scieee AI-readable full text Open interactive document viewer

Dual-SMU Synapse, Transistor and Solar Cell Characterization Tool for Keithley 26xx Series

Jehl li kao, Zacharie

Abstract

Keithley Dual SMU Parameter Analyzer - Software Description A Python-based graphical interface for electrical characterization using Keithley 26xx dual-channel sourcemeters. This software enables automated measurement and analysis of photovoltaic devices, transistors, and neuromorphic/memristive devices with advanced data processing and visualization capabilities. Core Functionalities 1. Current-Voltage (IV) Characterization Automated JV sweeps with configurable voltage range, step size, and measurement speed (NPLC) Multi-curve overlay plotting for comparative analysis Dual y-axis visualization showing current density and power density simultaneously Semilog plotting mode for analyzing devices across wide current ranges Hysteresis measurement with configurable forward/reverse sweep cycles Dark and illuminated measurements with photovoltaic parameter extraction 2. Photovoltaic Device Analysis Automatic PV parameter extraction: Open-circuit voltage (Voc), short-circuit current density (Jsc), fill factor (FF), and power conversion efficiency (PCE) Interpolation-based Voc calculation for improved accuracy Configurable irradiance settings (W/m²) for standardized testing Batch processing with multi-sample storage and CSV export 3. Transistor Characterization Automated gate-voltage sweep measurements for OFET/TFT devices Dual-channel operation with independent gate and drain control Output characteristics generation with multi-Vgs curve families Transfer curve analysis with data export capabilities 4. Neuromorphic Synapse Characterization Pulse-read sequences for electrical, optical, and memristor-based synapses Configurable stimulus parameters: voltage/current drive, pulse width, period, and amplitude Real-time conductance monitoring across pulse trains Synaptic metrics calculation: Paired-pulse facilitation (PPF), conductance change (ΔG), potentiation/depression quantification Safety checks for high-voltage operations 5. Spike-Rate-Dependent Plasticity (SRDP) Frequency sweep characterization (linear or logarithmic scaling) Rate-dependent learning curves showing ΔG vs. spike frequency Configurable frequency range (0.1 Hz - 1 kHz+) Multi-point analysis with automated data collection 6. Spike-Timing-Dependent Plasticity (STDP) Timing-dependent plasticity window measurement Pre-post spike pair generation with precise Δt control Bidirectional plasticity characterization (LTP/LTD regions) STDP curve plotting with automatic LTP/LTD region annotation 7. Simulation Mode Hardware-free testing with physics-based device models Exponential conductance change simulation for memristive behavior Realistic noise injection for measurement validation SRDP and STDP simulation engines for protocol development 8. Instrument Communication Multi-interface support: GPIB, RS-232, and LAN/Ethernet Automatic timeout management based on measurement parameters 4-wire and 2-wire sensing modes Current compliance protection (100 nA to 1.5 A range) Autorange capabilities for current measurement 9. Data Management CSV export with embedded metadata headers Multi-sample batch storage with unique cell identifiers Derived metrics calculated and stored automatically Timestamp tracking for temporal analysis Parameter presets for common measurement protocols (LTP, LTD, PPF, high-speed) 10. User Interface Modern CustomTkinter GUI with scrollable parameter panels Real-time plotting using Matplotlib with dual-axis support Parameter display panel showing calculated metrics live Preset selector for rapid protocol switching Measurement mode tabs: Diode, Transistor, Synapse, SRDP, STDP Technical Specifications Programming Language: Python 3.x Key Dependencies: PyVISA, CustomTkinter, Matplotlib, NumPy Target Hardware: Keithley 2636A/B Dual-Channel SMU Communication Protocols: GPIB (IEEE-488), RS-232, TCP/IP Data Format: CSV with metadata headers Measurement Resolution: 0.1-10 NPLC (power line cycles) Use Cases Memristor and resistive RAM device testing Artificial synapse characterization for neuromorphic computing Organic and perovskite solar cell characterization Organic field-effect transistor (OFET) analysis Optoelectronic device measurements Solar Cells Author: Zacharie Jehl Li-KaoContact: [email protected] Source code: https://github.com/SOLIS-project

