PEP-II Injection

1
PEP-II Injection Status, Improvements,Trickle,
Tuning
Franz-Josef Decker 9-Oct-2003
1. Status Fills from scratch times
overall 12 - 24 min Top-off 3-4
min clear aborts 3 - 7 min
Delivery 50 min injection 8 - 15 min
tuning 1 - 2 min
2. Improvements Injection Tuning Goal
factor of 2-3 better gt 20 for BaBar
now (0 later!) Tools faster
injection / less tuning less aborts less
inhibits (BaBar) / less feedbacks not keeping
track 3. Injection background 50 of
background due to injection
4. Trickle injection Continuous injection
into LER
2
Overall
Tuning
From G. Wormser
Start HER
Inject HER
Start LER
Inject LER
3
Injected positron charge 1.2 0.2 E10
4
Injection efficiency 80 (down to 40)
5
Injection rate 1.2 E10 80 at 15 Hz
3.1 mA/s fill rate gt Injection time 8 min
fill time for 1500 mA (Rest of 8-15 min must be
tuning or less injection!) or 24 min for
4500 mA, too bad ?! Improvements Increase
current 2.0 (-2.4) E10 gt95 2 (-4) bunches
35.6 ns apart (for by-2 pattern) change
pulser into kicker 30 Hz ? 60 Hz But
BaBar background inhibits some detector
parts (EMC) look still at background not
background /charge Changes in PEP-II have to
keep up with faster rate IP and IR quadrupole
moves, feed-forward would help (RF) More
improvements better diagnostics, injection
stability, less background
6
Injection Top-off and Fill from Scratch
Lost beams 75625 BaBar start 82511 (trapped
event)
Fill from Scratch 12 min 2 10 min
Top-off 3-5 min
Injecting 10 min
At least 2 (-5) min down
7
List of Injection Problems and some Solutions

1. Smoother fills (less saw-tooth) with trickle BIC
   should allow full low charge quanta (now 50).
   R.C.
2. Diagnostic injection rate per bunch?
3. Quanta difference for e by trigger delay for
   flux concentrator instead of pos-phase. M.St.
4. Energy feedback independent of quanta. L.H.
   
5. BG diagnostic from BaBar for each injection
   pulse, to find correlation, (GADC).
6. HER BG high and 5 ms structure, needs more ideas.
7. Injection or PEP-II problem (e.g. tune) maybe
   pseudo lifetime helps.
8. EMC rate-limit for HER needs revision BG per
   charge!
9. Injection bump not closed, needs setup time,
   diagnostics (BPMs) seems o.k.
10. Faster injection 1.5E10 ? 2.4E10 too fast for
    RF, FB (e.g. LEROgtneeds FFW), EMC-BG, BSOICS?
11. Match/mismatch at injection / clearance at
    septum.
12. How to set-up and maintain good injection x, x,
    y, y, E, ? PLIC, ?x, ?x, ?y, ?y, ?E, (?z ).
13. Fast feedbacks are they bad for the injected
    bunch and should be switched off for 10 ms.
14. Reliability/availability number ?
15. Faster injection by hardware multi-bunch, 60 Hz,
    pulsed magnet changed to kicker.
16. HER injection BG o.k. by itself gt fill out off
    collisions FFW.

8
List of more Injection Problems and some Solutions

1. Trickle protection (DALEK). R.I.
2. Quantify injection background ? BG dt /
   ?Istored history. R.C.
3. Beam loss during injection.
4. Injection timing jumps.
5. Change fixed quanta to requested quanta
   gt30-50 faster 10.50.250.125 10.875.
6. HER injection BG o.k. by itself gt fill out off
   collisions FFW.
7. IP move during injection LERO/HERO FFW
   (feed-forward).
8. BaBar only abort one beam gtwe should be able to
   fill against full ring.
9. Gated camera diagnostics.
10. e injection give e- BG?
11. Faster recover from aborts.
12. Diagnostics for injection E_off, phase_off.
13. Standardize?
14. Injection history x, x, y, y from BPMs
    diagnostic feedback needs coefficients.
15. BPM offsets into DB at injection point in y.
16. Injection time into history ( delivered
    time).

9
Injection Background
From M. Weaver
Injection backgrounds increase as a fill
progresses. The period from 0 to 240 seconds
consists of large charge quanta injection into
the HER and LER at 15 Hz each. The period from
240 to 320 seconds uses small charge quanta
injection into the HER. The period from 320 to
410 seconds includes 30 Hz injection into the LER.
10
High HER Injection Background
The background rise 5 ms after injection is not
fully understood. Possibly filamentation of the
stacked, injected beam with the stored beam.
11
Trickle Injection
Trickle injection Maximum possible improvement
22 (6.6 E33 (peak) / 5.4 E33 (average over
fill)) More realistic 5-15 (LER only) Trickle
test
1 hr
3 hrs
Average Luminosity with 3 min top-off time
12
More detailed trickle injection issues
Smoother fill (less saw-tooth) BIC should
allow full low charge quanta (now 50) (More
diagnostic injection rate per quanta we have,
but which make trouble injection rate per
bunch?) Discovered shortcoming of injection set
up Difference in quanta size should not be
possible to tune away ? reduce e by time
delay of capture cavity (trigger delay for flux
concentrator) instead of phase change of
pos-phase. Energy feedback should be
independent of charge ? four feedbacks for four
quanta (or even feed-forward?) Better BaBar
diagnostic of which pulse gave high background
(gated ADC), needed to find correlation. Otherwis
e real promising!