111th ATF2 weekly meeting

International phone/webex meeting

The LCLS BPM electronics and modified version for ATF2 EXT BPMs, Steve Smith (SLAC)

S.Smith briefly reported the LCLS BPM system with comparison to ATF2 system consisting of 12 stripline BPMs and suggestion of sharing responsibilities between SLAC and KEK for new electronics.

There are 200 BPMs with 100mm long stripline and 24mm aperture. The position resolution is 5um at beam intensity of 200pC, 1.25 x 10⁹ e/bunch. The LCLS electronics is based on digital processing with narrow band pass filter centered at 140MHz in the analog front end. The signals are digitized by 120MHz sampling ADC. It has online calibration system by launching tone burst into one strip to induce calibration signals on the two orthogonal strips, i.e. input to top (left) for left(top) and right(bottom) strips.

He suggested that SLAC builds processor chassis containing Analog Frontend without ADC board and KEK provides ADC / digital processing .

Q : Are you care of kicker noise ?

A : We will have no problem from external noise, because the kicker has no 140MHz signals.

Q : What is algorithm for position measurement ?

A : The position is calculated by amplitudes of signals.

Q : How much is the dynamic range?

A : The design one is 5um at 200pC. It can measure beam positions of 4 - 5 x 10¹⁰/bunch with attenuators if necessary. It can measure even from 20pC (1/10 of design) with resolution of 30um.

Q : What is the specification of digitizer ?

A : We have 16bit ADC with 120MHz sampling .

Q : Can you test the LCLS electronics at ATF2 in November ?

A : Yes.

Q : What is the production time scale of all BPM electronics?

A : It takes a couple of month. We may provide it in Feb. or Jan. 2010 . There has been some problem in calibration, i.e. bad parts which have been all replaced. LCLS people have been very busy for the repair works.

Q : What is the multi-bunch performance?

A : In case of 357MHz separation, it can measure average of whole bunches.

Q : How is for 150nsec separation ( ILC like)?

A : It can separate in each bunch, or we can use higher band pass filter for better separation.

Q : What is the calibration scheme?

A : We pulse input from top to induce left and right electrodes, and from left for top and bottom gains.

We actually resumed the discussion on how to proceed the new electronics after S.Bai's presentation.

Q : First to Yves, please explain status of checking the stripline BPM electronics ?

A : I checked all the head-amp and clipping modules . Three BPMs had mismatched input polarities in the clipping modules. One module labelled "filter skip is not present" will be modified to recover the low gain. In next week, I like to test the electronics with kicker noises and to see signals with new high pass filters of 25MHz, where typical frequency of signals is >100MHz. ( Further details will be reported at the weekly meeting, 9th September. )

Q : How do we proceed the new electronics of stripline BPMs ?

A : We (SLAC) will test the electronics in November, then like to provide the electronics for the 12 BPMs.

Q : Do you have a cost estimation of digitizers ? Also, we need the proposal of new electronics.

A : We can estimate the cost per channel from LCLS system, where there is 48 channels in total ( 12BPMs x 4 ). We will write the proposal.

C : SLAC would like to order components even before test results in November since it takes time to produce custom filters. So, SLAC would like to proceed the production.

Q : KEK can not fund the digital processing system as you suggested in this fiscal year. ( We understood that SLAC will provide whole system for the 12 stripline BPMs at previous meetings.) Is this supported by Japan-US cooperation ?

A : No, it is funded by SLAC/BDS money. We(SLAC) may fund the digital system too.

Attempts to model & explain the discrepancy between measured Twiss at the Post-IP WS and propagated ones in the re-matched optics, Sha BAI (IHEP)

S.Bai reported comparison of Twiss parameters between measurements at the post-IP wire scanner (PIP-WS) and propagated ones from measurements at the upstream in the re-matched optics on two days, i.e. 20th and 28th May, 2009. She also compared dispersions and beam sizes at PIP-WS between measurements and MAD simulation on the both days. Both dispersion and beam sizes were measured as a function of QD0 current.

Measured and propagated Twiss parameters have discrepancy in horizontal and vertical directions, where &alpha_x and &alpha_y have larger discrepancy on both days. The magnet currents were set close to the nominal ones except for three magnets of QF1X, QF6X and QF11X with largely different currents, i.e. -3.9%, -3.5% and -11%, respectively on 28th May.

The MAD simulation could not reproduce the measured dispersions in both directions on both days. The measured beam sizes could be fit by MAD with the magnet settings on both days, while the propagated Twiss parameters did not reproduce them at the PIP-WS .

Q : Why were the current settings of three magnets different on 28th May?

A : The currents were changed for correcting horizontal dispersions.

Q : Do they affect TWISS parameters at IP ?

A : I do not know.

Q : Why is it so poor to propagate to IP ?

A : I do not know. The current settings may not be correct ?

Q : Are the magnet currents wrongly set ?

A : I do not think so.

C : Of-course the most relevant check is not the propagation but a direct measurement (of Twiss parameters) at IP.

KEK site meeting

C: Knife edge target should be a razor or something glinded. Just hollowed metal plate cannot be used as a knife edge.

Q: When is the IP target prepared? The test of the mover (movement with vacuum, position reproducibility and so on) will be needed early.

A: Fabrication will be finished in the beginning of September.