Clear Performance and Demonstration of a novel Clear Concept for DEPFET Active Pixel Sensors Stefan Rummel Max-Planck-Institut für Physik – Halbleiterlabor 2nd International Workshop on DEPFET Detectors and Application 1 Rummel Stefan, MPI for Physics 2nd International Workshop on DEPFET Detectors and Application

Overview ●Introduction ●The clear process ●Clear performance ●The comparison method for clear performance measurement ●Investigations on mini matrices ●New effect: Gain modulation ●The capacitive coupled clear gate 2 2nd International Workshop on DEPFET Detectors and Application Rummel Stefan, MPI for Physics

DEPFET – Depleted Field Effect Transistor ● Combination of detector grade silicon with first p-FET amplification stage in each pixel ●Potential minimum for electrons is created under the channel by sideward depletion and an additional n-doping ●Electrons in the “internal gate” modulate the transistor current ●Signal charge is removed via a clear contact Large sensitive volume due to fully depleted bulk Low noise caused by a small input capacitance and internal amplification Transistor can be switched off by external gate – charge collection is then still active! 3 2nd International Workshop on DEPFET Detectors and Application Rummel Stefan, MPI for Physics

4 DEPFET – Matrix operation ●Column parallel architecture for fast readout ●Local charge to current conversion – no charge transfer ●Low power dissipation – only one row active while readout GATE SWITCHER CLEAR SWITCHER FRONTEND 2nd International Workshop on DEPFET Detectors and Application Rummel Stefan, MPI for Physics

Readout sequence Rummel Stefan, MPI for Physics Accelerators and Detectors at the Technology Frontier, Bad Honeff ●Sample – Clear – Sample ●Difference of two consecutive samples: I SIG = Sample 1 – Sample 2 = (I PED +I SIG )-I PED T ROW 5

Question – the g q mystery ●Single pixel measurements show up to ~500pA/e - for standard devices with 6µm gate length ●Large matrices with 6µm show however pA/e -. ●Charge loss is NOT affecting the performance (PRG, VLC, Bonn)‏ ●Why do we have pA/e - amplification in current standard matrices? ●Is an incomplete clear the reason? 2nd International Workshop on DEPFET Detectors and Application 6 Rummel Stefan, MPI for Physics

The clear process 7 2nd International Workshop on DEPFET Detectors and Application Rummel Stefan, MPI for Physics

The clear process GATE CLEAR GATE CLEAR Charge collection state Potential U CLEAR ~3V Barrier partially removed Potential U CLEAR ~6V Clear - charge is removed by drift Potential U CLEAR >10V 8 2nd International Workshop on DEPFET Detectors and Application Rummel Stefan, MPI for Physics

Clear- and system performance Incomplete clear causes: ●Loss of signal pulse height ●Increase of noise Full clear: Incomplete clear: Noise peak Signal peak 9 2nd International Workshop on DEPFET Detectors and Application Rummel Stefan, MPI for Physics

Methods to judge the clear performance ●Full clear ~ Minimum current noise level ●Does not take into account the signal ●My proposal: The comparison method ●Compare signal derived by CDS with and without clear in between ●Need calibrated amplifier ●Need flexible sequencer ●Focus on the signal 2nd International Workshop on DEPFET Detectors and Application 10 Rummel Stefan, MPI for Physics

The comparison method ●g q ICDS ~ Sample 1 – Sample 2 ●g q ~ Sample 3 – Sample 2 ●Define: g q fraction = g q / g q ICDS ●g g fraction = 1: complete charge is removed within 1 clear cycle! Well defined test charge: Fe 55 : 1639e, σ=13e 2nd International Workshop on DEPFET Detectors and Application 11 Rummel Stefan, MPI for Physics

Setup & Device under test ●Single pixel setup ●Clear swing up to 20V (ILC system + Switcher 2 ~14V)‏ ●Flexible sequencer ●Low noise ●Biasing identical to ILC system (Backside, Bulk, Transistor…)‏ ●DUT: Mini matrices ●6*8 pixel ●Layout identical to 128*64 matrices ●Pixels could be investigated in their natural environment  Excellent comparability! 2nd International Workshop on DEPFET Detectors and Application 12 Rummel Stefan, MPI for Physics

Results ●Full clear possible with voltages below 10V ●Incomplete clear can not be the reason for the small g q U14 E Type W/L 7.8/6 Ids 72µA 2nd International Workshop on DEPFET Detectors and Application 13 Rummel Stefan, MPI for Physics

Internal amplification ●For each point in parameter space the g q is available! ●g q changes with clear gate voltage! ●Decrease from 430pA/e - to 240pA/e - ! 2nd International Workshop on DEPFET Detectors and Application 14 Rummel Stefan, MPI for Physics

