Notes from Main Injector/Recycler Group Meeting Monday, 30 September 2002 Bruce C. Brown Terry Anderson -- Transfer Line Vacuum Window Issues The issues raise recently about the impact of scattering in the vacuum windows of the MI <=> RR transfer lines has resulted in an examination of options for changes. The removal of the window would require differential pumping to protect the RR vacuum from the gas load of the MI and transfer line vacuum. The possibility of putting an impedance limiting partial window (with a hole to equalize pressure rather than support a difference of one atmosphere) was briefly mentioned. The RR vacuum is now 1E-10 in or so in some places and one should design for the possibility of 1E-11. The RR20 (700) transfer line has readings of about 1E-8 Torr while the RR30 (800) transfer line has readings of about 5E-8 Torr. The configuration now in place has a pair of mirror magnets at the RR end of the line with vacuum window between them. If one were to remove the window, one would surely place a valve in that location. One would then wish to consider a massive differential pump at the MI end of this pair of mirror magnets. The impedance limiting window option would put a valve plus the window at the location of the present window. It could have a small hole to equalize pressure (avoid being blown through) and be very much thinner (lower radiation length => smaller scattering angle). The massive pump which was discussed has water as its largest gas load. If one uses 100" along the beamline and achieves 100" of effective perimeter (corrugations along the wall and one or more added 'walls' to provide suitable area. It can employ TSP (Titanium Sputter Pump) or NEG (Non-Evaporatable Getter) technology. With 1E4 sq. in of pumping and 1 - 3 liters/sec/in2 of pumping capacity (typical TSP design numbers) one pumps 1E4 liters/sec of water. The expected gas load (from existing pressure measurements) is 6E-7 Torr-liters/sec at 6E-11 in the new pumping device. At a location in the beam pipe 3 m toward the MI, one needs to achieve a pressure of 3E-8 to achieve these conditions. In addition to the massive pump one needs 300 liters/sec of ion pump. Detailed review will need to include issues of what gas load is experienced in various failure modes. For example if the ion pumps upstream go off (power failure) the massive pump will maintain RR vacuum isolation but only for a brief period based on the total gas load experienced. This will set a requirement for the valve closing speed (and other design issues). Discussion: Stan Pruss asked how many cycles are fast-acting valves designed to experience? Is it sufficient so that the value can be cycled for each beam transfer (perhaps every 10 minutes in some suggested operational modes) and left closed between transfers? Terry considered it. [Response next day: Cycling the valve 50000 times a year is not a good thing.] Carol Johnstone has examined the situation at all the beamline vacuum windows and offered to provide interested parties with a summary table. Bill Foster suggested that the same sort of problem had been examined and solved at Cornell to protect the vacuum at the interaction point. (Per SPIRES I found:) A NOVEL DIFFERENTIALLY PUMPED UHV FLANGE IN THE CESR INTERACTION REGION. By J. Cherwinka, Y. He, Y. Li, N. Mistry (Cornell U., LNS). PAC-2001-RPPH120, Aug 2001. 3pp. Presented at IEEE Particle Accelerator Conference (PAC2001), Chicago, Illinois, 18-22 Jun 2001. But Bill said his reference was: MASSIVE TITANIUM SUBLIMATION PUMPING IN THE CESR INTERACTION REGION N.B. Mistry, R. Kersevan and Yulin Li http://accelconf.web.cern.ch/accelconf/pac97/papers/pdf/8C006.PDF John Marriner -- Recycler Ring RF Report John's slides are available in the meeting notes web area. http://www-ap.fnal.gov/~bcbrown/MtgNotes/RecyclerRF_Report.ppt http://www-ap.fnal.gov/~bcbrown/MtgNotes/Long_Emit_Growth.pdf Also included in Beams Document Database as http://beamdocs.fnal.gov/DocDB/0003/000382/001/Long_Emit_Growth.pdf http://beamdocs.fnal.gov/DocDB/0003/000382/001/RecyclerRF_Report.ppt These notes will only include some of the details not written on the slides. Brian Chase observed that some of the issues about low level analog response could be addressed by employing improved electronic components. This has been addressed. High level work includes understanding of both linear and non-linear distortion where the non-linear distortion mostly produces rounding of the response to abrupt changes. The linear distortion includes the phase distortion vs. frequency. This produces a h=1 sine wave response which creates some of the more obvious problems we observe. Current planning suggests designing for linear response with of order 1% errors (6% errors now observed). A new document -- "RF Operations in the Recycler" has been prepared to provide specifications of the Recycler RF control sequences. In other machines such as the Main Injector or the Tevatron, the LLRF software works with 'states' which accomplish major tasks such as 'accelerate protons to 120 GeV for PBar production'.... whereas the major RR LLRF 'states" define changes which would correspond to transitions between the 'states' of the other machines. This software controls manipulations between major modes -- 'Squeeze Cooled Beam for Injection' -- a distinction worth keeping in mind. This document describes newly standardized barrier size and positions. The barrier buckets will be 48 buckets long and 2 kV high -- sufficient for 18 MeV. Recent measurements have included studies of the shape of proton bunches stored in barrier buckets. By measuring the time deltaT from 0 to full charge at the edge of a bucket, knowing the voltage of the barrier, one can obtain a measure of the longitudinal emittance. A plot of data analyzed in this way showed an emittance which rose from 14 eV-s (1 sigma or 66%) to about 35 eV-s with most of the growth in the first half of a 6000 second plot. The analysis used time information only from the beam monitor, not including the time information from th RF waveform. The value of deltaT grew to 230 ns. The available time in the barrier bucket is 900 ns. Discussion: This was done with MI ramping including $2B (protons to 150 GeV) cycles. It is known that the orbit effects of 150 GeV ramps are about x2 greater than the effects from 120 GeV ramps. The separation between orbit change effects and betatron acceleration effects have not been carried out. No measurements of the phase of the beam wrt the 2.5 MHz RF have been employed to quantify the orbit path difference effects (a beam phase signal is not created in existing hardware). Guan Wu affirmed that the programs for implementing ramp control of both tune and orbit against changes due to the MI ramps are available to be tested. Shekhar -- Plans Expect to do a PBar transfer off of the bottom of the Accumulator stack (5-6 mA of pbars) today. Plan to use protons with 53 MHz buckets for tuning tomorrow, then do an additional PBar transfer later tomorrow. This will be followed with a 120 - 150 mA (1.2 - 1.5E12 pbars) transfer on Wednesday. Other machines will have an access but we will keep our pbar stack through their Wednesday shutdown.