Booster RF Fanback System
Jim Steimel (12/92)
Introduction | Operations | Physics | Engineering | Testing
[ Booster Notes ] [ *Booster Home* ]



The purpose of the Booster RF fanback system is to provide a diagnostic for determining the phase and amplitude matching of RF at the cavities. Matching is very important for optimizing paraphase at injection and maximizing the total RF volts during acceleration. The signals are taken from the upstream gap monitors of the cavities, brought to the RCC racks through matched electrical lengths of cable and proper phase shifts, summed, and brought to the control room through matched cable.  These cables enter a box which provides the control room with the total RF Sum signal, the phase shift between A&B signals, and the diode detected RF Sum, A, and B signals.


During injection, there should be zero volts on the beam. This is accomplished by phasing the cavities so that the polarity of voltage that the beam sees reverses with each cavity. The paraphase module changes the phasing of the cavities after injection, and the beam starts to accelerate. It is important to retain the proper phase relationship throughout the Booster cycle using the paraphase offsets and curves.

The means to diagnose the cavity phasing is contained in MCRR7, and it looks like the following:

The A & B (50W ) outputs are 30-53 MHz signals which are proportional to the voltage provided to the beam from the A  cavities and the B cavities respectively.  RF Sum (50W ) is the difference between A & B because A & B are 180° out of phase during the accelerating cycle. The peak detected signals provide the envelopes of the A, B, and RF Sum signals respectively. RF Sum peak detected is also read by an MADC and Operations can be plotted on a console with B:RFSUM.  The ± 15V lights monitor the power to the peak detectors. A,B, and RF Sum (50W ) will still work if power is lost to the chassis.

The A & B (50W ) outputs are usually connected to a fast phase detector located in MCRR8. The phase detector outputs are high impedance and have a value from -10V to 10V with 0V corresponding to 0° and 180° phase shifts. The phase detector output can be plotted on the console with B:RFPDET.

A & B 50W outputs must be terminated with 50W terminators when not in use, otherwise the phase detector and the peak detected signals will not be correct. The RFSUM(50W ) output has a buffered output and can drive a high impedence load without affecting other measurements.  RFSUM(50W ) has a scale of 4.6 kV/turn / 1mV. B:RFSUM has a scale of
 0.81 kV/ turn/ 1 mV.

To maintain proper acceleration and control of the beam energy, a beam bunch must see the same RF phase at each cavity location. Since the beam has a finite transit time from cavity to cavity, the phase of the RF at each cavity at a given time must be different. The calculation is shown below.

vRF Cavity 1 = cos(2pfrft)
vRF Cavity 2= cos(2pfrft-f)
VRFCavity 1' = cos(2pfrf(t+td))
VRF Cavity 2' = cos(2pfrf(t+td)-f)
f = 2pfrftd
td = d/ v beam
v beam = 2pRBooster frf/h
f = d*h/ RBooster
Fig. 3 Cavity Phase Equations

The phase difference of the RF between two cavities is not a function of the RF frequency. It is only a function of the ratio of distance between cavities and the Booster radius. Thus, a wideband phase shifter, not a delay, is needed to properly phase the cavities. Figure 4 shows the phase relationship between all of the cavities in the Booster. 

The fanback system uses the downstream gap monitor in the Booster cavities to detect the RF voltage at the cavity gaps. The signals which are out of phase with cavity 9 by 28° or -152° are given a -28° phase shift. This provides two sets of signals which are 180° apart. The signals which are in phase with cavity 9 are summed together and provide the A Sum signal. The signals which are 180° out of phase with cavity 9 are summed together and provide the B Sum signal. Thus, the difference between A & B provides the RF Sum signal.

The RF Sum signal can also be used to estimate the total volts per turn seen by the beam. This information can be used to estimate the syncrotron frequency as a function of time in the cycle. RFSUM (50W ) has a scale of 4.6 kV/turn / 1 mV. B:RFSUM has a scale of 0.81 kV/turn / 1 mV.

The three key aspects for having a precise RF Sum signal are phase shifting, delay matching and amplitude matching. *Figure 7* shows a block diagram of the system.

The phase shift is produced by splitting the input signal with a 90° hybrid, attenuating one port, and vector summing the voltages. The -90° leg of the hybrid is attenuated enough, so that its vector sum with the 0° leg of the hybrid produces a vector which is -28° out of phase with the 0° leg. A detailed drawing of the components and their values is shown in the schematic.

The old system used a cable which was callibrated to give a -28° phase shift at 40MHz. This meant that the phase shift at 30MHz would be -21°, and at 53 MHz the phase shift would be -37°.  By replacing the cable with a wide band phase shifter, the tuning becomes much more accurate.

The attenuators in the block diagram are needed to amplitude match the signals which do not require a phase shift. Each attenuator matches the attenuation associated with the phase shifters. Their values are shown in the schematic.
All of the electrical delays from each cavity to the control room are carefully matched. The cables for the cavity gap monitors 1-16 to the RCC racks are matched, and the cables for 17 & 18 gap monitors are matched. Cables 17 & 18 are longer than cables 1-16 because of the extra distance they need to get to the RCC rack. The extra length for 17 & 18 are compensated by a shorter cable for the 17 & 18 signals from the RCCW rack to the control room. The electrical delays seen by each cavity from gap to RF Sum is shown on the *Figure 8* delay chart.

The outputs of the A & B (50W ) spigots are connected to the phase detector module with matched cable. Since the fast
phase detector is 0V for a 90° phase shift, matched 90° hybrids are placed on the inputs to the phase shifter to make the 0V crossing for a 0° & 180° phase shift.

The electrical delays of all the components of the system were measured with a network analyzer, and the results are shown in the graphs. Also, the attenuation and phase intercepts of the summer boxes are shown in the graphs.

Many of the cables and signal paths need to match in phase and attenuation. There are three ways to check this with a network analyzer. One way is to simply measure the delays and attenuations of the signal paths and require the numbers to be equal. Another way is to use the output of one signal route as the analyzer reference and the other signal route as the analyzer input. The measured response is the difference in phase, delay, and amplitude of the two routes. The third way is to use one signal route as the standard for a thru response callibration and measure the other signal route in place of the first. The output is equivalent to the second method.

All summer outputs are compared to each other, and all cables which go from the RCC racks to the control room are matched this way.

The "A Sum" and "B Sum" voltages were measured at each of the RCC racks.  Then, the output of RF.Sum (50W ) was measured with all but one cable removed from the control room sum box. Each cable was connected, the signal amplitude was measured, and the cable was disconnected until a ratio for control room amplitude to RCC rack amplitude was known for all A & B signals. The comparison is shown in Figure 6.

The final test checked the phase match between cables by looking at the phase difference between RF cavity signals at different times in the cycle. The total delay mismatch is determined by a linear extrapolation of phase difference between cables as a function of the RF frequency.

[ Booster Notes ] [ *Booster Home* ]




Security, Privacy, Legal