Minutes of 113^th Impedance Working Group meeting

Introduction and previous results

• BGI is used for bunch-by-bunch measurements of the transverse beam profile. Beam ionises residual gas and ionisation electrons are accelerated and guided into the detector (Timepix4)
• Known risks: beam coupling impedance and beam-induced heating.
• New design was developed by BI group, more compact than SPS BGI.
• First impedance results showed that a cathode made of steel is preferable wrt cathode made of ceramic with resistive coating (lower power loss)
• HV feedthrough and how it is terminated has an impact on beam coupling impedance.
  • If not terminated on resistive load with same characteristic impedance of the feedthrough (e.g. short or open circuit), signal reflections appear, associated with impedance contribution. 
  • Matched temrination assumed for following results

First impedance results

• Initial impedance result gives a total power loss of 95 W, up to max 113 W (rigid freq. shift +-20 MHz).
  • 66 W from broadband contribution below 500 MHz -> 98% dissipated through the HV feedthrough port -> resistive termination should be included to dissipate this power
  • 46 W from impedance in frequency range between 500 MHz and 1.4 GHz. 
    
    

Impedance mitigation

• Enlarged opening aroud HV pin and fused silica tube -> brodaband contribution at 882 MHz is damped
• Reduced volume with higher shoulders next to the RF shield -> reduced resonances at 808 MHz and 1.91GHz
• Tuning of remanent resonances with reducing volume of ceramic (macor) holders -> resonance at 808 MHz is completely damped, resonance at 1.91 GHz is drastically reduced with shunt impedance of 48 Ohm (initial was 639 $\Omega$) and reduced Q factor. 

Power loss of mitigated design

• Total power loss goes down to maximum 78 W
  • 65 W caused by broadband peak at 169 MHz -> 98% dissipated through HV feedthrough (matched load)
  • from preliminary results, the remanent 13 W are split between feedthorugh port (80%), cathode (14%) and ceramic holders (4%)

Conclusions and outlook

• Important to ensure impedance matching at HV feedthrough port -> resistive load to dissipate power
• Total power loss was higly reduced and impedance mitigated
  • Latest design with proposed modification is currently simulated and power loss will be provided. 
• Longitudinal and transverse stability to be assessed.

Discussion

Chiara P. mentions that BI is also investigating on the use of quartz for the holders. Michela shows that keeping the volume of holders the same but having a material with lower dielectric constant shows similar results wrt smaller macor holders. Chiara P. also mentions that it would be important to consider the longer eletrical path for HV reasons with including grooves on the ceramic holders. Michela clarifies that the holders were simplified for impedance simulations but the latest model received from L. Margerison with smaller holders has the longer electrical path.

James raises that the understanding on the HL-LHC BGI can be useful to better characterise and investigate on the SPS BGI and the impact of the feedthrough there. Chiara P and James mentions that it would be good to have power loss map for SPS BGI design. 

AoB - LHC BWS Observations beginning of the run 2026

Speakers: Jonathan Emery (CERN)

• BWS LHC - 2026 configuration:
  • Beam 1: 2 legacy and 2 hybrid (same as 2025), 1 new BWS prototype (B1H3)
  • Beam 2: 2 legacy and 2 hybrid (same as 2025)
• Measurements of resistance on the wire is done by injecting continuously 300 uA current and measuring the voltage. Readings of resistance is about 80 Ohm on the prototype. 
• Prototype has 8 temperature probes installed on the two sides, outside of the instrument, to have readings of temperature on the absorbers.
• Temperature logs are not yet in NXCALS but recorded locally

• Temperature is calculated from the variation of the resistivity on the wire

• Estimated temperature on the wire is higher with respect to the hybrid version. Correlation with the bunch length is observed -> Higher temperature on absorbers and on the wire when the bunch length is longer.

• Estimated rate, 1-1.5 degrees per hour.

• Not well understood correlation between higher temperature readings and longer bunch length (with larger spectrum)

Discussion 

Carlo attributes the strange correlation with bunch length to some issues which were seen recently with the blow up in LHC.

Rama confirms that there was this issue with bunch length spread in operation, observed mainly on beam 2. He suggests Jonathan to check again the data now, the issue was fixed on Friday after Easter.