RF Gun Control Concepts by rub18840

VIEWS: 0 PAGES: 6

									                                                                              3 August 2001 3:53 pm


                              RF Gun Control Concepts
                       G. Möller, W. Riesch, S. Simrock, T. Thon, H. Weddig

1.0 Requirements
The rf gun must demonstrate performance adequate for operation with the TESLA FEL. This
implies low emittance beam ( < 2 mm mrad) and operation at a frequency of 1299.9996 MHz.

Remember : THINK BIG

1.1 Amplitude and Phase stability
see XFEL Photoinjector Report TESLA-FEL 2001-03

Amplitude stability:      +-0.25 %           ( note: originally +-0.5 %)
Phase stability :          +-2.0deg.

Notes:
1. Stability intra pulse and pulse to pulse (for 1 day)
2. For gradients between 35-60 MV/m
3. Phase stability relative to M.O.
4. Repetition. rate 1 Hz - 10 Hz, pulse length 100us-800 us
5. Time for required for field stabilization during flat-top 50 us.

1.2 Operational
 1. Set constant forward power
 2. Control constant field
 3. Must be robust. Control must be guaranteed for large range of operating parameters
    such as rep. rate, pulse length and power.
 4. LLRF provides setpoints, application programs such as conditioning written by others.
 5. Measure resonance frequency with high accuracy
 6. Provide loaded Q with high accuracy
 7. Provide calibrated field with 5% accuracy
 8. ON-OFF knob.
 9. Measure resonance frequency of 0-mode. (delta_f = -5 MHz).
 10. Correct for directivity of directional coupler
 11. Support cavity conditioning needs.

1.3 Interlocks
Carsten Mueller (x 2612)
 1. rf gun reflected power

1.4 Monitoring
 1. klystron forward and reflected power
 2. circulator forward and reflected
 3. rf gun forward and reflected
 4. circulator load power
 5. preamplifier output

1.5 Temperature Controller
Olaf Krebs MKK ( x2018) + company (Frank Stephan)


 /home/simrock/doc/frame/requirements/rf_gun_req.doc                                        Page 1
                                                                               3 August 2001 3:53 pm


2.0 Control Issues
2.1 How to measure accelerating field
Only measurements of forward power and reflected power are available.
Derived measurements can be cavity field, detuning and loaded Q.

Measure amplitude of incident wave at directional coupler close to cavity. Crosscheck with
monitor for forward power at klystron and peak reflected power at cavity immediately after
turning off.


2.2 Sources of rf field perturbations
 1. Changes in resonance frequency (temperature). A change of the rf gun temperature of
   1 deg. C results in a frequency change of 20 kHz. This has to be compared to a
   bandwidth of 65 kHz (HWHM). A change of 1 deg. results in a phase change of 15 deg.
   and 3.5% amplitude. Recommendation: Digital rf control should monitor detuning and
   gun temperature and apply appropriate adjustment of incident wave.
 2. Changes in loaded Q
    - Does cathode change result in change of loaded Q ?
    - Expect changes of the order of 1% (changes in coupling) which result in field changes
       of same order of magnitude at constant forward power.
 3. Klystron power stability
    - Stability of klystron gain: ~V^(5/2) i.e. 1% change of klystron voltage results in 1.25%
       amplitude of incident wave,
    - phase stability of klystron: 1% change of klystron voltage result in 14.0 deg. phase
       change
Note: Stability of klystron voltage is specified to be < 1%

Recommendation: Digital feedback should measure HV and apply appropriate corrections
to klystron drive signal.

4. Beam loading
    - Beam loading is very small. P_beam = 10mA*3MV = 30 kW to be compared to 3-5
      MW of incident power. A change of 1% in power results in 0.5% in amplitude.

2.3 Miscellaneous
- second harmonic may be large at output of directional coupler
  (it may be necessary to insert lowpass to suppress harmonics and spurious signals)

3.0 Concept for field control
The control of the cavity field requires an rf signal which reflects the actual field experienced
by the beam. Usually the a field probe provides the signal needed for control. In case of the
Zeuthen rf gun there is no field probe available to ensure that the field on axis is not distorted
by the asymmetry of the probe coupler.

The lack of a cavity probe signal dictates the use of forward and reflected power to calculate
the actual cavity field.




 /home/simrock/doc/frame/requirements/rf_gun_req.doc                                         Page 2
                                                                             3 August 2001 3:53 pm




                                      Vcav                       Vref


                                            vfor
 Figure 1: Relation between cavity voltage and incident and reflected wave
Since the cavity is supposed to be operated on resonance, the reflected power will be
constant, and it will be sufficient to maintain constant forward power. For more precise
regulation the reflected power must be considered. This is especially true if the resonance
frequency of the cavity is not well controlled.


 1300MHz                                     Klystron


    ~          IQ-Mod.                                                          RF Gun


                                                        Pfor   Pref
                                            Pfor
                                                                                Temp.
                        LLRF                                                    Control
                                             Pref

  Figure 2: Block diagram of RF System for rf gun.