Full text

Version 1.0 SOLIS-Sourcemeter: User Manual for Keithley 2636A Current-Voltage Characterization Software Zacharie Jehl Li-Kao [email protected] October 10, 2025 Contents 1 Introduction 3 1.1 Supported Connection Types . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3 1.2 KeyFeatures ...................................... 3 2 Diode Mode: Photovoltaic and Rectifier Characterization 3 2.1 MeasurementPrinciple................................. 3 2.2 ConfigurationParameters ............................... 3 2.3 Photovoltaic Parameter Extraction . . . . . . . . . . . . . . . . . . . . . . . . . . 3 2.4 HysteresisMeasurement................................ 4 2.5 Dark vs. Illuminated Measurements . . . . . . . . . . . . . . . . . . . . . . . . . 4 3 Transistor Mode: Field-Effect Characterization 4 3.1 MeasurementProtocol................................. 4 3.2 ChannelConfiguration................................. 4 3.3 ParameterExtraction ................................. 5 3.4 DataExportFormat.................................. 5 4 Synapse Mode: Neuromorphic Device Characterization 5 4.1 Basic Synapse Measurement . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5 4.1.1 Pulse-ReadSequence.............................. 5 4.1.2 TimingParameters............................... 5 4.1.3 DriveTypes................................... 5 4.1.4 SynapseMetrics ................................ 6 4.2 Presets for Common Protocols . . . . . . . . . . . . . . . . . . . . . . . . . . . . 6 4.3 Spike-Rate-Dependent Plasticity (SRDP) . . . . . . . . . . . . . . . . . . . . . . 6 4.3.1 MeasurementProtocol............................. 6 4.3.2 SRDPParameters ............................... 6 4.3.3 DataInterpretation .............................. 7 4.4 Spike-Timing-Dependent Plasticity (STDP) . . . . . . . . . . . . . . . . . . . . . 7 4.4.1 TimingConvention............................... 7 4.4.2 Implementation................................. 7 4.4.3 STDPParameters ............................... 7 4.4.4 Expected STDP Window . . . . . . . . . . . . . . . . . . . . . . . . . . . 7 4.5 Potentiation-Depression Cycling . . . . . . . . . . . . . . . . . . . . . . . . . . . 7 4.5.1 CycleStructure................................. 8 SOLIS-Sourcemeter user manual 1 Version 1.0 4.5.2 Independent Parameter Control . . . . . . . . . . . . . . . . . . . . . . . . 8 4.5.3 CyclePresets.................................. 8 4.5.4 CycleMetrics.................................. 8 4.5.5 Real-TimePlotting............................... 8 4.6 Multi-Device Sequential Testing . . . . . . . . . . . . . . . . . . . . . . . . . . . . 9 4.6.1 Configuration.................................. 9 4.6.2 Multi-Device Workflow . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 9 4.6.3 StatisticalAnalysis............................... 9 5 Simulation Mode 9 5.1 SimulationModels ................................... 9 6 Measurement Optimization 10 6.1 NPLCSelection..................................... 10 6.2 Autorange vs. Fixed Range . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 10 6.3 4-Wire vs. 2-Wire Sensing . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 10 6.4 ComplianceSettings .................................. 10 6.5 TimeoutConfiguration................................. 10 7 Data Export and Analysis 10 7.1 CSVFileStructure................................... 10 7.1.1 DiodeMode................................... 10 7.1.2 SynapseMode ................................. 11 7.1.3 Cycle Characterization . