Participants: 49

Follow-ups:

• William Viganò identifies the best kit for the Nanocrystalline cores, let him know if someone is interested.
• Christos Zamantzas proposes to collect the measurements on the characterisation of the cables and connectors and write a paper together with BPM and BLMs.
• Michele Bozzolan needs to order cables for the CKC50 cables, let him know if someone is interested. 
• Marek Gasior advices to avoid LEMO connectors when possible and go for BNC instead if quick manipulation is necessary. However, if quick manipulation is not mandatory, better are screwed connectors, like SMA or N. N are also very robust mechanically.
• Thierry Bogey said that he has very good experience with SMB connectors, which are quick-mount and small.
• Manfred comments hat D.Valuch wants to build a laboratory for specific tests on EMI, discussed via the WG, let Manfred know if someone has ideas. 

BLM, William Viganò

• 1st installation in June 2020 -> September 2020 start to see signal peaks
• Oct 2020: inspections shown signal cables routed together with high power and high voltage cables, also in flooded areas.
• OP complained about TT2 BLM noise
• Noise affects mainly the PS transfer lines
• Two source of noise:
  • Asynchronous Noise in Switch Yard (SWY) due to power converters switching frequency.
  • Synchronous Noise due to the DFA pulsing (TT2 kickers).
• Mitigation measures:
  • Cables re-routing adding metallic flexible ducts (galvanised steel duct) that hold the signal and HV cables, grounded every 5 m, and is installed at 50cm distance from power cables (before was few cm).
  • 3 orders of magnitude noise reduction for the SWY channels.
  • Cost of about 15kCHF - removal and re-routing of cables - for 7 channels + added metallic duct. The cost mainly from installation, not from components. 
  • Collaboration with ABT: installation of Nanocrystalline core on 2 channels, then on other channels accordingly with the improvement obtained. 
  • 2 orders of magnitude noise reduction for channels equipped with Nanocrystalline core.
  • Cost of about 300CHF per channel + manpower.
  • The preferred solution is the cable re-routing, while the Nanocrystalline core is a backup solution where when there is limited available time and limited budget.
• Limitations :
  • Big ground loops are present in the SWY.
    • Ground connections are not clean but affected by voltage potential instability due to high current switching (both SWY and TT2 tunnel).
    • Request to have an EMI Grounding architecture at CERN (not only based on safety).

BPM, Michele Bozzolan

• VME self EMI, Linac 4 BPMs.
  • Conducted noise common to all 4 channels in the Analogue receiver board improved by use of SMD common mode chokes. Radiated noise reduced by improving the electrical contact of the metallic cover of the high freq. amplifier.
• Noise coupling to long cables, PSB extraction line BPMs.
  • Observed as large oscillation beam synchronous
  • Improved by cable change: 2xCC50 differential -> 1xCKB50 singled ended (factor 5)
• Kickers induced noise, PSB trajectory system (BTMS)
  • ABT reduced the EMI generated by the kickers
  • Interference reduced by still visible

·         Coupling mechanism to BPM to be identified

• Radial loop electronics, PSB trajectory system
  • 10kHz noise bursts 
  • Adding a buffer amplifier for isolating from the radial loop noise.
• Noise coupling to long cables, PS trajectory system (TMS)
  • 150 m CK50 
  • Similar to BLM: coupling between power cables and signal cables but in this case ferrites would not help, better cable isolation yes.
  • CKB50 has better transfer impedance wrt CK50 and better magnetic field shielding because of the ferromagnetic layer
• Connectors for the tri-axial cable CKC50 developed by BLM together with EN-EL are expensive. BPM they use a simpler version (coaxial).

LIU wire-scanners, Jonathan Emery

• Noise observed during the scan, as the wire is moving. In between scans the noise is lower.
• How the power is sent to the motor was changed between legacy BWS generations and new LIU systems.
• But cabled routed together (PL7SE, CK50 and NES18): PMT signal 1mA - 1V on CK50 and motor powering 400V, 60V.
• Initial 2021 commissioning:
  • Lower noise if PMT system isolated from the machine grounding
  • Main noise source identify at 16kHz with a reproducible phase, in the PS goes up to 1.3% (of the acquisition’s full scale) at the moment of the scan.
  • Studied the effect of ferrites, main peak reduced but harmonics appeared.
• Marek: Suggests to change the type of ferrites for higher frequencies.
• William: Number of loops and ferrite types could probably be optimised – if we would group the ferrites procurement we could probably have different types in stock to quickly experiment.
• Marek: add few ferrites (ceramic type) without any wiring for the higher frequencies, he has also some.

SEM, Jean Tassan-Viol

• Linac 3 SEM grid noise issue, between Switch Yard and Linac4. During years was ok ITF.MSG.05 and suddenly became noisy.
• Measured of noise in the air and in the metallic structures.
• Found: bad electrical connections of the metallic structures , ground not done for EMI reduction.
• Problem found:
  • Insulating washer under the BNC connector (not new), to break the ground loops but it was a bad idea.
  • Quick solution, add a clip in the coaxial.
• Marek advices to use BNC instead of LEMO.  Jean comments that now they don’t use LEMOs, now they don’t use for new things.
• Thierry Bogey also suggests SMA instead of LEMO.

BTV, Stephane Burger

• SPS/BA2/TT20 Feb 2021 huge noise on the cameras. Electronics checked, as well as grounding/shielding on cables/connectors/interfaces, etc. noise still present.
• Camera re-installation in March 2021 at BA80/TDC2 and same noise pattern found. 
• Source of the noise identified: New stepper motor driver as of switching mode (vs DC before the consolidation in LS2)
  • Mismatched pairs of cable used for the motor phases -> correction of pairing + choke on all drivers in TT20 crates and 2 drivers in TDC2 crates (out of stock due to electronics procurement issue).
  • Noise reduced but still a contribution.
• Next steps:
  • Install choke in the rest of drivers. 
  • Test of notch filter on the video input side is on-going.

BCT, Romain Brice Ruffieux

• 2019 PSB noise due to the distributor BI.DIS10
  • Several tests carried out to check the machine ground, vacuum chamber ground, surface ground, induced currents and other detectors equipment in the vicinity.
  • Ferrite on the BCT cables and on the distributor cables decreased the noise interference amplitude by 93%.
• 2022 LINAC4 BCT on 2 devices
  • Noise not as high as in the PSB.
  • New connexion box and a new amplifier designed by M.Gasior installed. 
  • Noise reduced on amplitudes by 70%, but the baseline fluctuation is still there.
• East Area BCT
  • Addition of 1 new BCT
  • Disturbance every 2usec (PS revolution)
  • OP machine optimisation reduced the error. 
• TT2 BCT
  • Similar to BLM System, fast pulsed deflector kicker DFA induced noise. 
  • Ferrites showed deterioration in some BCTs —> they have been removed
  • Improved grounding —> some improvement observed.
  • Removal of the superfluous head amplifier —> clear improvement but still residual disturbances on BCT126. Installed later an improved version. 
  • Low frequency noise improved by changing to thin SMA coaxial cables by thick N cables.