CODD system proposal

Creation of a beam synchronous reference frequency

This part derives a beam-synchronous reference frequency at h=512. Referring to the figure, a limiting log-detecting amplifier squares up the signal from a WCM. An approximation of the revolution frequency is provided by a frequency table, indexed by the B-field. Its value is multiplied (digitally) by the current h and converted into an RF signal by a DDS. The phase of this DDS is compared with the signal from the WCM, and a correction is numerically added to the table output. At point 1, we now have a digital representation of the revolution frequency, phase locked to the beam.

The value at point 1 is then digitally multiplied by 8 and applied to a 2nd DDS, yielding a beam-synchronous h=8 signal. This is then multiplied up to h=512 by a classical PLL that must cover about one octave. (At injection, β for ions is about 0.5, and it's essentially unity at ejection for protons.)

Prior to injection, and during harmonic changes, the B-field is constant and the h=8 RF signal provided by beam control can be relied upon to have a fixed phase relationship with the beam. The PLL can thus be switched to that frequency during those times. A programmable phase shifter (point 3) is set for each of those occurences so as to minimise the phase jumps seen by the PLL. The phase jumps can be observed at point 2.

The limiting amplifier delivers an RSSI signal which is essentially the logarithm of the beam intensity (point 4). A useful signal for troubleshooting. To aid in remote diagnostics, this signal, as well as that of the PLL phase detector, should be digitised and made available for readout by the DSC.

There are two problems with this scheme. First, the limiting amplifier connected to the PU introduces new odd harmonics, which weren't in the orginal PU signal, and it obscures others. Second, by mixing down to DC, it's no longer possible to fix the loop gain by servoing the IF amplitude. The phase error signal and the 'IF' are both at baseband.