Design and test of a high-speed beam monitor for hardon therapy

1
Design and test of a high-speedbeam monitor for
hardon therapy

• H. Pernegger on behalf of Erich Griesmayer
• Fachhochschule Wr. Neustadt/Fotec Austria
• (H. Frais-Koelbl, E. Griesmayer, H. Kagan, H.
  Pernegger)

2
MedAustron

• Austrian medical accelerator facility
• Cancer treatment and non-clinical research with
  protons and C-ions

Protons
C-Ions
3
Layout
Preliminary layout
4
Parameters

• Proton Carbon Beam
• Energy 60-240 MeV protons and 120-400 MeV/u
  C-ions
• Intensity 1x1010 protons (1,6 nA) and 4x108
  C-ions (0,4 nA)
• Beam size 4x4 mm2 to 10x10 mm2
• Setup
• 4 fixed beams and 2 gantries
• Field sizes 40x40 cm2, 25x25 cm2, 4x4 cm2 (fixed
  beams), 20x20 cm2 (gantries)
• Active scanning
• Extraction period 1 s to 10 s

5
High speed beam monitor

• Initial goal Develop a detector for beam
  diagnostic
• Measure intensity structure of extracted beam
  by counting individual particles (no integration)
• Short pulses with good time resolution for
  high-speed counting
• Resolve beam time structure (measure number of
  extracted particles for each revolution)
• 1D or 2D position sensitivity to provide beam
  profile
• Rates counting single particles at rates close
  to the GHz/channel-range

Maximum rates up to 6x larger during RF cycle
6
Beam Monitor Concept

• Segmented CVD diamond as detector material
• High drift velocity short charge lifetime give
  short signals
• Radiation hard
• Variable segmentation possible on thin solid
  stage detector
• RF-amplifier and parallel counting
• Direct amplification of ionization current pulse
  (no current integration)
• discriminator and pulse counter to parallel
  readout

7
Test of first prototype

• Tested a first prototype of detector and
  electronics at Indiana University Cyclotron
  Facility
• Tested with protons (worse case smaller signal)
• Tested in energy range for proton therapy
  (55-200MeV)
• Variable intensity
• Main focus measure analog signal characteristics
• Signal time properties
• Amplitude properties
• Energy scan and dE/dx in diamond
• Efficiency
• Tested with first prototype of
• 2 samples of CVD diamond
• First prototype for analog amplification stages
• First tests of digital electronics (in progress)

8
Setup and Samples used for tests

• 2 diamond samples with different pad size
  scintilator as telescopes
• 2.5 x 2.5 mm2 (in trigger) CCD 190 mm, D 500
  mm
• 7.5 x 7.5 mm2 (for analog measurements) CCD 190
  mm, D 500 mm
• Trigger scintilator (5x3mm2)

9
RF amplifier stage

• 3-stage current amplification
• Parameters (per stage)
• Bandwidth 2GHz
• Amplification 20dB, Noise 2.7dB
• Some signal estimates
• Max. current peak from diamond 1.7mA for MIP
• Max (theor.) SNR expected for 55 to 200 MeV
  protons 201 to 81

10
Digital readout stage

• Disriminator
• Discriminate on voltage and time difference
• Baseline restoration with delay line
• Implemented in PECL

• Counting readout
• Count in fast 8-bit and latch to 24 bit counters
• Allows to store full pulse trains for fast rate
  vs time measurement in SDRAM

11
Measured pulses

• Single signals in diamond (protons at 55 MeV)

12
Signal Time Properties

• Rise time 340ps Duration 1.4ns

• Average pulse shape

• Pulse duration (FWHM)

13
Diamond signal amplitudes

• Amplitudes in the full energy range
• r.m.s. noise 18 mV

14
dE/dx in CVD Diamond

• Compare measured signal to calculated dE/dx
  behaviour in diamond
• Normalized at 104 MeV for uncertainty in absolute
  calibration

15
First results on Signal-to-Noise
200 MeV 104 MeV 55 MeV

• Measured most probable S/N ranges from 151 to 71

16
Preliminary results on Efficiency

• Defined as signal with
• amplitude gt 3 x snoise
• tsignal in /- 3ns window of trigger time
• Measured efficiency of 99 to 94 (noise limited)

100
90
17
Next steps

• Diamond and dedicated electronics seems to be
  ideally suited for beam diagnostics
• Achieved very promising results for beam
  diagnostics with protons
• SNR 71 to 151 in the typical energy range for
  proton therapy
• Risetime of 350ps and pulse width 1.4ns
• Efficiency 94 to 99 (electronics noise limited)
• Since then
• Worked on optimizing SNR for even lower signals
  (MIP range) and achieved lower noise with
  modified electronics
• Recently tested with C ions (3 weeks ago) and
  large surface (3x1cm pad)