1. Time-Resolved Images of Coherent Synchrotron Radiation Effects
P. Emma et al, SLAC
Presented at FEL-2012

2. Motivation & Idea CSR is one of the limiting effects for FELs
“…would like to see the time-resolved effects of CSR” - Joe Bisognano, 2008 LBNL FEL Workshop
“Can we use a skew quad?” - Oleg Shevchenko (BINP), 2008 LBNL FEL Workshop
Yes… Skew quad is a poor man’s transverse RF deflector, with a few differences
Skew quad installed at LCLS BC1 in 2009 (PAC’09)
Creates vertical dispersion after chicane revealing x vs t space (CSR effects)
Measurements & simulations discussed here…

3. Add a CSR-Diagnostic Skew Quad in BC1
screen
BC1
Skew quad maps the time coordinate of the pre-BC1 bunch onto the vertical axis of the screen

4. How it Works (first measurements Dec. 15, 2009)
OTR11 screen in BC1
~1% chirped energy spread at 220 MeV and large dispersion at chicane center
Skew quad generates y' from x:
Dy' = x/f
OTR12
OTR11
skew
Dy' = x/f
x ~ t ~ E
OTR12 screen after BC1
(skew quad OFF)
0.05 mm rms
OTR12 (skew quad ON at kG)
4 mm
(~1 ps)
Ipk = 250 A (after BC1)
y ~ time

5. Simulation of skew quad effect at OTR screen with and without CSR
OFF
CSR
Skew
quad ON
CSR
OFF
spot streaked on screen
x-y spot on screen
Simulation of skew quad effect at OTR screen with and without CSR
Skew
quad
OFF
CSR ON
Skew
quad ON
CSR
CSR blows up x spot size
CSR is time resolved

6. BC1 Area in LCLS (S-band & X-band RF)
(L1X)
135 Me13V
OTR12
SQ11
OTR11
BPM

7. measured temporal profile is used in simulations
Simulated pre-BC1 temporal profile
Measured pre-BC1 temporal profile
sz  0.5 mm
sz  0.5 mm
ignore red Super-Gaussian fit
Temporal profile measured at 135 MeV using RF deflector

8. X-band RF (L1X) varied: LiTrack simulations too weak (16 MV),
too strong (22 MV), and
just right (19 MV)
HEAD
L1X = 16 MV (too weak)
LiTrack simulations
L1X = 22 MV (too strong)
TAIL
L1X = 19 MV (just right) – linearizes phase space, but not shown here.

9. L1X Voltage = 16 MV (too weak)
Scan L1S phase holding BC1 energy constant
L1X Voltage = 16 MV (too weak)
-29.0°
Measured on OTR12
-31.0°
HEAD
-31.5°
BPM
after
BC1
-32.0°
-32.5°
-33.0°
-33.5°
-34.0°
-37.0°
TAIL
skew
OTR12 E X
OTR11
BPM
Energy loss in the final BC1 bend kicks to the right here
HEAD t

10. L1X Voltage = 22 MV (too strong)
Scan L1S phase holding BC1 energy constant
L1X Voltage = 22 MV (too strong)
-26.0°
Measured on OTR12
-28.5°
-29.0°
HEAD
-29.5°
-30.0°
BPM
-30.5°
-31.0°
-31.5°
-32.0°
-32.5°
-33.0°
-33.5°
-34.0°
TAIL
skew
OTR12 E X
OTR11
Energy loss in the final BC1 bend kicks to the right here
HEAD t

11. L1X Voltage = 19 MV (just right)
Scan L1S phase holding BC1 energy constant
L1X Voltage = 19 MV (just right)
-27.0°
Measured on OTR12
-30.0°
HEAD
-30.5°
BPM
-31.0°
-31.5°
-32.0°
-32.5°
-33.0°
-35.0°
TAIL
skew
OTR12 E X
OTR11
HEAD
Energy loss in the final BC1 bend kicks to the right here t

12. Now Switch OFF L1X RF
Measurements (left) and simulations (right) with X-band RF switched off. L1S phase settings differs from simulation by 2°.
acceleration
measured
simulation

13. Summary
Skew quad is simple, inexpensive, and adds time-resolved measurements of bunch
Note this resolves the pre-compressed bunch length coordinate, not the compressed one
Reasonable agreement with simulations, but preliminary results
Should repeat and run more careful simulations as well

14. At BC1 Extreme Compression with Laser Heater OFF
e/e0 > 10

15. Emittance Growth with CSR Measured in 2009… (good agreement with Elegant and CSRtrack)
Phys. Rev. ST Accel. Beams 12, (2009)

16. Blow-up reasonably well modeled at extreme compression
-26.0°
-22.0°
-22.0°
-26.5°
Vertical (time) is pre-BC1 bunch-length coordinate, not post-BC1, so blow-up is not so familiar wrt usual simulations
Elegant
Measurement
(L1X = 19 MV)