LLRF field control

In order to find the maximum achievable CW field stability HZB and Cornell started a collaboration to test the Cornell LLRF system at HoBiCaT. It is also the testbed to implement different microphonics compensation algorithms and determine the maximum loaded Q achievable with still stable operation regarding the required field stability. Figure 3 shows a scheme of the Cornell LLRF system.


Scheme showing the working principle of a digital LLRF system.


Still, modern ADCs cannot sample the RF at 1.3 GHz directly with the needed resolution of 14-16 bit. Thus the field information of the current amplitude and phase is measured by sampling the IQ field components – thus the field’s envelope- at four times an intermediate frequency (IF), which is a product of mixing the fundamental frequency with a by IF shifted second carrier. These field components are then processed by a simple PI control loop embedded on a FPGA chip.

QL sf (Hz) sf (deg) sA/A
5.107 9.5 0.008 1.10-4
1.108 7.9 0.009 2.10-4
2.108 4.2 0.024 3.10-4

Table 1 summarizes the achieved best field stabilities for high loaded Q operation under given microphonics detuning. Those values were obtained by 2-D proportional and integral gain scans of the field control settings. Figure 4 shows some time domain data of the cavity field phase for different gain settings. Current LLRF systems used at HZB are based on analogue technique (BESSYII) or digital technique based on VME or custom made rack electronics. For BERLinPro HZB plans to join the mTCA community to use this new standard of mTCA for physics. For more information visit DESY’s mTCA website given in the box to the right.


Residual RF phase error at 10MV/m for different gain settings.