A proposal for an improved laser system for the CEBAF photo-injector.

Thomas Jefferson National Accelerator Facility

Operated by the Southeastern Universities Research Association for the U.S. Department of Energy

A proposal for an improved laser system for the CEBAF photo-injector.

John HansknechtElectron Gun Group



A timeline of laser systems at Jlab

• Feb 1995. A 5mW He-Ne laser source produced the first photo emitted beam at Jlab. The beam was “DC” and the chopper “chopped” the beam for the three halls. There was no “tune” mode and viewer limited mode was achieved by inserting a neutral density filter on a pneumatic cylinder to reduce laser power.

Pros: We made polarized beam!

Cons: 1. Most of the beam produced was

thrown away on the chopper.2. Controls were not suited for

production beam.



1996

• April 1996. JLab source group, under the guidance of Dr. Charles Sinclair, was the first in the world to demonstrate high frequency polarized synchronous photoinjection from GaAs.

• The laser driving the gun was “state of the art”. A diode laser was rf gain-switched at 1497 MHz and subsequently amplified by a tapered-stripe laser diode amplifier [1].

• The laser provided tune and viewer limited pulse structures. These “Macro-pulses” were relatively easy to create electronically and met the requirements necessary for all beam diagnostics.

• This laser system was subsequently copied by other labs for use on their electron guns.

[1] M. Poelker, Appl. Phys. Lett ., 67, 2762 (1995).



The 1996 CEBAF Laser Table



1996 Laser Pulse Structure

Pros:

• Better photocathode lifetime vs. beam from a “DC” laser.

• Gain switching was simple.

Cons: “All for One & One for All”

•The chopper is still required to intercept beam for amplitude control.

• The current drawn from the photocathode needed to be 3 times the current requested by the highest current hall.

•Wavelength not tunable. Diodes and amplifiers were only available at two important wavelength ranges.



1997 Diode Laser system improvements

• The diode laser system was modified to provide 3 separate lasers, each pulsed at 499MHz and phased 120° apart.

• Space constraints forced the source group to design a new compact “seeded amplifier”.

• This design change was also needed to provide the ability to quickly swap lasers among the two wavelength selections.



1997 Laser Table schematic



Key points for 3 laser operationsBeam combining methods

http://www.jlab.org/accel/inj_group/laserparts/Beam_combining_tutorial.pdf



1997 laser table (open in lab)



1997 Laser Pulse Structure

• Beam amplitude is customized for the specific halls at the laser rather than at the chopper.

• Individual hall laser can be shut-off if hall does not want beam.• Individual Tune and viewer limited modes.• Most efficient use of the precious resource of electrons. -

Longest lifetime of photocathode.Beam is now being routinely delivered for physics from Bulk GaAs

T-Gun Broken.



1997 and 1998

New problems introduced:• ASE (Amplified Spontaneous Emission) is not our

friend:1. leakage to unintended hall

2. Polarization dilution 3. Tune mode cross-talk Laser powers of individual lasers are subsequently dropped to limit

ASE. • We are laser power limited further than before. • Beam coincidence • Vendor delivery problems• Dripping sweat kills lasers Changes are needed… soon

100uA 35% polarization delivered to HAPPEX from Bulk GaAs



1999 brings major changes

• Vertical Gun replaced with 2 horizontal guns (no more kneeling on the floor for laser work)• An air-conditioned laser hut (inexpensive plastic curtain) is

constructed to contain the laser table. (no more sweat dripping on the lasers)• Safety – No more vertical laser beam.• Ti-Sapphire lasers are in testing phase. New Strained layer

photo-cathodes require 10X more power than bulk GaAs for same current. We are severely power limited.

• No real changes to lasers other than physical layout.

50uA 70% polarization delivered to HAPPEX from Strained GaAs



1999 - Ready for some serious physics now



1999 laser table schematic



Y2K – Introducing the Actively-modelocked Ti-Sapphire laser (another first)

Jlab Patented Technology – C. Hovater, M. Poelker

(C. Hovater and M. Poelker Nuclear Instruments and Methods in Physics Research A 418 (1998) 280-284)



Y2K- Actively modelocked Ti-Sapphire laser schematic

Jlab Patented Technology – C. Hovater, M. Poelker



Y2K laser table schematic



Y2K Laser System

Pros:• Ti-Sapphire laser is wavelength tunable to reach peak polarization

of photo-cathode material• Ti-Sapphire laser provides high power as compared to diode laser

systems - (can deliver more Coulombs between cathode activations)

• Ti-Sapphire laser has no ASE

Nov. 2000. Delivered high current and high polarization to two halls simultaneously. (GEn & GEp) This would not have been possible without the new Ti-Sapphire laser.



