

45 Next Tests Test with klystron to full voltage Test with full pulse length High power test with 10 Hz, 128 kv, 5 MW Klystron Measurement of the electro magnetic radiation Long term tests with modulator

46 The pulse cables behave in the way it is foreseen. The functionality is now proven for XFEL