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 11 7.1.4 Multi-DeviceData ............................... 11 7.2 ExtractingMetadata.................................. 11 8 Contact and Support 12 SOLIS-Sourcemeter user manual 2 Version 1.0 1 Introduction This software was developed as a deliverable of the MSCA Staff Exchange Project SOLIS (https://cordis.europa.eu/project/id/101183049). SOLIS-Sourcemeter is a measurement suite for electrical characterization of semiconductor devices using the Keithley 2636A dual-channel SourceMeter. The software supports three primary measurement modes: diode/photovoltaic characterization, transistor characterization, and neuromorphic synapse measurements. All measurements utilize buffered acquisition for speed and reliability. 1.1 Supported Connection Types •GPIB (IEEE-488) •RS232 (Serial) •LAN (TCP/IP) 1.2 Key Features •Real-time JV curve plotting with dual-axis power density display •Automatic photovoltaic parameter extraction (Voc,Jsc, FF, PCE) •Semi-logarithmic current visualization •Hysteresis characterization for ferroelectric and memristive devices •Advanced synapse characterization: SRDP, STDP, cycling endurance •Multi-device sequential testing •Hardware-free simulation mode for protocol development 2 Diode Mode: Photovoltaic and Rectifier Characterization 2.1 Measurement Principle The software performs voltage-sweep current measurements to generate current density vs. voltage (JV) curves. For photovoltaic devices under illumination, the software automatically calculates standard solar cell parameters. 2.2 Configuration Parameters Parameter Description Typical Range Starting Voltage Initial sweep voltage −0.5 to 0 V Ending Voltage Final sweep voltage 0.8 to 1.2 V Voltage Step Voltage increment 0.01 to 0.05 V NPLC Integration time (power line cycles) 0.01 to 10 Compliance Current limit 1 nA to 1.5 A Surface Area Active device area 0.01 to 10 cm2 Irradiance Incident light intensity 100 to 1000 W/m2 Table 1: Diode mode configuration parameters 2.3 Photovoltaic Parameter Extraction For illuminated measurements, the software calculates: SOLIS-Sourcemeter user manual 3 Version 1.0 1. Open-Circuit Voltage (Voc): Interpolated voltage where J= 0 Voc =Vlow −Jlow Vhigh −Vlow Jhigh −Jlow (1) 2. Short-Circuit Current Density (Jsc): Current density at V= 0 Jsc =−J(V= 0) [mA/cm2] (2) 3. Fill Factor (FF): Ratio of maximum power to theoretical maximum FF = |Vmp ·Jmp| Voc ·Jsc ×100% (3) 4. Power Conversion Efficiency (PCE): PCE = |Pmax| Pincident ×100% = |Vmp ·Jmp| Pincident ×100% (4) where Pincident is the irradiance in mW/cm2. 2.4 Hysteresis Measurement Enable hysteresis measurement to characterize voltage-dependent polarization or ion migration effects. The software performs forward and reverse voltage sweeps for the specified number of cycles. This is important mostly for: •Perovskite solar cells (ion migration) •Ferroelectric devices •Memristive systems 2.5 Dark vs. Illuminated Measurements Dark JV: Characterizes diode properties without photogeneration. The software does not include fitting yet. Use for: •Ideality factor determination •Reverse saturation current extraction •Shunt and series resistance analysis Illuminated JV: Measures photovoltaic performance. Ensure accurate calibration of whatever light source you use for precise PCE calculation. 3 Transistor Mode: Field-Effect Characterization 3.1 Measurement Protocol Transistor mode performs output characteristics (ID-VDS) measurements at multiple gate voltages (VGS). The software sweeps drain-source voltage while incrementally stepping the gate voltage, generating a family of curves. 3.2 Channel Configuration The selected channel controls VDS (drain-source), while the opposite channel controls VGS (gatesource): •Channel A selected: SMU A = drain, SMU B = gate •Channel B selected: SMU B = drain, SMU A = gate SOLIS-Sourcemeter user manual 4 Version 1.0 3.3 Parameter Extraction From the output characteristics, extract: •Linear regime: Mobility, threshold voltage •Saturation regime: On/off ratio, subthreshold swing •Contact resistance: From low VDS behavior 3.4 Data Export Format The CSV file contains columns for each VGS value with paired voltage-current density data. Surface area is embedded in the header for conductivity calculations. 