Choose timezone
Your profile timezone:
PS Landau RF System Study Group
→
Europe/Zurich
864/2-B14 - SALLE J.B.ADAMS (CERN)
Present: R. Calaga, H. Damerau, N. Esfahani, G. Favia, E. Jensen, A. Lasheen, L. De Mallac, M. Morvillo, M. Paoluzzi
- Heiko presented some news and updates with respect to the previous meeting.
- Concerning the 'lambda/4' resonator design with on-axis tuner two main issues became clear.
- The gap short-circuit of the existing 40/80 MHz cavities would be mechanically very difficult to fit with the Landau cavity.
- The complexity of a design in vacuum would be very significant.
- Heiko suggested to therefore define a 'conservative' default option:
- Assume a harmonic number ratio of 4, corresponding to a cavity in the 40 MHz range.
- The total voltage of about 50 kV will be distributed over two cavity units, i.e., 25 kV per cavity.
- With these relaxed constraints the cavity can be in air with a ceramic gap.
- Possibly reuse existing ceramic gaps from the PS 200 MHz RF cavities. Michele will investigate the quantity and type (Al2O3/BeO) of spare gaps in stock for the 200 MHz cavities.
- The conservative design significantly simplifies the mechanical design, in particular a gap short circuit.
- Improved reliability is expected thanks to the moderate gap voltage and the possibility to run with a single cavity in 'degraded mode'.
- Nasrin showed the 2nd iteration of the
- Having re-iterated the mechanical constraints, the length of the bare Landau cavity should not exceed about 720 mm to mechanically fit into the short straight section.
- Dielectric breakdown or surface field emission is expected to occur at gap voltages significantly higher than the operating voltage required.
- According to the hatch chart, multi-pacting may occur across the ceramic gap in a voltage range of 0.16 to 4.7 kV. Its is assumed that this can be removed by conditioning.
- Different types of capacitive loading (meander-line) and mushroom-like structure have been studies. The meander-line capacity is not an efficient means for capacitive loading as the a large volume is taken away from the cavity. The mush-room structure has been retained for the simulations.
- With the G-510 garnet material quality factor of about 8000 is achieved with an R/Q of about 36...37 Ohm. During the Skype meeting with the Fermilab colleagues it was suggested to stay with AL800 garnet material.
- Quality factor and R/Q are hence very similar to the 1st iteration design, showing the robustness of the approach. The present design has been optimized for high shunt impedance.
- Seen that is may be more efficient to increase the drive power, avoiding a rather complicated cavity geometry, Nasrin will investigate adding again more ferrite and reduce the capacitive loading.
- The next steps also include further investigations of losses in the ferrite (thermal stability) and considerations of higher-order modes.
- Michele commented that the air gap between the ferrite ring and the inner/outer conductors should be included in the model.
- Erk remarked that the shunt impedance would be reduced when adding ports the connection to the amplifier, etc.
- Mauro explained that no issues were observed with the gaps for the Finemet cavities, they are straightforward to produce, which removes the need to stick to the gap design of the 200 MHz cavities.
- Michele suggested to study also a double-gap resonator structure, starting from the existing 20 MHz cavities. This would simplify the coupling to the amplifier (step-ratio of 1:1). Additionally, it has a mechanically compact ceramic gap.
- Mauro presented an alternative option for the Landau cavity based on Finemet material. It does not require any tuning as the quality factor can be adjusted to cover the frequency range needed.
- The shunt impedance of a Finemet core has a wide maximum in the range of 30-40 MHz. This 'resonance' is caused by its inductance and capacity. When putting Finemet cores in series, the inductance adds up while the capacity stays quasi constant, which lowers the natural resonance frequency.
- An additional coil can be installed in parallel to the gap which (1) pushes the resonance frequency of the system upwards and (2) increases the quality factor of the resonator. Both effects are in the right direction for the Landau cavity.
- The Finemet cores can be powered by solid-state amplifiers individually or in parallel, but powering with tubes is excluded due to their large output capacitance.
- The main limitation of the Finemet cavity approach is the cooling of the cores. When doubling the surface compared to the Finemet rings used in the PSB cavities, the maximum power loss is estimated at 2 kW per ring.
- Considering a 40 % duty cycle and a shunt-impedance of 210 Ohms per ring, the achievable voltage is estimated to 1.45 kV. The quality factor of the high current coil across the gap being an issue, Mauro suggested to use coaxial line stubs (with some dampers against higher-order modes they may introduce).
- Based on this design, a cavity in a short straight section of the PS could be constructed with a total shunt impedance of about 3.4 kOhm, proving almost 25 kV per short straight section. However, the total RF power is estimated to 80 kW, hence a total of 160 kW for two cavity units.
- Fast RF feedback could be considered, but would require to install the solid-state RF amplifiers close to the cavity, and due to the large bandwidth of the system, only 2-3 impedance reduction is hoped for.
Follow-up
- Giorgia will perform beam-loading and feedback estimates for both options, to investigate the additional power required to counter-act the beam-induced voltage.
- Alexandre will check the consequences on longitudinal beam stability, comparing the narrow resonance of the ferrite cavity with the rather large bandwidth of the Finemet cavity proposal.
- Michele will check the inventory of existing ceramic gaps from the PS 200 MHz.
There are minutes attached to this event.
Show them.