Final Design Review for The SNS BLM System R' Witkover, D' Gassner

1
Final Design Review forThe SNS BLM System R.
Witkover, D. Gassner

• Part III Ion Chambers

2
Choice of Detector

• Ion Chamber
• Good uniformity, linearity, stability
• Resistant to radiation
• Scintillator/Photomultiplier
• Wide dynamic range, fast response
• Requires frequent calibration
• Susceptible to radiation Damage
• Pin Diode
• Good uniformity, linearity, stability
• Not well suited to pulsed beam
• 107 Max pulse rate ? 1000 cnts during macropulse
• Ion Chamber is best choice

3
FNAL Ion Chamber

• Designed by Shafer for Tevatron 1982
• Sealed glass, Argon filled
• Packaging improved for RHIC to reduce noise

4
FNAL Detector Production Problem

• Poor uniformity with preferred polarity
  (electrons to center electrode) so RHIC used
  opposite
• BUT Slower ion transit time and earlier
  saturation

5
Problem Solved

• Notching of outer electrode eliminated crimp at
  support lower gradient

6
Original BLM Detector Design For SNS

• Advantages
• Low Cost ( 450 assembled)
• Long history of Design/Manufacture (1982 -)
• Proven record in Accelerators (Tevatron, RHIC)
• Disadvantages
• Recent problems with sole source vendor (2001)
• Poor unit-to-unit response for preferred bias
  polarity in later units
• Slow Ion Collection time (600 µsec at 3000 V)
• Poor Collection Efficiency (saturation) at high
  loss rate (35-40 for 1 local loss at 3000V)

7
Improving Ion Transit Time

• Ion speed vion µ0 E(P0/P) µ0
  (V/d )(P0/P)
•  Where µ0 Ion Mobility, V Voltage, P/P0
  Pressure (Atm)
•  
• Transit Time t d/ vion d2/µ0 V
  (P0/P)
• For plane geometry d Electrode gap
• For cylindrical geometry d2 (a2 b2)/2 ln
  (a/b)
• Where b Inner radius and a Outer radius
• The new design uses the same outer diameter of
  1.5 inches but an inner diameter of 1 vs 1/4-inch
  and active length of 6.7 vs 4 inches to get the
  same sensitivity.

8
Efficiency and Ion Transit Time vs Inner Radius
9
Calculation of Collection Efficiency

• Difficult calculationsolutions available for
  certain regimes
• Static
• Duration long compared to ion transit time
• Electron-Ion pairs continually created drift in
  opposite directions
• Pulsed
• Duration short compared to ion transit times
• Electrons rapidly collected leaving only ion
  cloud
• Solutions shown due to Boag and others

10
Collection Efficiency for Continuous Loss
Specific Charge density Equivalent Gap
I sig Chamber signal current  a Outer
radius b Inner radius  a Townsend
recombination coefficent k1 Electron
mobility k2 Ion mobility e Electronic charge
Efficiency
11
Collection Efficiency for Continuous Loss
12
Collection Efficiency for a Pulsed Beam
Specific Charge density
Collection Efficiency
13
Collection Efficiency for a Pulsed Beam
14
New BLM Detector Design

• Advantages
• Better HV design
• Should improve unit reproducibility
• Should allow higher bias Voltages
• Better Collection Efficiency for high loss rate
• Much faster Ion Collection time
• Low-skill design- can be made by multiple
  vendors
• Disadvantages
• Development time and funding not in original
  scope
• Unforeseen production problems may arise
• Increased cost
• Long-term stability unknown

15
SNS Prototype Ion Chamber
16
Simulated Linac/HEBT Response
17
New Design Prototypes

• 3 prototypes built
• 1 Unable to hold 3 kV with Argon filled with N2
  
• Installed in Linac for tests
• Tests done at linac at 200 MeV, 2-4 kW local
  loss
• 2 Redesigned ceramic insulators, Signal
  electrode. Higher Voltage feedthru
• Held 3.7 kV with Argon
• Installed in Booster to AGS (BTA) Line Tests done
  at 1.2 GeV 5-10 W local loss
• 3 Shown to vendors as model

18
Beam Test of FNAL SNS Prototype BLM
19
Detector Saturation Tests
20
Pulsed Beam Tests in BNL BTA Line
21
X-ray Sensitivity of RHIC BLM
X-rays produced by RHIC 200 MHz RF cavities
Vertical Scale is in kV And Pico-Amps
22
Future X-ray Tests

• Future tests required
• 1/8, ¼, 3/8, ½ thickness lead shields
• Estimates from LANL and SNS suggest 10s of R/hr
  X-rays at 30 cm
• BNL test Signal 1000 pA 51 R/hr during
  pulse.

23
Status of Detector

• New detector design shown to 3 vendors
• LND (Long Island, NY)
• ADC (Ithaca, NY)
• Far West (San Diego, CA)
• LND built prototype undergoing testing
• ADC .preparing estimate
• Far West. preparing estimate
• Preliminary estimate at BNL Material and labor
  costs 800, vs 450 for FNAL design. Industrial
  production cost will probably be higher.