4 Synapse Mode: Neuromorphic Device Characterization Synapse mode implements pulse-read sequences for characterizing artificial synapses and memristive devices. The software supports basic potentiation/depression measurements as well as the more advanced plasticity protocols. 4.1 Basic Synapse Measurement 4.1.1 Pulse-Read Sequence The fundamental operation sequence is: 1. Apply stimulus pulse (voltage or current) on stimulus channel 2. Wait for pulse duration (twidth) 3. Turn off stimulus 4. Wait for read delay (tdelay) 5. Apply read voltage on read channel 6. Measure current after settling time 7. Calculate conductance: G=I/Vread 8. Return to idle state 9. Wait for period completion 4.1.2 Timing Parameters Parameter Description Typical Range Stim Width Pulse duration 10 to 1000 ms Stim Period Time between pulses 50 to 2000 ms Read Delay Post-pulse settling time 5 to 100 ms Settle Time Measurement stabilization 2 to 20 ms Table 2: Synapse timing parameters Critical constraint: tperiod ≥twidth +tdelay +tsettle +tmeasurement (5) 4.1.3 Drive Types •Voltage-driven (V): Apply fixed voltage, measure current •Current-driven (I): Apply fixed current, measure voltage SOLIS-Sourcemeter user manual 5 Version 1.0 Voltage drive is preferred for most memristive devices; current drive is suitable for phasechange memory (PCM) characterization, which is not part of the SOLIS project. 4.1.4 Synapse Metrics The software calculates: •Paired-Pulse Facilitation (PPF): PPF = G2 G1 ×100% (6) Measures short-term plasticity; critical for temporal information processing. •Conductance Change (∆G): ∆G=Gfinal −Ginitial (7) ∆G%=Gfinal −Ginitial Ginitial ×100% (8) •Dynamic Range: Gmax/Gmin ratio (on/off ratio) 4.2 Presets for Common Protocols Preset Voltage Width Period Pulses LTP Moderate +1.0 V 100 ms 200 ms 50 LTD Moderate −1.0 V 100 ms 200 ms 50 PPF Test +1.5 V 10 ms 30 ms 2 High Speed +1.0 V 10 ms 20 ms 100 Table 3: Preset configurations for basic synapse measurements 4.3 Spike-Rate-Dependent Plasticity (SRDP) SRDP characterizes how synaptic weight change depends on input spike frequency which is a critical property for rate-coded neural networks. 4.3.1 Measurement Protocol The software sweeps through user-defined frequencies, applying a fixed number of pulses at each frequency and measuring the resulting ∆G. fi=(Logarithmic: 10log10(fmin)+i· log10(fmax)−log10(fmin) N−1 Linear: fmin +i·fmax−fmin N−1 (9) where i∈[0, N −1] and Nis the number of frequency points. 4.3.2 SRDP Parameters •Freq Start: Minimum frequency (1-100 Hz typical) •Freq End: Maximum frequency (10-1000 Hz typical) •Freq Points: Number of frequencies to test (5-20 recommended) •Log Scale: Logarithmic spacing (recommended for wide frequency ranges) SOLIS-Sourcemeter user manual 6 Version 1.0 4.3.3 Data Interpretation Plot ∆Gvs. frequency to identify: •Potentiation regime: Positive ∆Gincreasing with frequency •Depression regime: Negative ∆Gmagnitude increasing with frequency •Saturation: Plateau at high frequencies •Critical frequency: Transition point between LTD and LTP 4.4 Spike-Timing-Dependent Plasticity (STDP) STDP characterizes synaptic weight change as a function of relative timing between preand post-synaptic spikes, and it is the fundamental learning rule in temporal coding. 4.4.1 Timing Convention ∆t=tpre −tpost (10) •∆t>0: Pre-synaptic spike precedes post-synaptic →LTP (potentiation) •∆t<0: Post-synaptic spike precedes pre-synaptic →LTD (depression) 4.4.2 Implementation For each ∆tvalue: 1. Measure initial conductance G0 2. Apply Nspike pairs with timing offset ∆t 3. Measure final conductance Gf 4. Calculate ∆G=Gf−G0 The stimulus channel delivers pre-synaptic spikes; the read channel delivers post-synaptic spikes. 