Low Level RF Control

Paul Corredoura


The talk began with describing the differences between the PEP-II LLRF system and what exist currently in the damping rings. The PEP-II system resides mainly in a 13 slot VXI crate as a family of modules which employ baseband IQ (in-phase and quadrature) signal processing. The basics of IQ techniques are described including phase/amplitude detection and the implementation of a built-in network analyzer using dedicated hardware and Matlab. The system is completely programmable via a well designed EPICS interface.

Issues are listed which must be dealt with before a system could be built for the damping rings. The main difference is the pulsed aspect of the damping rings. The measurement of the cavity loading angle for the tuner loops would need to be synchronized with beam storage. The arbitrary complex baseband signal generator used in the network analyzer could be used to modulate the RF system to shift RF frequency, bunch munch and pre-phase the cavity RF before beam injection but again this function would need to be synchronized with the beam injection/extraction. All these changes are very reasonable but they would need to be done before building any new hardware.

The overall topology of the system is described including the shielded RF rack, VXI crate, Allen Bradley PLC system and the local panel. A very busy block diagram showing all the main components and how they are interconnected is presented. This is followed by a signal flow block diagram describing how RF signals are handled in the system to implement direct RF feedback, s-band beam phase feedback, klystron phase compensation, and the baseband network analyzer (complex signal source as well).

Some of the actual PEP-II RF modules are described. Pictures of the hardware, test data and the associated EPICS control panels are presented. Examples of what automated calibration routines and diagnostics can be created with the network analyzer and Matlab scripts including automated tuner loop and direct RF feedback configuration are shown.

The baseband PID controller for the direct RF feedback loop is described. This technique allows for large loop gain (>40 dB) at the carrier frequency and increased bandwidth over the same loop implemented at the RF carrier frequency.

Examples of using the built-in network analyzer as a spectrum analyzer is presented demonstrating a dynamic range > 100 dBC. Several example EPICS panels showing that controls are entered in real engineering units. For example gain of the direct RF loop is set by typing in the desired gain in dB and the loop phase in degrees.

The talk concludes with a list of channels which would likely be monitored if an Allen Bradley system were used to handle the slow interlocks. Based on this list, the previously described RF system and a "new" S-band beam phase module, the cost of designing and building such a system is estimated (including software).