Tom Powers LLRF Systems for Next Generation Light Sources LLRF Workshop October 2011 Authored by Jefferson Science Associates, LLC under U.S. DOE Contract No. DE-AC05-06OR The U.S. Government retains a non- exclusive, paid-up, irrevocable, world-wide license to publish or reproduce this manuscript for U.S. Government purposes.

DISTRIBUTION STATE A T. Powers, LLRF Workshop 2011 Cornell X-ray light source Near perfect energy recovery, i.e. first and second pass 180° out of phase from each other. Moderate Gradient of 15 – 20 MV/m Requires on the order of 2 MW of linac RF power to produce 500 MW of electron beam power. Image Copied from

DISTRIBUTION STATE A T. Powers, LLRF Workshop 2011 Standard Linac Approach Accelerate off-crest in L1 to induce a phase-energy correlation along the bunch Use harmonic RF to linearize the longitudinal phase space (Extract energy from beam) Implement a laser heater (LH) to increase intrinsic energy spread (optional) Partially compress the bunch in BC1 Intermediate acceleration in L2 Fully compress the bunch in BC2 Accelerate to final energy and de-chirp in L3 C. Tennant, IPAC 2011

DISTRIBUTION STATE A T. Powers, LLRF Workshop 2011 Combination Recirculator and Linac C. Tennant, IPAC Inject beam at E inj 2.Accelerate through linac (f L1 ) to induce a phase-energy correlation along the bunch 3.Perform the first bunch compression and linearization in recirculator arcs 4.De-chirp the beam through linac on the second pass by running near zero-crossing (f L2 ) 5.Accelerate on-crest through the afterburner (f L3 ) 6.Perform the final bunch compression

DISTRIBUTION STATE A T. Powers, LLRF Workshop 2011 General Phase and Gradient Stability Requirements Gradient and phase stability determined during beam physics sensitivity studies. Subject to be different from machine to machine. In general Phase stability 0.03º to 0.01º rms (60 fs to 18 fs at 1500 MHz) (folks talk about 10 fs for synchronization of final beams to end stations.) Gradient stability 0.05% to 0.005% rms ? ? cavity by cavity stability or ensemble of cavities ? ? Temperature stabilized LLRF systems probable. For slow drifts remember one bad cable can ruin years of design and implementation.

DISTRIBUTION STATE A T. Powers, LLRF Workshop 2011  Loop phase and amplitude control signals available in control room.  Tuner control algorithm -- may need fast adaptive controls for pulsed machines or ERLs.  Loop gains, bandwidths, etc. control available in the control room.  Filter for rejection of the first π mode frequency below the fundamental frequency.  First fault buffers.  Interface to fast feedback control system.  Quench detection  SEL for cavities with high loaded-Q  One button turn on of cavities even if they are not properly tuned. Some of the “Necessary” Features

DISTRIBUTION STATE A T. Powers, LLRF Workshop 2011 Other uses of LLRF Drive and seed laser phase control. Using optical detector along with a PZT and pico motor controls. Lock laser fundamental frequency to sub-harmonic of cavity RF. Switch to or augment phase feedback loop with a higher harmonic of laser frequency. Check harmonic content, phase noise, etc. as an on line task. Receiver for beam based phase feedback for seed laser phase. As part of a fast feedback system for energy stability. For commissioning the cryomodules and their tuners after installation into the machine. * Caution must be used when choosing frequency for optical and electron beam based detectors in order to avoid interference from high power RF systems.

DISTRIBUTION STATE A T. Powers, LLRF Workshop 2011 ERL or Beam Off Crest Effects Ideally energy recovering linacs (ERLs) operate with second pass beam 180° out of phase with respect to first pass beam. In real machines you don’t always get that. Errors in path length Intentional modes of operation with different phase settings. Changes in second pass phase due to issues like lasing and not lasing in an FEL. Often times one needs to run cavities off crest in machines that are not ERLs. Some examples are: Putting a chirp on the beam. Using cavity focusing and non-relativistic beams in an injector. Bunch compression using 3 rd order harmonic cavities.

DISTRIBUTION STATE A T. Powers, LLRF Workshop 2011 Basic Equations For RF Source Power

DISTRIBUTION STATE A T. Powers, LLRF Workshop 2011 THE EFFECTS OF TUNING ON OFF-CREST CW BEAM LOADING On beam turn on the forward power increases the phase shifts and microphonics effects are multiplied The tuner operates with a goal of making ψ Kly equal to zero by shifting the frequency by δf S which compensates for the I 0 R C sinψ B term. Thus ψ Kly → 0 and P Kly is minimized to : Where δf M is the frequency shifts due to microphonics Thus, assuming that you can wait for the tuner, in this case: 0

DISTRIBUTION STATE A T. Powers, LLRF Workshop 2011 THEORETICAL EXAMPLE OF TUNERS COMPENSATING FOR OFF CREST BEAM LOADING CEBAF 7-Cell Cavity L = 0.7 m (r/Q) = 960 E = 8 MV/m Q L = 2 x 10 7 δf = 10 Hz I 0 = 10 mA Pass one ψ B = -10º Pass two ψ B = 166º Resultant Beam 0.7 mA at 78°

DISTRIBUTION STATE A T. Powers, LLRF Workshop 2011 THEORETICAL EXAMPLE OF TUNERS COMPENSATING FOR OFF CREST BEAM LOADING

DISTRIBUTION STATE A T. Powers, LLRF Workshop 2011 PREDICTED AND MEASURED FORWARD POWER IN AN ERL The solid lines indicate the predicted values based on: Q L = 2 x 10 7 E = 5.6 MV/m. Δf = 10 Hz Test Process: Tune the cavity with no current. Disable the mechanical tuners. Ramp the current up and record the forward power and phase. Repeat with Tuners enabled.

