1. Experiments on Fast Ion Instability at PLS
Mar. 6, 2007
T.-Y. Lee, H. S. Kang, and J. Choi
Pohang Accelerator Laboratory

2. Brief History of Fast Ion Instability (FII)
T. O. Raubenheimer and F. Zimmermann: Prediction of FII. Linear theory.
Phys. Rev. E 52, 5487 (1995).
2. J. Byrd et al. : First observation of FII at ALS.
Phys. Rev. Lett. 79, 79 (1997).
3. M. Kwon et al., J. Y. Huang et al.: Streak camera images of FII at PLS. This work was done in collaboration with KEK.
Phys. Rev. E 55, 7550 (1997). Phys. Rev. Lett. 81, 4388 (1998).
4. It turned out that FII was not harmful on B-factories.
5. Interests in FII is revived for the ILC damping ring design.
T.-Y. Lee
Pohang Accelerator Lab.

3. Fast Ion Instability
Ions are created by passing electrons and cleared after the final bunch.
Electron bunch train
Ion clearing gap
Tail
Head
T.-Y. Lee
Pohang Accelerator Lab.

4. Fast Ion Instability The number of ions grow along the bunch train.
1. The FII growth rate depends not only the vertical beam size but also
N: number of electrons in a bunch, P: vacuum pressure,
n: number of electron bunches, bunch spacing….
T.-Y. Lee
Pohang Accelerator Lab.

5. Pohang Accelerator Laboratory
160-m long 2.5-GeV S-band PLS Linac
2.5-GeV 3rd generation Pohang Light Source (PLS)
T.-Y. Lee
Pohang Accelerator Lab.

6. PLS Storage Ring Parameters
Energy (GeV)
2.5
Ring circumference (m)
280.56
Lattice
TBA
RF frequency (MHz)
Bunch spacing (ns) 2
Harmonic number
468
Natural emittance (nm rad) 18
Number of superperiods 12
T.-Y. Lee
Pohang Accelerator Lab.

7. The purpose was to verify experimentally.
PLS Experiment in 1997
The purpose was to verify experimentally.
Main tools: Streak Camera, observation of
ion frequency
T.-Y. Lee
Pohang Accelerator Lab.

8. Streak Camera Images Normal vacuum pressure Ion pumps turned off
0.5 nTorr
2.2 nTorr
Induced by CO-ions
T.-Y. Lee
Pohang Accelerator Lab.

9. Streak camera Images with He Injection
Normal condition
0.2 nTorr Helium
3.34 nTorr Helium
Instability was also excited by Helium gas injection in one place of the ring
T.-Y. Lee
Pohang Accelerator Lab.

10. Ion Frequency Observation
This figure shows FII depends on N and n separately not on the total current.
It also shows that the n-dependence is more important. For larger n, N becomes less important.
P=0.5 nTorr
T.-Y. Lee
Pohang Accelerator Lab.

11. New PLS Experiment Motivation
Due to the revived interests in FII, we planned to do more experiments.
PLS has a Revolver In-Vacuum X-ray UNdulator (RIVXUN). The minimum gap is 5 mm. The length is 1.2 m
If the beam orbit is distorted when the RIVXUN gap is lowered, the vacuum pressure of the undulator area is increased up to one order of magnitude higher. (Undulator SR hits internal structure).
It is possible to control the local pressure step by step.
T.-Y. Lee
Pohang Accelerator Lab.

12. RIVXUN: Revolver In-Vacuum X-ray UNdulator
Permanent magnet structure is a revolving type with four arrays, which provides 4 different undulator periods of 10, 15, 20, and 24 mm.
Array
Undulator Period [mm]
Number of period
C, c 10
101
B, b 15 67
A, a 20 50
D, d 24 42
Undulator magnet length is 1.2 meter
Magnet material : Nd2Fe14B
designed at Spring-8
Kitamura et al. NIMA 467, 110 (2001)
T.-Y. Lee
Pohang Accelerator Lab.

