RAL Template

1
Photoinjector Laser Development CARE PHIN
Meeting DESY HamburgNovember 2004 Graeme Hirst,
Marta Divall, Ian Ross, Emma Springate, Guy
Suberlucq, Roberto Losito
2
Outline

• Oscillator and preamplifier
• New system layout (advantages and disadvantages)
• High-power pump diodes
• Pumping distribution
• Head design
• Response to advisors suggestions
• Milestones and deliverables
• Summary

3
Initial system layout (May 04)
1.5 GHz NdYLF oscillator preamplifier
3-pass NdYLF amplifier x100
10 W 6.7 nJ/pulse
1 kW 0.7 ?J/pulse
200 ?s, 5-50 Hz
2332 e- bunches 2.33 nC/bunch
Feedback stabilisation
15 kW 10 ?J/pulse
2332 pulses 370 nJ/pulse
2?
4?
Optical gate (Pockels cell)
Energy stabiliser(Pockels cell)
Beam conditioner
4
Oscillator and preamplifier
Order was placed September 28th 2004, with
delivery in February 2005 Acceptance tests will
take place at CERN Support will be provided at
RAL after shipping HighQ will provide an
all-optical timing jitter measurement
5
Amplifier heat management

• Thermal lensing
• Fracture limit of YLF for rod geometry is 22
  W/cm

With 37 kW peak pumping power at 50 Hz the
thermal load is 21 W/cm assuming homogeneous
pumping along the length of the rod.
Fracture limit improvements for etched
rods Redesign the system with less power in the
final amplifier

• Possible collaboration with Vienna University of
  Technology
• Development of etching technology for
  strengthening of YLF laser crystals
• Development of ion exchange technology for
  strengthening of YLF laser crystals
• Combination of etching and ion exchange
  techniques for strengthening of YLF laser crystals

6
New system layout
1.5 GHz NdYLF oscillator preamplifier
3-pass NdYLF amplifier x300
10 W 6.7 nJ/pulse
3 kW 2 ?J/pulse
200 ?s, 5-50 Hz
2332 e- bunches 2.33 nC/bunch
Feedback stabilisation
15 kW 10 ?J/pulse
2332 pulses 370 nJ/pulse
2?
4?
Optical gate (Pockels cell)
Energy stabiliser(Pockels cell)
Beam conditioner
7
Calculated gain curves
20E02
16E02
12E02
8E02
4E02
0
0
200
400
600
8
Consequences of the new design

• Advantages
• Less power in the second amplifier
• safer thermal conditions
• 50 Hz operation is possible
• Higher overall efficiency
• Smaller gain and higher saturation
• in the second amplifier
• Similar power level in each amplifier allows
  easier thermal lens compensation
• (The two crystals can be rotated by 90o to each
  other)

• Disadvantages
• Two completely new amplifiers
• Higher cost

New specification for the diodes
9
Pump diodes
1st amplifier 15 kW peak power, 400 ?s, 50Hz
pumping
2nd amplifier 20 kW peak power, 200 ?s, 50Hz
pumping

• Asking for
• 18 kW peak power for long lifetime 109 shots
• 2.5 duty cycle for flexibility in pumping time
• Homogeneous pumping along the 7 cm
• length of the rod
• 5 fold geometrical configuration

• Asking for
• 22 kW peak power for long lifetime 109 shots
• 2 duty cycle for flexibility in pumping time
• Homogeneous pumping along the 11 cm length of
  the rod
• 5 fold geometrical configuration

Specification is being finalised for the
diodes Homogeneous pumping has to be
investigated A survey of the market has to be
carried out
10
Diode geometry
Mixed polarization - and to c-axis
Polarization onlyperpendicular to c-axis
More sensitive towavelength and sotemperature
Some parallel to c-axis ? higher absorption
efficiency
Too high power/length? several stacks needed or
biggerspacing
More flexible in power/length
The divergence angleis ideal for 1 cmdiameter
rods
Special microlensarrangement neededfor 6-8o
divergence
Difficult to calculatepumping distribution
Easy to calculate pumping distribution
11
Pumping distribution
Changing the position of the diodes
Measurements
Changing the temperature /- 5oC
LasCad calculations
MathCad calculations
12
Pump diodes

• Lensed option is expensive and does not guarantee
  good performance
• Buying more arrays to fill the length of the rod
  is cheaper than having a custom bar spacing
• The estimated cost for the two amplifiers
  together
• max. 400.000

The tender will be out by November 1st
13
Amplifier head design

• Simple water and electrical connections
• More accurate positioning of the
• focusing optics
• Easy assembly
• Possibility of rotating the rod in situ
• Similar design for the two amplifiers

14
Advisors suggestions

• Utilizing the energy stored in the laser rods
  during the total pump-up phase in a more
  efficient way would probably allow to reduce
  the required amount of diodes and therefore the
  total costs of the laser system. It would also
  solve the problems with
• fracture limit of the laser rods. This option
  should be taken into account because it is of
  large interest for other institutions.
• With burst-mode less pumping time is needed.
  However to get the same energy/pulse from the
  system (peak power) the same
• pump power is needed. Diode duty cycles are
  standardised, and diode lifetime depends on the
  number of shots, not the
• energy/pulse, so the qcw solution does not
  affect the cost. Burst-mode would help with
  thermal management, but in the new
• design the fracture problem is solved, and with
  the same thermal power/unit length in each
  amplifier the thermal lensing can
• easily be compensated. With steady-state
  operation the output is more stable against slow
  input variations, while in the
• burst-mode case these can cause high gain
  overshoot at the front of the train.
  Sophisticated feedback stabilization and timing
  control are therefore needed for safe operation.
• It may be worthwhile to consider an oscillator
  running in burst-mode instead of CW mode.
• This type is laser is not common on the market.
  Stable operation is guaranteed in a cw
  mode-locked laser.

Ingo Will, Armin Liero, Dieter Mertins, and
Wolfgang Sandner, IEEE JQE 34 (10) OCT
1998 Feedback-Stabilized NdYLF Amplifier System
for Generation of Picosecond Pulse Trains of an
Exactly Rectangular Envelope
15
Timescale and deliverables
16
Summary

• Tender has been carried out
• Purchase will be February 2005
• Laser oscillator test results

• New design for 50Hz operation
• Survey of high power pump diodes
• Tender action under way for the first amplifier
• Simultaneous full market survey for the second
  amplifier
• Preliminary designs for the head
• Fracture limit measurements
• Test results by autumn of 2005

The system to be delivered by 31st of March 2006

• Beam conditioning after the amplifiers
• Comparison of different nonlinear crystals
• Thermal management
• Test results by June of 2005

• Test of existing system
• Purchase faster electronics
• Compensation for nonlinear operation
• Test results by February of 2005