Wieman: 1LBNL

Micro-vertex

STAR Collaboration Meeting Aug 2003

LBNLHoward Wieman, Fred Bieser, Robin Gareus (Heidelberg), Howard Matis,

Marcus Oldenburg, Gulshan Rai, Fabrice Retiere, Kai Schweda, Hans-Georg Ritter, Eugene Yamamoto

UCIYandong Chen, Stuart Kleinfelder

BNL Instrumentation DivConsulting

OSUIvan Kotov

PurdueDennis Reichhold

Wieman: 2LBNL

Today's topics

• Monolithic APS CMOS, new development of photo-gate – Purpose

– Modeling

– First silicon results

• Mechanical work– Vibration measurements

– Light weight silicon support, another approach

– Thin beam pipes

Wieman: 3LBNL

• CDS removal of fixed pattern noise and KTC reset noise on the chip (standard diode requires CDS off chip)

• Increase signal by reducing signal spreading to adjacent pixels. The photo gate permits large geometry without adding capacitance to the sense node.

Photo gate purpose - addresses standard diode limitations

P-

P

P+

Standard diode geometry

Standard APS diode structure

Wieman: 4LBNL

photo gate

source follower gate

reset gate

transfer gate

row select gate

sense node drain

• Large photo-gate to collect large fraction of the charge on a single pixel, directly on the p- epi layer

• Small transfer gate also directly on p- epi layer

• Small drain (minimum capacitance) connected to source follower gate (sense node)

Photo-gate geometry

20 m

x -2 m- 1 m 1 m5 nm

8 m

0.1 m

0.4 m

P epi 1.4x1015 1/cm3

N+ 1x1020 1/cm3

photo gate transfer gate

drain

x = 0.4 and 0.8 m

(simulation quantities)

Wieman: 5LBNL

Photo-gate issues in standard CMOS

• No double poly process – possible poor transfer between gates because of low transverse field

• Floating n well between gates, a bad solution to the transfer problem with single poly

• Sub-micron process may solve problem

single poly, limitation of CMOS

double poly, standard for CCDs

photo gatetransfer gate drain

floating n well

Wieman: 6LBNL

Photo gate/transfer gate operation

photogate

transgate1.8 V V

drain2.4 V

photogate

transgate

drain

V phg = 0.8 Vtransfer mode

V phg = 2.4 Vcollection mode

Wieman: 7LBNL

Photo gate/transfer gate with 800 nm separation

photogate

transgate1.8 V V

drain2.4 V

photogate

transgate

drain

V phg = 0.8 Vtransfer mode

V phg = 2.4 Vcollection mode

Wieman: 8LBNL

Drain current after light injection

• 400 nm between photo gate and transfer gate, no floating n well

• Nano amp drain current

• Rapid electron transfer - complete in 60 ns

photo gate

transfer gate drain

Light injection

60 ns

Wieman: 9LBNL

First silicon tests, comparing photo-gate with standard diode structure

Photo-gate directly to sense node drain

DC bias:V photo-gate 0.6 VV drain 2.4 V

Output signal for Fe55 X-ray test

Issue:

ADC

Linear

ADC

Log

diode

Photo-gate 1

Photo-gate 2

diode - fullcharge collection

Why is the signal spread out – is it surface traps under the gate?

Wieman: 10LBNL

Mechanical Concept• Single end

support for rapid installation and removal

– For beampipe bake out

– Insurance for beam excursion damage

• Readout electronics in end support module

Low Mass Carbon fiber tube

Thin silicon ladders under tension

Aluminum Kapton cables under tension

Wieman: 11LBNL

Tension concept

Wieman: 12LBNL

Tension concept

Wieman: 13LBNL

50 m Silicon

9

Thinned and polished wafers, a standard industrial process

Thinned Silicon wire bonded to cable, both supported under tension

Used in wind tunnel test

Wieman: 14LBNL

Air cooling and vibration

• 1-2 m/s air cools 100 mW/cm2

• TV holography shows 2 m vibration for tensioned silicon structure

TV holography vibration map

Lucite wind tunnel Thin silicon ladder,tension support

Laser light source

Camera

Wieman: 15LBNL

Thin stiff ladder concept

• Under construction

• Will be tested for vibration in our wind tunnel

carbon composite(75 m)

aluminum kapton cable(100 m)

silicon chips(50 m)

21.6 mm

254 mm

Wieman: 16LBNL

Self supporting ladder concept

Outer shell carries extended beam pipe gravitational loads

Free end access for detector insertion

Wieman: 17LBNL

A thinner beam pipeElastic Bucking InstabilityEuler 1757

Eigen solution

Critical buckling force

The challenge: increase moment of inertia along y without increasing material

16 cm

4 cm

Aluminum 150 m thick

critical pressure 5.5 atmospheres

Increase I with corrugations

0.03 atmos. with no corrugation

Wieman: 18LBNL

Conclusion

• Present monolithic CMOS APS detector technology suitable for slow 10-20 ms readout (examples at LBNL/UCI and LEPSI/IRES)

• Fast readout not ready yet, but progress is being made

• Radiation hardness good enough for RHIC

• Mechanical concepts progressing

• R&D is fun, but…

• Proposal due this December

Wieman: 19LBNL

Test of diode variation

• No Fe55 signal

• Will test with more statistics

p- epi

np p

Puzzle:

Wieman: 20LBNL

Rejection of primary tracks

• For large rejection ratios must set cut at several times multiple scattering angle

• At these large angles single coulomb scattering dominates and materials contribute linearly

Beam pipe x/X0 = .17 %Si x/X0 = .05 %C Comp x/X0 = .03 %Al Kapton x/X0 = .03 %

Leakage fraction

0

0.6

1.2

total Be beam pipe 600 micronsSi 50 micronAl Kapton cable

Wieman: 21LBNL

Photo-gate Fe55 X-Ray test

• Charge collection histogram for CMOS APS – comparing two photo-gate structures to standard diode structure

• Photo-gate coupled directly to drain – no transfer gate• Photo-gate shows more complete charge collection, but…

ADC ADC

Log Linear

diode

Photo-gate 1

Photo-gate 2

full diodecharge collection

Wieman: 22LBNL

Drain current after light injection, floating n well delay

V

photo gate

transfer gate drain

floating n well

Light injection

Potential before and after injection

V=Q/C (very small)

drain currentlogdrain current

linear

200 s 200 sSmaller the signal the higher the effective resistance and the slower the transfer

Wieman: 23LBNL

Radiation resistance

• Radiation tests of the LBNL APS at the 88” cyclotron

55 MeV protons 30 year RHIC operation at 4 x design luminosity

Wieman: 24LBNL

Properties, LBNL APS

MIP (most probable) 440 e

Node C, measured with Fe55 Xray 6.1 fF

Gain ~26 V/e

Noise ( 1 pixel, CDS, Ileak subtr ) 17 e rms

Signal/Noise (9 pixel sum, CDS) 9

Signal/Noise (potential, single pix) 26

kTC reset noise (measured) 50 e rms

kTC reset noise (expected) 30 e rms

Leakage Current 0.9 fA

Wieman: 25LBNL

Silicon Cost

Ladders per Wafer 5

Ladders per Detector 24

Yield 60%

Number of Detector Copies 4

Number of Wafers 32

Wafer Cost Each 2-5 k$

Wafer Costs 64-160 k$

Mask Cost 150-200 k$

Total 214-360 k$

8 inch wafers

20 mm x 170 mm ladders