Two questions that frequently come up are “What is the difference between the front end phase and the ADC timing? How do I set them up correctly?”

In this note I'll describe what these critical settings control and how to properly adjust them.

Bunch-by-bunch feedback based on iGp12 typically uses FBE-LT front end to detect beam signals. The front end performs BPM pulse stretching/replication (bandpass filtering in the frequency domain) and downconversion from the detection frequency to baseband. It is important to understand the signal flow in this setup to be able to configure its settings.

The figure above shows a simplified block diagram of the feedback input section with the signal shapes sketched at various stages.

One important distinction must be drawn between the longitudinal and the transverse planes. In the longitudinal plane the input signal is the sum of all pickups (to eliminate sensitivity to transverse position). Thus, the front end always operates with a significant input amplitude. In the transverse case, the input signal is the difference of top/bottom or inner/outer pickups. Signal amplitude in this case depends on the orbit offset and can often be close to zero. Consequently, some of the tuning shortcuts, possible in the longitudinal plane, are impossible in the transverse case.

The main goal of the tuning procedure is to iteratively adjust the local oscillator phase (front end phase) and the ADC timing to maximize the sensitivity to beam motion at the synchrotron or betatron frequency. Note that DC offsets, observed in the mean waveform, are not always related to this maximization process, especially in the transverse case.