Design and characterisation of a Highly Miniaturised Radiation Monitor

(HMRM)

N. Guerrini1, R. Turchetta1, D. Griffin1, T. Morse1, A. Marshall1, O. Poyntz-Wright1, S. Woodward1, E. Daly2, A. Menicucci2, H. Araujo3, E. Mitchell3

1.STFC - Rutherford Appleton Laboratory, UK

2.ESA - Noordwijk, The Netherlands3.Imperial College London, UK

International Workshop on Semiconductors Pixel Detectors for Particles and Imaging

PIXEL 2012 – Inawashiro, Japan6th September 2012

2

• Introduction

• Application and scope

• Radiation monitor architecture

• ASIC design and characteristics

• First results

• What’s next

• Conclusions

OUTLINE

6th September 2012

3

INTRODUCTION

6th September 2012

Data on the ionising radiation environment are very important to ensure an efficient design and operation of spacecrafts. Monitors currently in use weight more than 2Kg, require several Watts of power, are quite large (>10cm in each dimension)

STFC and IC (within the framework of a ESA technology development contract) aim to develop a “chip sized” (that is smaller, lighter and less power hungry) prototype radiation monitor suitable for application in:• Coarse radiation housekeeping• Save and alerting function• Support to platform and payload systems

4

INTRODUCTION

6th September 2012

Radiation Monitor requirements:

• Minimum number of external components

• Dosimeter function

• Particle Rate Meter function

• Particle Species Identification function

• In-flight calibration procedure

• Power dissipation < 200mW

• Weight <20 g

• Radiation TID tolerance of at least 100kRad

• Latch up free

• Operating temperature range from -40 to +80°C

Monitor based on Active Pixel SensorsLow cost (CMOS process)Low complexity (functionalities on integrated circuit )

56th September 2012

---

-

+ ++

+ +

+

--

+-

Low energyparticle

High energyParticle (MIP)

High energyParticle (MIP)

Low energyparticle

• When the energy deposited in the detector by a low energy

particle is the similar to the energy deposited by a MIP

Ambiguity

• With APS sensors and judicious selection of silicon substrate

thickness we can think to sample the dE/dx curve and uniquely

identify the particle species and energy.

APPLICATION

66th September 2012

1. Conceptual design: APS telescope with low threshold for e/p discrimination– Parameters: number of sensors and shields, aperture size– Variables: energy thresholds, sensitive area, particle ranges

2. Optimisation with preliminary simulation geometry of particle telescope– Parameters: dimensions/materials for wafers, shields, casing, window, algorithms– Variables: thresholds, effective area, efficiency and purity for PID

3. Validation with full simulation geometry in 5 reference orbits– Parameters: algorithm tables (on-board and ground segment)– Variables: ID purity, pile-up probabilities, dosimetry functions, spectral reconstruction

G4HMRM v1.0CADd[1] d[2] d[3] d[4]S1 S2 S3 S4

1 32

APPLICATION

protons

7

APPLICATION

6th September 2012

SIMULATED PERFORMANCE SUMMARY

Electron range: 0.063 to 6 MeV

Proton range: 1.3 to 500 MeV

Maximum omni/d flux: 1x108 cm-2s-1

Dosimetry

Count rate on front sensor

Dose rate on all sensors

Also cumulative lifetime doses

On-board alerting functions

3 programmable particle ‘channels’

Offline spectral reconstruction

Electron/proton energy spectra

electrons

8

ARCHITECTURE

6th September 2012

Three main mechanical elements of the HMRM chassisLID: Provides the prescribed geometry for detector telescope. Mounts the window (for visible blindness)BODY: Main structural housing around electronicsBASE: structural interface to spacecraft

Fabricated in Titanium - Mass 52g (including fasteners and connector)

96th September 2012

Flex rigid PCB designFour rigid boards

Interface PCBInterface connector PCBFPGA PCBDetector PCB

ARCHITECTURE

6th September 2012 10

ASIC DESIGN

• 0.18 µm CMOS Image Sensor technology

• 20 µm 4T-pixels in a 50 x 51 array

• Snapshot and correlated double sampling

(CDS)

• Frame rate up to 10,000 fps

• Column-parallel 3-bit single-ramp ADC, with

in-column trimming

• Digital readout, plus analogue readout for

debugging

• Integrated DAC for voltage/current

generation

• Band gap

• Temperature sensor 

CMOS sensor for a Highly Miniaturised Radiation Monitor (HMRM).

