1. RF Synchronization Activity
at SPARC
                  
Gallo
and
D. Alesini, M. Bellaveglia, R. Boni, G. Di Pirro, A. Drago, A.Ghigo
P. Baldini, L. Cacciotti, M. Scampati, A. Sprecacenere, S. Quaglia
                   
6th SPARC Injector Review Committee - Frascati, 14 June 2005

2. SUMMARY General Layout;
Demodulation of “square” pulsed RF signals: bench qualification;
Phase monitoring of short pulses: bench qualification;
Control of the phase noise from RF power stations: bench test of a fast intra-pulse phase lock system;
To do list and Conclusions

3. SPARC RF DISTRIBUTION: SCHEMATIC LAYOUT

5. SPARC RF PHASE STABILITY GOALS:
SPARC phase I:
± 3° between the Laser pulse and the Linac RF
(RF gun mainly)
SPARC phase II:
± 0.5° between the Laser pulse and the Linac RF
(RF gun and RF compressor mainly)

6. Phase Lock to the RF Reference Line
All the Linac RF signals will be phase-monitored respect to the RF ref. Line. The resolution of the phase measurements must be << 1 ps ( << 1° 2856 MHz)
The slow phase phase variations will be sampled at 10 Hz (the Linac rep. rate) and corrected acting on the phase shifters (included those placed in the high power waveguide branches).
The fast phase variations (i.e. the jitter with spectral components beyond 10 Hz) are much more difficult to correct. The possibility of implementing intra-pulse, fast phase lock to reduce the fast phase jitter introduced by the RF stations (klystrons + drivers) is well advanced.
Phase Lock to the RF Reference Line

7. RF Demodulation Channel based on I&Q Detector
passive, low cost, 4 quadrants, amplitude independent
1 I&Q Demodulator 0.5 k€
½ PCI Sampling Board 0.4 k€
1 RF Isolator k€
RF Attenuators k€
1.2 k€
1/4 PCI Samp. Board 0.2 k€
1 Peak Detector k€
1 Directional Coup k€
0.7 k€
CW RF
Total cost:
1.9 k€/channel (High Quality)
1.2 k€/channel (Standard)
Pulsed RF
RF Isolator

8. PULSAR I&Q DEMODULATOR
MODEL ID I (custom)

9. PULSAR I&Q DEMODULATOR
MODEL ID I : CALIBRATION

10. PULSAR ID I4-0428
Long Term Acquisition

11. PULSAR ID I
ADLINK MS/s
PCI Sampling Board
Phase Resolution

12. LASER to RF Line Syncronization
The phase of the Laser pulse respect to the Linac RF reference line will be continuously monitored and the “slow drifts” (i.e. the phase noise in a frequency band limited to ½ of the Linac rep. rate) will be compensated (shifting the phase of the RF ref. Line seems the simplest way to do it).
The “fast phase jitter” (i .e. the noise in a band larger than ½ of the rep. rate) of the Laser pulse can’t be compensated and must be kept inside the acceptable limits by specifications (1°nrequested to the manufacturer).

13. Monitoring the synchronization of the bunch from the Gun and of the laser pulse

15. HOMDYN Time-of-Flight Simulations

16. Bunch Monitor Cavity TM010 – 2856 MHz R/Q = 28.5 W Q0 = 15000 (Al)
Qext = 30000
Vp 1 nC) = 4 V
= 0.5 ms
ftun  10 kHz
Bunch Monitor Cavity
TE111 – 2985 MHz
R/Q = 0.5 W
Q0 = (Al)

17. Bench measurements of short pulse synchronization

19. Results of the bench measurements of short pulse synchronization

20. BENCH MEASUREMENTS OF A FAST INTRA-PULSE PHASE-LOCK CONTROL
An intra-pulse phase-lock system is under study to actively correct the fast phase jitter coming from the RF stations.
Since the RF pulse is about 4.5 ms long, the rise time of the system should be of the order of 1 ms, corresponding to a bandwidth of about 1 MHz.
A fast phase modulator is needed, while the overall delay of the connections (including the group delay of the RF station) should be limited to  150 ns.
The speed of the OpAmps manipulating the signal has also to be adequate.

21. PHASE STABILITY MEASUREMENTS
DAFNE LINAC – STATION B
August 2003
mixer sensitivity = 5.6 mV/Deg

22. ± 2.5° ± 1° Long Term (90 min) phase variation: Short Term
(16 shots) phase variation:
± 1°

23. KLYSTRON PHASE LOCK

24. FAST PHASE LOCK FEEDBACK SYSTEM Sketch of the Experimental Set-up
“Noisy” RF Station

25. PULSAR ST-G9-411 Phase Shifter
CALIBRATION
FREQUENCY
RESPONSE

26. Lock Amp: Circuit sketch Fast Phase Lock Feedback System:
Open Loop Gain
(amplitude and phase)

28. OPEN LOOP
CLOSED LOOP

29. OPEN LOOP + noise
CLOSED LOOP + noise

30. To do list and Conclusions

31. DAFNE LINAC KLYSTRON MEASUREMENTS
The phase jitter at the output of one of the DAFNE TH2128C Linac klystrons has been measured to figure out the amount of expected noise in the SPARC case and to gain some experience in this subject.

32. Open/Closed Loop Phase Pulses
256 frames with  10°pp ,
10 kHz phase noise
 10°pp
Single shot
phase pulses

33. KLYSTRON PHASE NOISE MEASUREMENTS CONCLUSIONS
The measured phase jitter at the output of the Linac station-B klystron is about ± 2.5° in long term and ± 1° in short term.
The main contribution seems to come from the driver amplifiers (planar triode technology). In the SPARC case the driver amplifiers will be of a different and better performing technology (pure class A, solid state) with tight phase jitter specification (1° rms). The expected phase jitter values of the SPARC RF stations are consequently lower.