Gain modulation ●Possible reasons: ●Charge loss? ●Decrease of current due to parasitic steering of clear gate? ●Dependence on device especially W ? 2nd International Workshop on DEPFET Detectors and Application 15 Rummel Stefan, MPI for Physics

Gain modulation - charge loss ●Spectra show no indication of 80% charge loss! ●Inner pixel -0.5V clear gate voltage show also no indication of charge loss! ●Effect must be g q related! Clear gate: -2.5V Clear gate: -0.5V 2nd International Workshop on DEPFET Detectors and Application 16 Rummel Stefan, MPI for Physics

Gain modulation – current change ●Observed current variation <4% ●Could not explain g q variation ●Gq ~ sqrt(Ids) ●A change of the current due to clear gate change Region of full clear 2nd International Workshop on DEPFET Detectors and Application 17 Rummel Stefan, MPI for Physics

Gain modulation – device dependence ●Devices with different W were investigated ●Pixel with higher width shows smaller modulation 470pA/e - – 320pA/e - ●Modulation looks like W related U25 E Type W/L 10.8/6 2nd International Workshop on DEPFET Detectors and Application 18 Rummel Stefan, MPI for Physics

Gain modulation with Clear Gate ●Reduction of parasitic capacitances ●Would explain W dependence ●Probably the candidate… 19 Potential GATE CLEAR GATE CLEAR U CLG ~0V U CLG ~-3V CLEAR CLEAR GATE GATE Potential 2nd International Workshop on DEPFET Detectors and Application Rummel Stefan, MPI for Physics

Optimal operation point ●Full clear - low clear gate voltage ●High internal amplification - high clear gate voltage ●Tradeoff neccesary! ●Higher clear high voltage improve the gain ●Switcher 4/5 allows operation in optimal region! Clear gate voltage 20 2nd International Workshop on DEPFET Detectors and Application Rummel Stefan, MPI for Physics

Clear - Summary ●Comparison method is working and confirmed that a full clear is possible with ~10V clear voltage ●New effect observed: Gain modulation ●Switcher 4/5 will allow to operate the detector with optimal settings! ●Close to ~ 500pA/e - will be available – 100e - RO noise! (without L scaling)‏ 2nd International Workshop on DEPFET Detectors and Application 21 Rummel Stefan, MPI for Physics

Capacitive coupled clear gate (CCCG)‏ 2nd International Workshop on DEPFET Detectors and Application 22 Rummel Stefan, MPI for Physics

The capacitive coupled Clear Gate ●Reducing the potential barrier via build in capacitor ●DC potential defined by resistor ●Potential change given by: (capacitive voltage divider)‏ C gd C gs R cg C cg-clear U clear U clear gate Drain Source Gate C gd C gs R cg C cg-clear U clear U clear gate C cg-drain C cg-source Drain Source Gate C cg-gate 23 2nd International Workshop on DEPFET Detectors and Application Rummel Stefan, MPI for Physics

Direct measurement of the coupling ●Direct measurement possible with test structure with additional transistor 2nd International Workshop on DEPFET Detectors and Application 24 Rummel Stefan, MPI for Physics

Coupling ●Measured and couplings extracted from layout are in agreement ●Clear Gate – Clear: ~210mV/V ●Clear Gate – Gate: ~160mV/V ●Typically voltage swing of Clear and Gate are similar ~ 10V ●Twofold improvement: ●Clear Gate is lowered when row select is applied ●Clear Gate is increased above the DC value when reset pulse is applied 25 Row select - Gate Reset – Clear pulse Clear Gate DC Value Time Amplitude 2nd International Workshop on DEPFET Detectors and Application Rummel Stefan, MPI for Physics

Cd 122 spectrum ●Internal amplification significantly improved to 470 pA/e - ! ● Readout noise lowerd to 220e - Capacitive coupled part 470pA/e - Uncoupled part 350pA/e - Cd 122 : 22keV 128 x 64 Pixel matrix Spectrum taken after optimization 26 2nd International Workshop on DEPFET Detectors and Application Rummel Stefan, MPI for Physics

Summary ●Clear performance: ●Full clear possible with ~10V Clear high voltage ●New effect: g q modulation ●Even without CCCG the new generation of Switcher chips will allow to obtain a g q of ~500pA/e - CCCG: ●The capacitivly coupled Clear Gate scheme is working ●Consitent picture – couplings, gain ●The internal amplification is significantly improved – 470pA/e Clear swing of 10V ●The readout noise is lowered to 220e - from 330e - ●CCCG desings will be also present in PXD6 27 2nd International Workshop on DEPFET Detectors and Application Rummel Stefan, MPI for Physics

2nd International Workshop on DEPFET Detectors and Application 28 Rummel Stefan, MPI for Physics

2nd International Workshop on DEPFET Detectors and Application 30 Rummel Stefan, MPI for Physics