Independent Measurement of Two Beams in an IP Feedback BPM (response to a question asked at LCWS05 )

Size: px

Start display at page:

Download "Independent Measurement of Two Beams in an IP Feedback BPM (response to a question asked at LCWS05 )"

Julian Hunter
5 years ago
Views:

1 Independent Measurement of Two Beams in an IP Feedback BPM (response to a question asked at LCWS05 ) March 22, 2005 Steve Smith

IP Feedback in 2-mr Crossing Scheme Both incoming and outgoing beams traverse the first BPM Can we cleanly measure beam-beam deflection (outgoing

2 IP Feedback in 2-mr Crossing Scheme Both incoming and outgoing beams traverse the first BPM Can we cleanly measure beam-beam deflection (outgoing beam) shortly after passage of incoming beam through this BPM? Sneak preview of conclusion: Yes, it works with a substantial safety factor IF we re careful!

3 IP Feedback (TDR version)

4 IP Feedback Measure beam-beam deflection angle Correct beam-beam offset bunch-by-bunch Several BPMs on each side measure incoming beam One BPM on each side measure outgoing beams Recommend stripline BPM for BPMs which must measure both the incoming and outgoing beams. How well can two beams be separated in one BPM? Estimate using SLC experience Calculation

5 Directional Stripline BPM An unloaded stripline BPM with matched feedthroughs at both ends is directional (for β beam = 1) In principle, it gets perfect separation of beams going opposite directions. Have achieved about 30 db isolation at ~ 40 MHz (SLC IP example) Presumably can do better with careful design, machining, matching At issue is matching impedances to a few percent through feedthroughs, cable at low RF frequencies. 2% impedance matching yields incoming-outgoing isolation of 100. We assume here a factor of 30.

6 Directional Stripline BPM Acrobat Document

7 SLC IP BPM Directionality e beam electron direction output Vpeak ~ 250 mv electron beam positron direction Vpeak ~ 8 mv

8 Requirements Require system to have sufficient dynamic range to handle (average) offset beam amplitude ratio of ~2 for closest-stripline between beam at 1/3 Radius and centered beam OK Only issue is leakage of jitter of incoming beam into estimate of outgoing beam position.

9 Separation of Incoming/Outgoing Beams Separate incoming & outgoing beam signals with 4 tools: 1. Directionality of stripline Assume factor of 30 rejection of incoming signal 2. Temporal difference between incoming/outgoing BPM z=4 m => Dt=27 ns Assume incoming signal envelope has decayed by 1/10 3. Phase/orthogonality of sampled waveform i.e. choose analog filter so that the tail of the incoming beam signal is near its zero crossing when the outgoing beam signal is at its peak assume a factor of 3 isolation. 4. Prediction/feedforward compensation of incoming beam leakage into outgoing beam signal Using other BPM measurements of incoming to estimate incoming beam position in stripline BPM Assume a factor of 5 in isolation

10 Temporal Separation With high-bandwidth signal processing used in the SLAC linac and SLC, the incoming beam signal can be arranged to have essentially decayed away by the sample time for the outgoing beam. Assume a factor of ten. Example: SLC IP BPM (Simulation)

11 Phase relationship Signal shaping can be chosen so that the tail of the incoming beam signal is near its zero crossing when the outgoing beam signal is at its peak. Get maybe a factor of 3 contribution to isolation. Procedure: measurement is done sampling signal at its peak. Choose analog filter so that 27 ns after peak signal is (approximately) crossing zero.

12 Prediction/Compensation Contamination of the measurement of the outgoing beam on each strip, due to the incoming beam leaking through filters 1-3 above can be predicted from the beam position of the incoming beam in this BPM. The incoming beam in this BPM is measured cleanly. Other upstream BPMs may contribute to the estimate as well, Especially the next two nearest BPMs which are probably high resolution cavity BPMs The prediction coefficients are learned by running single beams through the system a few pulses once a week if everything is stable over a week a few pulses/hour if not. Of course if one can entirely predict the first BPM from the readings in the 2nd & 3rd, you didn't need the first ones anyway. Depending on what unique information on beam parameters is provided by the close-in BPMs, this technique could be powerful in removing contamination from the incoming beam, especially because as it takes advantage of the power of the first 3 techniques to gain leverage, one doesn t need to know the incoming beam position at BPM1 very precisely if the other 3 filters keep the contamination small. We know of no examples of doing this in practice.

13 Conclusions Required resolution: σ y < 10 microns (single bunch) Budget half of this for other noise contributions Leaves incoming/outgoing contamination budget σ y < 5 µm Combine directionality, temporal separation, and phase isolation Get rejection factor of r = Then we meet requirement, before applying prediction/compensation technique, if incoming beam jitters by less than 5µm/900 ~ 5 mm (beam is unlikely to jitter this much!) 1 30 We have been conservative, can do better in the rejection factors Can still apply the prediction/compensation tool. Conclude we have tools to control incoming beam contamination of the beam beam deflection measurement Must design and fabricate directional stripline BPM and its interface to the outside world carefully. And choose processing scheme carefully = 1 900

Next Linear Collider Beam Position Monitors

Next Linear Collider Beam Position Monitors NLC - The Project Beam Position Monitors Steve Smith SLAC October 23, 2002 What s novel, extreme, or challenging? Push resolution frontier Novel cavity BPM design for high resolution, stability Push well