EMI Steering Group Report

Transcription

1 TTF EMI/EMC Working Group Meeting 29/05/2006 H. Kapitza (FLA) EMI Steering Group Report 1. Epcos EMC Filter Seminar: On May 4, 2006 there was an EMC Filter Seminar held in the Novotel HH-Schnelsen, organized by the Epcos company, the former passive components branch of Siemens. The event was surprisingly neutral, with some very nice talks and good documentation. Is there a problem to join such events? 2. Next meeting: Monday 10/07/06 at 11:00 here. Focus: Progress on today s topics.

2 TTF EMI/EMC Working Group Meeting 29/05/2006 H. Kapitza (FLA) XFEL Tunnel Grounding (2) Topics: 1. Cross Section Update 2. Floor Inclusion 3. Mockup Tunnel 4. Next Steps

3 XFEL Facility Ground Due to its length >> λ(2 MHz)/10 = 15 m the XFEL facility needs a hybrid ground system where many local single-point grounding areas are connected to a common facility ground. In order to push its impedance as low as possible, the common facility ground (plane) should include as much through-going metal structure as possible (tubes, pipes, cable trays, remesh,...). Since these components will not really form a plane, aim at a mesh size λ(2 MHz)/50 = 3 m.

4

5 Röntgenlaser XFEL Tunnelquerschnitt Modulstrecke Bauliche Planung

6 Vorschlag fuer XTL Bodenplatten T. Hott, 22 Mai 06 Fahrweg Installationsseite Gitter (alle m) - 3t belastbar Kantenschutz Gewindeloch - M6 oder M8

7 XFEL Mockup Tunnel On May 3, 2006 there was a meeting of all parties interested in an XFEL mockup tunnel. Originally I expressed interest in operating a klystron there, fired by a stretched out pulse cable. During the meeting the following became clear: Most parties are interested in logistics checks, with access as freely as possible. Klystron operation will hinder this substantially. The mockup will be rather small, like 30 m long or so. For usage and cost reasons that makes sense. But this is not the place to test stretched out pulse cables. There are better test scenarios for the pulse cables. Currently no EMC use of XFEL mockup tunnel.

8 Next Steps For the following points some understanding and some numbers are still missing. They will be presented next time: 1. Energy radiated from the pulse cables and its absorption by the tunnel mesh. 2. Pulse cable cutoff frequencies and consequences for the cable trays.

9 Umbau TTF2 Injector Rack Area A third hut for the synchronisation laser will be set up (end of July). The rack area will be partly rearranged and enclosed. At the same time an adequate ground plane will be prepared. Its current state is desolate. The power supply situation must be inspected and probably revised (power from two HVen). Meeting with J.Schäfer (MKK) on At the same time remove old TTF1 ground system relics.

10

11

12 EMI shielding for the new Injectorlaser

13 EMI shielding for the new Injectorlaser Right now no special shielding agains EMI, all cable connections are routed through two holes in the wall.

14 EMI shielding for the new Injectorlaser Also the cable connections in the laser hut are not routed ideal to minimize crosstalk

15 Current problems- from outside 230V Mains Current on compresse air

16 Current problems- from outside 230V Mains Current on network cable

17 Current problems- from outside Laser oscillator during maintenance day Laser oscillator during normal run Suspect: Modulator/Klystron

18 Current problems- from inside Pockelscell driver Kabel tray

19 Current problems- from inside PTO direct from photodiode via RG58 Output from ontable LEMO cable, not connected to the photodiode The pockelscell outpus pulses (~4 kv) are visible - crosstalk

20 Shielding - floor Old hut New hut

21 Shielding walls, roof

22 Cable connections routing through wall

23 Cable connections Ethernet Old scheme: Standard twisped pair connection from the switch in the injector area to every device New proposal: One fiber link from the old control room in the laser, a switch "satellite" in the laser hut. But the satellite offers 24 ports, overkill? contact person: Thorsten Witt, Is it possible to use this satellite for both huts? Alternative: Media converters to get fiber links from the Injctor area. Timing Old: New proposal: Fiber optic connection into the hut, timing distribution in the laser hut. RF Old scheme: RF-Cables from the master into the laser hut. New proposal: RF-Cables decoupled by a DC-block into the hut. BIS (temperatures, (door-)interlocks Old scheme: Cables from the devices to the BIS-box in the hut New proposal:

24 Cable connections Motor-Control Old scheme: CAN-Bus controller in the injector area, cables to each motor. New proposal: Beckhoff controller box in the laser-hut. Connected via ethernet. Problems with different hardware in old/new laser? Distance measurement Old scheme: Pick-up connected by cable to the measurement transceiver in the injector area. New proposals: Not urgend needed, to be planned but not connected. Is there a possibility to put a filter in the cable connection? Mains Old scheme: Standard connection New proposal: Use a line filter or transformer. BIC Old scheme: Ground free differential signal, opto coupled inputs New proposal: -

25 Does this effort make sense? The hut is not designed as EMI shield

26 Some links Company for cable connections (Hamburg): Some considerations can be found in the logbook:

27 EMI Sources and Noise in the Laser Hut Efforts in the injector rack area must be coordinated with efforts in the nearby laser hut. Just as the injector rack area, the laser hut has a complete safety ground system but no low impedance high frequency instrumentation ground system. Significant EMI currents were measured on interconnecting 27 MHz reference cables. Add a low impedance instrumentation ground system to this area: Lay down a cm wide, 0.4 mm thick copper conductor under the racks and under the optical table which are then low impedance connected to it. Signal cables should be routed close to this ground plane (which includes the table). The same is true for existing safety ground conductors which should/must be kept.

28 Near the power entrance to the hut the instrumentation ground should connect to an external reference, e.g. the injector rack area ground plane. Cables exiting the hut or connecting to other ground systems or longer than λ/10 should be fitted with DC blocks. Some power supply considerations: If possible use the same main power source for the injector racks and the laser hut with a common safety ground point. Separate power for EMI sources, e.g. flash lamp drivers, and instrumentation racks. Provide rack power filters as proposed for the injector rack area.