EuPRAXIA_PP WP9 fifth meeting


Zoom link:

Meeting ID: 954 4945 8932
Passcode: 6211

Present: Federico Nguyen (ENEA), Sergey Antipov (DESY), Edda Gschwendtner (CERN), Andrea Latina (CERN), Enrica Chiadroni (La Sapienza), Luigi Faillace (INFN), Thorsten Lamb (DESY)


RF system choices for the EuPRAXIA 1st site -- E. Chiadroni (La Sapienza) & L. Faillace (INFN)

The SPARC_LAB injector shall produce ultra-short, ca. 7 fs, high brightness electron beams. It consists of a 1.6 cell UCLA/BNL-type standing wave RF gun followed by a 11-cm-long X-band linearizer, and 4 travelling wave SLAC-type S-band structure. The first two S-band cavities are surrounded by solenoids for the operation in the velocity bunching scheme.

  • The S-band lengths, and voltages have been optimized to provide the right separation and duration of the driver and witness bunches. The first S-band cavity operates at 90 deg off crest for RF bunch over-compression.
  • The X-band linearizer has an iris radius of 4 mm. It is operated in the SW mode thanks to the low required power.

The injector is followed by an X-band booster. The X-band cavities are powered by an 25 MW RF power source, consisting of a Solid State Pulsed Modulator and an X-band klystron. Installation of a PSI design BOC pulse compressor to further increase the peak power by about a factor of 3 is considered as an option.

So far only short prototypes of X-band structures have been tested with RF power. 

  • A 20-cell structure has been tested at low power, in preparation for tests at high power
  • A full-scale 90-cm-long prototype is in production

Sergey raised a question whether the operating gradients in the RF cavities are too relaxed. 

  • Luigi explained that while the X-band cavities have demonstrated much larger gradients, those were achieved in shorter samples. As EuPRAXIA@SPARC_LAB is the first project employing long multicell structures on a large scale, 60 MV/m has been chosen as safe, conservative target.
  • Enrica pointed out that the first two S-band modules operate at a reduced voltage to achieve required over-compression and temporal separation of the two bunches.



Laser (to RF) Synchronization -- T. Lamb (DESY)

In order to achieve low drift operation the oscillator synchronization is only one piece of the puzzle - one also has to address RF distribution and laser stabilization and transport. Otherwise it might not be justified to purchase and install a complex and expensive synchronization system. 

  • The reference oscillator will determine the jitter or the facility. For smaller facilities commercial main oscillators are available. 
  • Classical distribution is done with RF cables. This is a cost-efficient method suitable for smaller facilities. The distribution can be actively stabilized, although the solutions are not commercially available presently. The downside is that RF cables pick up phase drifts, which can be demanding. A figure of merit is 20 fs/m/K. Climatization, temperature, and humidity may have an impact on the distribution. 
  • Alternatively, optical distribution can be done. This is done at a larger facilities like European XFEL.

When it comes to the synchronization itself, two solutions are available:

  • Direct conversion: robust and easy to implement, allows changing the laser timing with RF timing shifter but susceptible to long-term drifts. At DESY LUX facility 11 fs in-loop and 23 fs absolute jitter has been achieved with this method.
  • MZM-based laser-to-rf synchronization, where an optical pulse train is used to sample the 0-crossing of the RF phase. This method provides the best absolute performance with <  7 fs in-loop jitter reachable for single output. 1.9 fs short-term and 10 fs long-term stability achieved at DESY XFEL with dual output. The method requires a temperature stabilized and humidity sealed setup and works only for a fixed working point. 

Finally, laser-to-laser can be synchronized to 1 fs with an optical cross-correlator. At DESY 200 as has been achieved in a lab, and 1.1 fs across the 4 km long run of XFEL.

Sergey inquired about the approximate cost of the direct conversion system.

  • After the meeting Thorsten communicated that a commercially available RF sync setup plus a drift-free L2RF phase detector would amount to 60 kEuro (RF distribution extra). 
  • DESY LbSync group can readily provide advice and assistance in the specification of components in all synchronization related questions.
There are minutes attached to this event. Show them.