Y2K Laser System

Cons:• Ti-Sapphire lasers are extremely sensitive to alignment and

cleanliness. 1um changes of cavity will affect lock. Small changes in room temperature will change alignment of the cavity.

• The cavity length sets the fundamental repetition rate of the cavity. Injection modelocking relies on many parameters being “perfect” to achieve a good pulse structure on the output.

• If laser phase lock is lost, the beam can be sent to the wrong hall(s).• Phase noise makes e- beam difficult to transport• Difficult to produce a “Tune” structure. Diode used: colinearity,

phase differences and amplitude differences caused problems.• Laser “on-call” is a full time job



2 Halls from 1 Ti-Sapphire Laser (Nov 2000)

plate

Polarized beam splitter

plateGlan ThomsonPolarizer

Atten C

Polarized beam splitter

Glan ThomsonPolarizer

plate

Atten A

500Mhz pulse trainLinear polarized from Ti-Sapphire laser

Path Length of assembly is adjustable so the returning beam will be delayed by 120 degrees of Rf phase

100% reflector

2% reflector

Small A-C lossLarge B loss

Atten B

plateGlan ThomsonPolarizer

100% reflector

Beam from Diode Laserfor Hall B. Phased 120 degrees from A and C

Either

ToPockelsCell and Cathode

100% reflector



2001 Laser System g0 preps



2001-2002

• New technology available- (SESAM technology) Commercial Ti-Sapphire laser that provides superior reliability and performance over our injection-seeded and AOM mode-locked Ti-Sapphire lasers.

• g0 31MHz experiment would not have been successful were it not for this laser.

• Laser was so successful that we purchased several for 499MHz as well.

• Laser table controls upgraded to provide fine control for the correction of current and position asymmetries.



2002 – Present Laser Table



Enhancements



2004- A Safer and Cleaner environment



Current Standings

Pros:• The year is 2005. The injector is generally not the source of problems

for the accelerator. • The injector is providing the highest polarization, highest current

synchronous photo-injected beam ever delivered in the world.• We are now capable of delivering parity quality beam to 3 experiments

simultaneously. (sort of)• The injector area is Safe, Clean, and Cool• A new load-locked gun is coming soon that will allow rapid exchange

of previously prepared photo-cathodes.

Cons:

• The source group has lost several key personnel and the present budget does not support replacement. Innovation & improvements now take a back seat to maintenance of the existing system. Everything takes longer.

• Although the Time-Bandwidth Products, Inc lasers are vastly superior to our previous lasers, they are still temperamental and require an expert to maintain.



Laser Specific issues that need to be addressed

1. Wavelength tunability and time required for changes 2. Phase noise & phase lock 3. Vendor spares4. Beam Colinearity5. Polarization dilution6. Tune mode quality7. Parity system quality8. 1.497 GHz stable laser for Accelerator tuning has been requested9. Laser Power limit10. Amplified Spontaneous Emission (ASE)11. Laser sensitivity to its environment and laser safety.12. Time consumed to replace lasers13. Only 2 Laser “experts” in the group. 14. Need to align multiple items on laser table.15. Confusion to operators when lasers are changed and have different

performance characteristics.



Let’s build a new laser system from scratch.

• Our first point of action is to select a wavelength. Our recent success with the “super-lattice” cathodes proves that 780nM is ideal for high polarization and excellent QE.

• We are sticking with this decision and will place this cathode material in both guns, so our laser system will deliver 780nM light. (scratch #1 from the list)

• We liked our gain-switched diode technique because of its phase noise and phase lock attributes. Let’s start with three laser “seeds” and gain switch them. (scratch # 2 from the list)

• We will be using lasers and fiber laser amplifiers designed for the cable TV and communications industry. (scratch #3 from the list)

• We are going to use new fiber laser technologies that allow us to combine all lasers in a single fiber with identical polarization and collinear travel. (scratch #4 and #5 from the list)

• We will use fiber-based electro-optic modulators with GHz bandwidth, so we should be able to produce any desired tune mode or other modulation on the beam with high quality. (scratch #6 and #7 from the list)



New laser system design. Lets see what we have so far.

