D.T. Corcoran, G.C. Harper, D.W. Storm and K.B. Swartz

The concept of using a high energy buncher operating at a frequency equal to the sum of the linac frequency and the low energy buncher frequency was presented last year¹. The construction of the new resonator by modifying one of our low- resonators was described in that report. The resonator was successfully plated with a lead-tin mixture (1.8% tin by weight) using a technique similar to that used at Stony Brook². This resonator exhibited substantially more severe multipacting than the low beta resonators used in the linac. We were able to overcome this multipacting using the freon conditioning technique first described by Noé. The resonator has a low-field Q of 1.8 × 10⁸ and can reach 2.5 MV/m at 10 W. This is not as good as most of our low- resonators, but is more than sufficient for the buncher.

In order to operate the resonator in the linac, new electronics were developed which produce a clock signal at 13/12 times the linac frequency . The same unit produces the signal for the low energy buncher at 1/12 × . The power rf electronics we use in the linac are sufficiently narrow-band that they will not operate with this buncher, so we obtained an appropriate 100 W amplifier and circulator. The resonator control electronics are the same as was used with the previous high energy buncher.

With the previous high energy buncher operating at , the linac pulses between the bunches produced by the low energy buncher had small amounts of beam present, typically 1% of the main bunch in each of the 11 intermediate bunches. These small bunches resulted from the high energy buncher acting on the continuous beam that lies between the bunches produced by the low energy buncher. Because the new buncher is only in phase with the linac when the main bunch is present, much less beam is captured into the intermediate bunches. The reduction is typically a factor of 10 or more, as is illustrated in the figure. We thank Felix Liang for the figure.