DISTRIBUTION STATE A T. Powers, LLRF Workshop 2011 Predicted and Measured RF Drive Phase In an ERL

DISTRIBUTION STATE A T. Powers, LLRF Workshop 2011 Multiple Cavities Using a Single Source It can be a desirable to use a single source to drive multiple cavities. Reduced cost per Watt at higher RF –Power levels. Availability of klystrons or IOTs at desired levels for multiple cavities. Unavailability of klystrons, IOTs at desired power levels for single cavities. Reduced number of LLRF systems to drive cavities.

DISTRIBUTION STATE A T. Powers, LLRF Workshop 2011 Multiple Cavities Using a Single Source It can work in a straight forward manner when: The cavities are operated near crest. The beam is not sensitive to minor variations in gradient and phase. Loaded-Qs are well matched. Gradients are close to the same for all cavities. The loaded-Qs are relatively low as compared to pressure sensitivity and microphonics. You have the advantage of a large number of cavities and individual errors are corrected by statistics.

DISTRIBUTION STATE A T. Powers, LLRF Workshop 2011 Multiple Cavities Using a Single Source It can present problems when: The beam is sensitive to errors in gradients or phase. Detuning becomes significant as compared to the FPC bandwidths. Cavities are operated at different gradients. Cavities have different loaded-Qs Cavities are operated at different beam phases with respect to crest. Induced phase shifts are comparable to that required for putting an energy chirp on the beam. While linacs are an area where this concept can be very practical, injectors are an area where the problems become important especially when space charge and cavity induced beam focusing are important.

Stability in 6300 sec. LLRF stability study with 7 cavities operation at 25MV/m Field Waveform of each cavity - Vector-sum stability: MV/m ~ MV/m (~0.03%) - Amplitude stability in pulse flat-top: < 60ppm=0.006%rms - Phase stability in pulse flat-top: < degree.rms vector-sum gradient amplitude stability in pulse flat-top phase stability in pulse flat-top 7-cavity operation by digital LLRF A. Yamamoto, SRF Advances in ILC-SRF 19

DISTRIBUTION STATE A T. Powers, LLRF Workshop 2011 Stability in 6300 sec. LLRF stability study with 7 cavities operation at 25MV/m Field Waveform of each cavity - Vector-sum stability: MV/m ~ MV/m (~0.03%) - Amplitude stability in pulse flat-top: < 60ppm=0.006%rms - Phase stability in pulse flat-top: < degree.rms vector-sum gradient amplitude stability in pulse flat-top phase stability in pulse flat-top 7-cavity operation by digital LLRF Individual cavity gradients vary during the pulse. Even with this, the vector sum of 7 cavities has very good phase and amplitude stability (i.e. the system works as designed.) Note that the gradients have not yet achieved their steady state values as would be necessary in a CW machine. If you fix the gradient what happens to the phase errors?

DISTRIBUTION STATE A T. Powers, LLRF Workshop 2011 Some Questions What happens when cavities are operated CW? What happens when you have to tune the machine at one current and operate at another current?

DISTRIBUTION STATE A T. Powers, LLRF Workshop 2011 Simulation Method for CW Applications  Use the basic complex RF voltage to complex gradient equation to calculate the field in each cavity, including beam phase and cavity detune angle.  Sum the real and imaginary parts of the electric field signals for each of the cavities.  Compare the vector sum to the desired vector sum and calculate the error in the vector sum.  Add, with gain, the complex error to the complex RF voltage from the current pass.  Use this sum to calculate gradient in each cavity.  Repeat until the real and imaginary parts of the vector sum error are below a threshold. Where ψ is the cavity detune angle.

DISTRIBUTION STATE A T. Powers, LLRF Workshop 2011 Phase Error In When The Phase of the Beam Relative to the Cavity Field In 1 of 3 Cavities Is Different

DISTRIBUTION STATE A T. Powers, LLRF Workshop 2011 Phase Error In When The Phase of the Beam Relative to the Cavity Field In 1 of 3 Cavities Is Different Tune Beam Ops Beam

DISTRIBUTION STATE A T. Powers, LLRF Workshop 2011 Gradient Error When the Beam Phase in 1o f 3 Cavities is Different

DISTRIBUTION STATE A T. Powers, LLRF Workshop 2011 Gradient Error in When the Loaded-Q of 1 of 3 Cavities is Higher Than the Others

DISTRIBUTION STATE A T. Powers, LLRF Workshop 2011 ERROR IN GRADIENT AND PHASE WHEN 1 OF 3 CAVITIES IS DETUNED Q L = 2e6 F 0 = 1497 MHz (r/Q) = 960 Ω/m L = 0.5 m β = 3000

DISTRIBUTION STATE A T. Powers, LLRF Workshop 2011 Comments on Errors  Beam phase being different between cavities comes up in at least three places. Injectors where cavities are frequently operated with beam phases that are tens of degrees different. When setting the phase of cavities within a linac or an injector. Effects change with different beam loading conditions.  Detuning Effects Effects become more important at increased values of loaded-Q. Cavity detuning is a function of helium pressure which is seldom completely stable. For beams operated off crest the cavity tuners respond to changing beam conditions. Determining the proper forward to reflected power phase set point is, in general, a +/- 3° operation.

DISTRIBUTION STATE A T. Powers, LLRF Workshop 2011 ? ? ? Conclusion? ? ? I hope that I have given you something to think about with respect to the requirements for LLRF used in the next generation light sources as well as some topics for further discussion during this workshop.