Pre-Amp

Pre-Amp

Pre-Amp



More design thoughts

Since combining these lasers appear to be so easy, let’s throw in a fourth laser at 1.497 GHz. We will use a fiber based MEMS optical switch to efficiently swap from the 3 laser system (500MHz) to the single 1.497 GHz laser with the press of a button.

(and scratch # 8 from the list)



Laser system design continued:

Pre-Amp

Pre-Amp

Pre-Amp

Pre-Amp



Now we need some Power

• Thus far our system has four lasers that provide clean gain-switched light. We need to amplify it further to get sufficient power for operations.

• Erbium-Ytterbium fiber laser amplifiers are now commercially available. Their power level capabilities have been growing exponentially over the past few years. They will meet our immediate demand (for a price), but will become more powerful and cheaper as the technology and market demand grows.

• Previous worries about delivering quality single-mode TEM00 beam from a fiber are gone. New “Panda” fiber designs transmit pure single mode beam without worry.

We are specifying an amplifier that should triple our present deliverable laser power. (scratch # 9 from the list)



The final pieces of the system



The final system operation• Light was produced with proper pulse structure, intensity control, modulation

control.

• Light from all lasers was amplified and some ASE is present in amplified 1560nM light.

• Non-linear Second Harmonic Generation (SHG) crystal is used to frequency double the light from 1560nM down to 780nM. Non-linear gain of SHG crystal will cut off and not pass the low levels of ASE. (scratch 10 from the list)

• Light for all halls through the SHG crystal is linearly polarized and perfectly round. Now we can place a LP optic immediately before the helicity control “Pockels Cell” to obtain the highest purity polarization possible.

• All components are designed with quick disconnect polarization maintaining fiber connectors. When connected there are no laser safety issues except for the area of the fiber to air launch to the SHG crystal and subsequent beam delivery optics. An “expert” can be anyone with training on the system. (scratch the entire list)



Final laser system operation continued:

• The commercial fiber lasers, amplifiers and modulators often come with monitoring ports installed. There will be multiple points within the system to verify system operation and laser beam quality.

• We will be producing much more light than is needed, so we will now be able to afford placing fast photo-diodes and power taps at the output for phase feedback monitoring and control

• The system will consist of 19” rack mounted drawers that can easily be interlocked to power off when the lid is opened and thus eliminate any laser hazard.

• The main laser system could be remotely located (upstairs in the service building) and the main delivery fiber can be fed to the tunnel through a conduit.

• Operators will be able to select any laser for any hall. There will no longer be any confusion over the capabilities of a given laser.



System is compact and easy to swap components



Possible Pitfalls

• This is pure R&D. To the best of our knowledge it has never been done before and may not work exactly as planned.

• Communications laser companies are making big $$ from communications users. They have no interest in pursuing our little project, but have been helpful in offering to sell us components.

• There is a possibility that our mode of operation and PSS/FSD protection could change. Example: If ASE passes amplifier when a given Hall is in Beam Sync, we would need to secure all halls by securing the main laser amplifier until the chopper slit could be fully inserted. This would be very similar to how we used to run the thermionic beam.

• We may find temperature induced phase or mode variations in the fiber system that we have never experienced before in a free space system.

• The halls will lose their ability to independently move a PZT mirror for their hall. It is envisioned that one PZT mirror would serve all halls and the 30Hz PZT functions.



Laser system wide view



Fiber laser launch on table



What’s next?

• Matt Poelker and I will be performing laser studies and procuring components.

• Our group is short staffed and we need another PhD. One might consider finding one with experience on fiber lasers.

• We need input as early as possible from anyone who has any special needs for the beam. (i.e. special modulation schemes) So these can be designed into the system.

• If we want a quality product in the shortest amount of time we will need to form a team that consists of:

1. Electrical engineering support 2. Software support (drivers and screens) 3. EECAD and FAB support 4. Rf Engineering support 5. PSS/MPS system support and…………….



Actually much less, but we do need the labs commitment for funding of the project….

One Million Dollars!

Which we shall call…“The Alan Parsons Project”



Actual cost rough analysis (optical components)

$15K to 30K



Rough cost analysis

$57K

$6K

A proposal for an improved laser system for the CEBAF photo-injector.

Documents

Transcript of A proposal for an improved laser system for the CEBAF photo-injector.