1

ESS MEBT Layout & Chopper Rise Time- From the Point of View of Beam Physics -

Ryoichi Miyamoto (ESS/AD/BPG)On behalf of the ESS/ESS-Bilbao MEBT team

Jan 26th, 2015MEBT Chopper Workshop

Outline

● Layout(s)● Chopper rise/fall time● Conclusions/Questions

2

2 layouts: chopper & quad integrated vs separated

● ESS MEBT lattice structure:

(Triplet) + (Chopper+Dump) + (Triplet) + (Diagnostics Box) + (Quadruplet)● The chopper quad was originally considered to surround the chopper

(like LINAC4) but this has been questioned.● Two configurations are tried and the difference is minor from the point of

view of beam physics, as seen in the following slides.

Integrated (2014.v2) Separated (2015.v0)

3

Summary of two layouts

● Integrated– Chopper specs:

● Voltage: ≥ 4.5 kV● Plate length: 500 mm● Plate gap: 20 mm● Rise/fall time: ~10 ns

– Tuning of the chopper section slightly harder.

– Beam quality● Transverse: roughly the same● Longitudinal: slightly worse

– Slightly higher voltage and shorter plate better and the above situation may change. (Good Tunability and beam quality if 7-8 kV is achieved.)

● Separated– Chopper specs:

● Voltage: ≥ 4.0 kV● Plate length: 500 mm● Plate gap: 20 mm● Rise/fall time: ~10 ns

– Tuning of the chopper section slightly easier.

– Beam quality:● Transverse: roughly the same● Longitudinal: slightly better

– Slightly higher voltage and shorter plate better and the above situation may change.

4

Comparison of two layouts: distributions (MEBT out)

● The transverse better for the “Integral” and the longitudinal better for the “Separated”.

Integrated Separated

5

Comparison of two layouts: distributions (DTL out)

● Through the DTL, the transverse become roughly the same. The longitudinal is still better for the “Separated”.

Integrated Separated

6

Comparison of two layouts: densitiesIntegrated Separated

7

Comparison of two layouts: emittance and haloIntegrated Separated

(Matching not fine-tuned for these layouts yet.)

8

Rise/fall time and partially-chopped bunches

● Partially-chopped bunches are of concern for beam losses and a major factor to determine the rise/fall time.

● No clear threshold in terms of the beam losses and it seems rather a political decision of how much we are allowed to use the 1 W/m.

● SNS and JPARC are having no issue.

4.2 kV

1.5 kV

9

Losses for various partial voltages

● Losses mainly in DTL and almost none in the SC sections (even with errors).● The bad voltage range is ~1-2.5 kV.● Rise/fall time definition with 10-90% seems good enough. (Certainly 5-95%.) ● Lattice errors enhances the (confidence level) losses by roughly an order or magnitude.● Practically no loss with the 2nd and 3rd scrapers. (Should think the pessimistic case?)

10

Total losses during 10, 20, and 30 ns rises

● Linear 10, 20, and 30 ns rises are considered.● 10 ns → ~1.1 kV per RF cycle → 1 bunch in the bad range of 1-2.5 kV. ● Some differences between 10 and 20 ns but 20 and 30 ns are quite similar.● Should be verified by including the errors.● Hard to draw the threshold. (Do we care the peak or total ??)

11

Conclusions/Questions● Chopper parameters wanted by a beam physicist

– Voltage: ≥ 4 kV– Plate length: ≤ 500 mm– Plate gap: 20 mm– Rise/fall time: ~10 ns (20-30 ns OK ??) for 5-95% (10-90% seems OK)

● Two layouts are compared and for an around 4 kV voltage range the separated chopper and quad configuration seems better (although the difference isn't major).

● Which is the recommended technology choice for our MEBT chopper (and quad) to satisfy above parameters in terms of– Feasibility– Reliability– Design effort– Cost– …

● Is there anything else to do on the beam physics side?– Compromise between the voltage and rise/fall time?

12

Losses due to 1.5 kV partially-chopped bunch (old)