Since the rf gun requires watercooling (rf losses up to 50 kW) and exhibits a temperature
sensitivity of 20 kHz/deg. C, a water temperature control scheme has been implemented.
Presently the water temperature is controlled for a narrow range of operating parameters
while in the future it will be necessary to control the cavity detuning with high precision and
maintain stable control for a large range of operating conditions (rep. rate, pulse length, and
gradient).




 /home/simrock/doc/frame/requirements/rf_gun_req.doc                                       Page 3
                                                                                       3 August 2001 3:53 pm




                          P=3-5 MW                                 35-60 MV/m              RF GUN
       A&P or I/Q
       Controller                                                       D=35 dB
                                              D=25dB

                                                  50dB             50 dB
                                         Pr       Pf                   Pr         Pf




                                                        A&P or
                                                        I&Q
                                                        detector


          Feedforward         Gain
                                                Setpoint Vector

      Figure 2: Basic Feedback/Feedforward scheme


3.2 Design Choices and criteria
Analog versus digital
   - if repetitive error sources are dominant use of feedforward is recommended
   - if large feedback gain is needed (random errors) select analog feedback (minimizes
     loop delay to about 1us.
I/Q versus A&P
   - Amplitude detector provides more precise measurement of amplitude than I/Q
     detector.


                                                       Ampl.
  RFin                                                               Amplitude and Phase
                                                       Phase         detector


                                    LO



                                     Q

                    90o                                        LO (1300 MHz)

  RFin              0o
                                                                            I and Q detector

                                     I
          Figure 2: Basic rf signal detection schemes for forward and reflected power



 /home/simrock/doc/frame/requirements/rf_gun_req.doc                                                 Page 4
                                                                           3 August 2001 3:53 pm


3.1 Field control
1. Initially stabilize forward power
2. Later control cavity field. Calculate from forward and reflected power. V_c = V_f - V_r

Recommendation:
1. Digital control concept. Desired sampling rate 10 MHz (1 MHz may be ok) and latency
  1 us (4us may be ok). Dominating error sources are of repetitive nature or predictable
  (function of klystron voltage or gun temperature). This implies the use of feedforward
  which can be best realized in a digital system.
2. I/Q Control for incident wave or calculated cavity field. Control filter: Proportional
    +integrator; unity gain 100 kHz
3. Amplitude monitoring with precision temperature stabilized detector. Used for
    crosscheck and slow corrections.
4. Monitoring of all forward and reflected waves (vector !) with I/Q detectors.
5. Monitoring of gun water temperature
6. Monitoring of cavity detuning and loaded Q. Need 10 MHz sampling rate (compare to
    cavity time constant of 7 us).
7. Real time measurement of high voltage pulse (1MHz). Will be used for fast correction of
    feedback.
8. Control system access to actuators and sensors for gun water temperature control.


3.2 Resonance control of cavity
Gun (water) temperature control by MKK and external company.

The rf gun temperature control must deal with large variations in average power. Presently
the rf gun at DESY is operated at trep=1 Hz, tpulse= 100 us, Ppeak=2.3 MW corresponding to
an average power at 230 W. TESLA design is trep=10 Hz, tpulse= 1000 us, Ppeak=8.0 MW
corresponding to an average power at 80 kW.

For resonance control the following is recommended:
1. Instead of temperature regulation the cavity detuning should be controlled.
2. Detuning control must work for large range of parameters (average power !) including
    missing rf pulses and frequent changes of parameters.
3. Rf gun should be modified to incorporate a motor driven frequency tuner (stub-tuner or
    deformation of cavity walls) to provide fast and definite control of cavity frequency.

4.0 Hardware Choices
1. Phase detector : AD8302
2. Amplitude detector (log. amp) : AD8314
3. PSK demodulator (MAX 2105) for I/Q detection up to 1200 MHz
   digital, only 30 dB signal/noise ratio
4. Amplitude detector HP diode 333 C or similar




 /home/simrock/doc/frame/requirements/rf_gun_req.doc                                     Page 5
                                                                                               3 August 2001 3:53 pm


5.0 Task List


                                                 Table 1:

                                                   cost
                    Item                 P. w.                                     Remark
                                                  [kDM]

1.    Prec. detuning detector              2          1     Test at DESY Gun. Need 100 Hz resolution i.e.
                                                            0.036 deg. / us. Goal use AD8302. Cal. mV/deg.

2.    Schottky detector                   1           1     Independent amplitude and phase detector

3.    DSP board (C67) M67 +Omni-                      45    Development of next generation digital feedback
      bus modules ADC +DAC                                  based on C67.

4.    Monitoring of various signals                         forward, refl., temp., klystron volt. etc.

5.    DOOCS acc. actuators + sen-                           for study of improved gun resonance control
      sors of water temp. control

6.    Real time monitor for klystron
      high voltage

7.    Develop concept of improved
      gun resonance control

8.    Specification for A&P control
      algorithm for DSP code.

9.    Interface DOOCS -- and PC
      (compact PCI)

10.   RF Gun conditiong requirements




/home/simrock/doc/frame/requirements/rf_gun_req.doc                                                           Page 6

								
To top