4.4.3 STDP Parameters •∆tStart: Minimum timing offset (typically −50 ms) •∆tEnd: Maximum timing offset (typically +50 ms) •∆tPoints: Number of timing points (10-20 recommended) •Spike Pairs: Number of paired spikes per timing point (20-100) •Pre-spike Level: Pre-synaptic pulse amplitude •Post-spike Level: Post-synaptic pulse amplitude •Pair Period: Time between consecutive spike pairs 4.4.4 Expected STDP Window Classic STDP follows exponential decay: ∆G(∆t) =    A+exp −∆t τ+∆t>0 (LTP) −A−exp ∆t τ−∆t<0 (LTD) (11) where A+,A−are amplitudes and τ+,τ−are time constants (typically 10-20 ms). 4.5 Potentiation-Depression Cycling Cycling characterization assesses device endurance and reproducibility by alternating potentiation and depression phases. SOLIS-Sourcemeter user manual 7 Version 1.0 4.5.1 Cycle Structure Each cycle consists of: 1. Potentiation phase: Apply positive stimulus sequence 2. Depression phase: Apply negative stimulus sequence 3. Inter-cycle delay: Recovery time before next cycle 4.5.2 Independent Parameter Control Potentiation and depression phases use separate parameter sets, allowing asymmetric characterization: •Different pulse amplitudes (e.g., +1.5 V pot, −1.0 V dep) •Different pulse widths (e.g., 50 ms pot, 100 ms dep) •Different number of pulses •Different read voltages (e.g., 0.15 V pot, 0.08 V dep) 4.5.3 Cycle Presets Preset Cycles Pot/Dep Inter-cycle Application Standard Cycle 5 ±1.0 V 1000 ms Balanced protocol Fast Cycle 10 ±1.2 V 500 ms Rapid screening High Endurance 20 ±0.8 V 2000 ms Long-term stability Asymmetric 5 +1.5/−1.0 V 1500 ms Unbalanced dynamics Table 4: Cycle preset configurations 4.5.4 Cycle Metrics The software calculates: •Mean ∆G(potentiation): Average potentiation across cycles •Mean ∆G(depression): Average depression across cycles •Reproducibility (CV): Coefficient of variation CV = σ(∆G) |µ(∆G)|×100% (12) Lower CV indicates better reproducibility. •Dynamic Range: ∆Grange =⟨Gpot,final⟩−⟨Gdep,final⟩(13) •On/Off Ratio: On/Off = ⟨Gpot,final⟩ ⟨Gdep,final⟩(14) 4.5.5 Real-Time Plotting The software provides real-time plot updates during cycling, allowing immediate observation of: •Cycle-to-cycle variability •Fatigue effects (degradation over cycles) •Asymmetric behavior between potentiation and depression SOLIS-Sourcemeter user manual 8 Version 1.0 4.6 Multi-Device Sequential Testing Multi-device mode enables automated characterization of multiple devices without manual intervention. 4.6.1 Configuration •Number of Devices: Total devices to test (2-20 typical) •Device Name Pattern: Base identifier (e.g., ”Device” generates Device 1, Device 2, ...) •Inter-device Delay: Recovery time between devices (1-5 seconds) All devices use the same cycle parameters (potentiation, depression, number of cycles). 4.6.2 Multi-Device Workflow 1. Configure cycle parameters 2. Enable multi-device mode 3. Set device count and naming 4. Run measurement 5. Software sequentially tests each device 6. Aggregated results display device-to-device variability 4.6.3 Statistical Analysis Multi-device measurements enable: •Device-to-device variability assessment •Fabrication yield estimation •Statistical reliability metrics •Outlier identification It is better used with a multiplexer on a synaptic network for example. 5 Simulation Mode Simulation mode allows protocol development and testing without hardware connection. Enable via the ”Simulate (No Hardware)” checkbox in Synapse mode. I made it at first to test the logic without having to connect the hardware, but it can also be useful to help designing experimental protocols. 5.1 Simulation Models •Basic Synapse: Exponential conductance change with learning rate Gn+1 =Gn+ (Gtarget −Gn)×α(15) where α= 0.05 (5% step size). •SRDP: Logarithmic frequency dependence ∆Gnorm(f) = log10(f+ 1) log10(100) ×A(16) •STDP: Exponential timing window ∆G(∆t) = (0.5 exp(−∆t/20 ms) ∆t>0 −0.3 exp(∆t/20 ms) ∆t<0(17) All simulations include Gaussian noise (±2%-5%) to approximate realistic behavior. SOLIS-Sourcemeter user manual 9