13. Gas Desorption by Photons
To reduce the resistive wall impedance,
the permanent magnet array is covered with
a 50 m-thick Cu sheet coated with 50 m-thick Ni.
The bakeout temperature for
the vacuum chamber: 200°C
the magnet arrays: 125°C
Synchrotron radiation should be blocked by photon stops
Gas desorption by stray photons
- Photon desorption
- Electron-stimulated desorption
Pre-cleaning by stray photons: Aging process
Out-vacuum
In-vacuum
Aging by stray photons
Enough : Continuous
Not enough: Intermittent
Gas desorption by photons
weak
strong
T.-Y. Lee
Pohang Accelerator Lab.

14. Ion Instability: Measured vacuum pressure in Revolver
The Revolver vacuum pressure increased by 10 times when the gap was changed from 20mm to 6mm.
Even this local high vacuum pressure gives rise to FBII
Beam current : 100mA
1.3x10-9 Torr
T.-Y. Lee
Pohang Accelerator Lab.

15. FII Experiment in an In-Vacuum Undulator
Electron bunch train
Revolver Undulator
Ion clearing gap
Tail
Head
Relatively short ion clearing
gap is enough.
T.-Y. Lee
Pohang Accelerator Lab.

16. FBII at 6.4 mm Undulator Gap
Streak camera IMAGES shows that this ion instability is Fast Beam Ion Instability: the tail part of a long bunch train oscillates vertically.
There was no appreciable difference at the different fill patterns, because of localized ions.
Streak camera image
1 us
Vertical direction
The tail part of the bunch train is oscillating vertically.
T.-Y. Lee
Pohang Accelerator Lab.

17. FBII at 5mm Undulator gap
1 us
5ms
Beam loss is mostly at the tail
T.-Y. Lee
Pohang Accelerator Lab.

18. Ion Instability during the revolver gap change
Above gap 7mm, no instability and no lifetime change
Below gap 6.4mm, transverse ion instability appeared and then beam loss occurred.
Beam loss occurred as well as lifetime decreased rapidly
~ 5 Hours Electron Beam 5 mm Undulator Gap !
T.-Y. Lee
Pohang Accelerator Lab.

19. Ion Instability: Bunch Current
Before the instability
After the ion instability
1 s
1 s
The bunch current of long bunch train was scraped off to a triangular shape.
The physical aperture of the storage ring reduced to the Revolver gap.
T.-Y. Lee
Pohang Accelerator Lab.

20. Measurement of Beam Oscillation with turn by turn DBPM
Gap: 6.0 mm
Gap: 7 mm
Gap: 6.5 mm
Gap: 6.4 mm
Gap: 20 mm
300 um
T.-Y. Lee
Pohang Accelerator Lab.

21. Vertical oscillation when the gap is 6 mm
The digital BPM data in the figure is an average value of all bunches for each turn
Vertical oscillation amplitude of the tail of the bunch train is twice the peak value in the fuigre.
1 ms
3 ms
2 ms
Growth time of instability : 1 – 3 ms
Transverse damping time of PLS: 8 ms
T.-Y. Lee
Pohang Accelerator Lab.

22. Vertical Beam Size Dependence
The vertical bema size dependence is the key issue in relation to the ILC damping ring.
It is very difficult to decrease the PLS vertical beam size, while it is easy to increase it using skew quadrupoles.
By using these quadrupoles, the vertical beam size was increased 80 % as measured by X-ray diagnostic beamline. With the same conditions of pressure and bunch current, FII was not observed.
Eventually, we want to draw the FII threshold on P-s graph or n-s graph.
T.-Y. Lee
Pohang Accelerator Lab.

23. Summary and Future Plan
FII has not been observed in any existing ring in the vacuum pressure of normal operation (sub nTorr) except the ion frequency. It has been observed only with deliberate vacuum pressure elevation.
The purpose of this experiment was to probe the threshold of FII by slightly changing the pressure and the vertical beam size.
FII can be excited by high pressure localized only in a small area of a ring. This indicates that FII is potentially more dangerous than previously thought.
Vertical beam size dependence of FII was observed qualitatively.
More experiments will be done with refinement. Different bunch spacings (4 ns, 6ns..) will be tried.
Thank You for Your Attention!
T.-Y. Lee
Pohang Accelerator Lab.