Main specifications.

6th September 2012 11

ASIC DESIGN

• Simplest active pixel • Higher noise• Well proven, high radiation

tolerance

• Most popular active pixel• Low noise• Good radiation tolerance

The radiation tolerance requirements for HMRM are well within the 4T pixel measured range.

3T vs 4T

6th September 2012 12

ASIC DESIGN

Landau distribution(from beam test results)

S/N =118

Noise (in e-, before board noise correction)

Noise histogram(measured with

Photon Transfer Curve)

Most probable noise: 3.6e-

Average noise: 4.5e-

Conversion gain at output: 65.0μV/e-

Linear full well capacity: 17,900e-

CMOS 0.18 µm – 4T sensor

12 µm epi thickness with 1kW cm resistivity

15 µm pixel

FORTIS

6th September 2012 13

ASIC DESIGNRadiation Hardness

6th September 2012 14

ASIC DESIGN

• 60 sensors manufactured on 12 µm, low resistivity epitaxial substrates

• 60 sensors manufactured on 12 µm, high resistivity (>1kOhm cm) epitaxial substrates

• Design for radiation tolerance: Total dose Single event upset immunity

6th September 2012 15

ASIC DESIGN

The sample values from the pixels and the reset values are readout in snapshot mode and stored onto sampling capacitances at the top and at the bottom of the array.

The sampling control logic allows to sample and reset the pixels independently, hence controlling the pixel shuttering.

Row1 Readout

Pixel Sampling

Frame Time = 102us

2us

100us

Row2 Readout

Row24 Readout

Row25 Readout

RowN Readout

4us

RST

TX

RST

TX

6th September 2012 16

ASIC DESIGN

• One comparator per column (readout on both sides)

• Seven 7-bit programmable thresholds

• One 8-bit DAC for each comparator trimming adjusting

The ADC ramp is generated from the 7 programmable thresholds.

The comparator output is sampled 7 times (for each ramp level). The result is then converted into a 3 bits value.

In this way the output data will tell us between which thresholds the pixel value is.

6th September 2012 17

ASIC DESIGN

• 3-bit encoder per comparator

• Data read on a 9 bit bus 2*17=34 clock cycles for readout

• plus one clock cycles for temperature sensor data (9-bit)

• Test input to the entire shift register

• Triple majority voting (TMV) system for single event upset immunity.

• TMV used for programming and readout shift registers, except in shift register controlling the analogue readout.

• Enclosed geometry layout for total dose tolerance (entire chip).

18

RESULTS

6th September 2012

Photon Transfer Curve method, measured through analogue test output:• Gain = 59.9 µV /e- (exp.

51.2)• Noise = 16.8 e-rms (exp.

13.5)• Cin = 2.1 fF (exp 2.3)• Full well = 14,600 e-• Full well (lin) = 6600 e-

Expected performance on data path:• Gain = 40.3 µV /e-• Noise = 13.9 e- rms• S/N (m.i.p.) = 69

19

FUTURE WORK

6th September 2012

• Complete digital characterisation with light

• Test of 2nd iteration (end of September)

• Tests under irradiation (2nd iteration)

• Temperature tests (2nd iteration)

20

CONCLUSIONS

6th September 2012

• CMOS image sensors provide new solutions to radiation detection

• High radiation tolerance

• Highly Miniaturised Radiation Monitor (HMRM) based around a

CMOS image sensor now in development

• Measured noise = 16.8 e- rms

S/N (m.i.p.) = 69

• HMRM instrument about to be tested

with sources

• 2nd iteration on (end of September) its way with reduced

power consumption and improved cross-talk and reliability.

216th September 2012

Thanks!!!http://dsc.stfc.ac.uk/Default.aspx