• Some estimates of the various kinds of switches was attempted
• Thyratron costs are around $8K small quantity and ROM (rough order of magnitude) estimate is that this could drop 2X in large quantity
• IGBT’s should be about $200 each in quantity for the 2SKV unit. A Cassel switch (72 units) would be $14.4Kfor IGBT’ s only.
• A hard tube is estimated to cost around $40Kfor 5OOKV @ 55OA; should be less for 275A. A series IGBT switch (200 units) is $40Kfor IGBT’ s only. In these designs the transformer is eliminated which helps some to ofSset higher switch costs.
• Participants were encouraged to think of scenarios which could drive 2, 4 or even 8 klystrons. Iffeasible this could reduce capital cost, but efficiency of line, hybrid and induction designs may be lower, resulting in higher operating costs. Tradeoffs need to be evaluated.
• Conclusion: None of the designs appear to offer major cost advantages if driving 1-2 tubes. Driving more loads and streamlining of assemblies for easier manufacturing seems to be the only hope for significant cost reduction over current Baseline design (Line Type, 2 tubes)

slide 5 (back to top)

Potential Industry Collaboration

• Invited feedback from vendors on how best to collaborate
• Suggested participation in future design reviews.
• Suggested a regular news brieftng via the Web.
• Requested interest if any from system integrators who would be interested in providing a complete modulator
• Two responses: Northrup Grumann & Maxwell. (More needed.)
• Specific collaboration agreed to with Diversified pertinent to Hybrid Type switch test mockup and SLAC transformer.
• Strong interest from Litton in pursuing hard tube development
• Should explore similar collaborations with KEK and Japanese industry.

Conclusions

• Some new insights into advantages of each design especially overall efficiency potential
• Results should help us make an optimum R&D plan for continued modulator development

slide 6 (back to top)

MODULATOR SUMMARY

• Linac Baseline Design: Line Type 2-Pack - Held workshop to connect with industry & researchers - Advanced the prototype tank design - Made specifications for major components

• Thyratron, capacitor, pulse transformer, charger - Contacted industry for quotes - Completed first level cost estimate & WBS entry - Industry collaboration to evaluate Commercial IGBT stacked switch in process

slide 7 (back to top)

MODULATOR SUMMARY

Linac R&D Design Solid State 2/4/8 Pack

• Evaluated IGBT/Metglas stack of 6
• Achieved sub-100 nsec switching performance ofIGBTs
• Calculated & measured efficiencies/losses
• Calculated performance of 2/4/8 Pack designs
• Simulated overall performance and built costmodels
• Investigated protection issues for driving multiple loads
• Best economics for &Pack, w/>75% Efficiency
• Other configurations possible:
• 4-Pack (2T), 2-Pack (4T), Parallel combinations of IGBT drivers to match IGBT current capability to load

slide 8(back to top)

MODULATOR SUMMARY

S&L Band Baseline Cost Models

• Postulated Linac Line Type design to drive wider pulses, higher power.
• Baseline Cost Model includes dual PFNs and Thyratrons.
• Earlier model assuming 150 MW tubes abandoned.
• Current model assumes driving single 65 MW or 75 MW tubes with one modulator as in ZDR (See table next slide).

R&D Induction Model Potential

• Reliability improvements with solid state would be highly desirable, but have not been examined seriously.
• Driving considerably higher average powers economically appears problematical because size, weight and cost of Metglas scales with Volt-Seconds.

slide 9 (back to top)

MODULATOR SUMMARY

S&L Band Progress

• Not developed beyond basic concepts
• Based on power consumption, these modulators will cost proportionately more than the X-Band design driving two Klystrons (1 S-2.5 times average power)
• Technically somewhat easier because peak voltages, currents are lower.
• Feasibility of Induction Modulator for S&L needs to be investigated in FY99. -
• Line Type needs to be modeled further and possibly some component work with vendors (e.g. Transformer)

slide 11 (back to top)

MODULATOR SUMMARY

R&D Plan for X-Band Development - Goals:

• Improve reliability using solid state versus Thyratron
• Improve Efficiency from -60% to >75 %
• Reduce cost per Klystron by A= 2X
• Support Klystron development testing program
• Support Main Linac Full Power Test program with S-Pack prototype(s) in early CYOl

slide 12 (back to top)

MODULATOR SUMMARY

R&D Plan for X-Band Development

• Build prototypes in -12 months of Line Type and Solid State Induction type.
• Test both with particular attention to Klystron protection.
• Partnership with LLNL-LBL on Solid State R&D program
• LLNL has relevant experience with Induction Linac modulators and drivers (FETs, primarily)
• Solid State proto will contain 50 cores configured to produce 500 kV at 1 kA with 2T secondary, to drive up to 4 Klystrons (4-Pack)
• Other configurations will be studied (e.g. driving same number of cores with fewer IGBTs with cores in parallel)
• In FYOO, depending on success, double to 100 cores, 500 kV at 2 kA with 1T secondary, capable of driving an &Pack
• Line Type will serve as fallback as well as a test vehicle for new 75 MW Klystrons
• Coordinate Modulator development program at Full Power Test demonstration of a Linac Nonet (l/9 of a Sector)
• Construct a new Modulator Nonet section separate from existing NLCTA modulator installation, to preserve NLCTA as a component and structure test facility.
• Plan for operation in early CYOl .
• MK Protection, LLRF and Controls interfaced to existing NLCTA system need to be developed in parallel.
• Reliability engineering & Klystron protection studies need to proceed in parallel for both modulator types.

slide 14(back to top)

MODULATOR SUMMARY

Production Plans & Issues

• Modulator production requirements have to match availability of accelerator housings on the one hand, and Klystron availability on the other.
• A period of rampup is needed at the beginning of which at least two vendors are qualified to participate in a pre-production run to build a significant number of test units.
• This period will also require some additional 75 MW Test Stand capabilities which can be partly met by the early run(s).
• At peak production, program requires installation of the equivalent of an &pack every two (2) working days (e.g. 40 ea. 2-Packs per month). See curves on next slide.
• Final production shown is for maintenance spares of -115 2-Pack equivalents