Transcript Electrical Standards - Pylon Electronics Inc.

Electrical Measurement
Techniques
for Nanometrology
Speaker/Author: Richard Timmons, P.Eng.
President, Guildline Instruments
[email protected]
Tel: 1.613.283.3000; Fax: 1.613.283.6082
2007
Guildline Instruments Limited
Presentation Overview
DC Electrical Measurements

Nanoscale Range
Low And High Resistances
Low Currents
Low Voltages
Theoretical Frameworks
Techniques And Tips To Improve Accuracy
2007
Guildline Instruments Limited
Electrical Standards - Resistance
All Electrical Standards Traceable


To National Metrology Institutes
Via17025 Accredited Calibrations
DC Resistance Standards

1 µΩ (10-6) to 10 PΩ (10-16)
Uncertainties Range from 0.2 to 5000 ppm


Research Into 0.1 µΩ and Smaller Values
Temperature Stabilized Standards
Better Than Traditional Oil Based Standards
Best Uncertainties 0.2 ppm, Annual Drift < 1.5 ppm
Temperature Coefficient < 0.005 ppm

2007
Intrinsic Standard Is Quantum Hall at 12906.4035Ω
Guildline Instruments Limited
Electrical Standards - Current
Current

Current Shunts 1 µAmp to 3000 Amps
Best Uncertainties 1 ppm to 500 ppm
Stable, Linear Performance With Respect to Power

Primary Standard
Current Balance Between 2 Coils of Known Mass
and Dimensions With Uncertainty of 15 ppm
Practical Realization of Ampere

2007
From 1A = 1V / 1Ω With Better Than .001 ppm
Uncertainties
Guildline Instruments Limited
Electrical Standards - Voltage
Voltage

Typically 1 V to 10 V
Best Uncertainties < 1.0 ppm

Intrinsic Standard Is Josephine Junction Array
Typical Output In mV to 1V Range With Best
Uncertainties In the 0.01 to 0.001 ppm Range
Current Research on Stacked Josephine Junction
Arrays to Get Higher Voltages
Precision Voltage Dividers Used to Transfer To
Range of Nanovolts to Kilovolts
2007
Guildline Instruments Limited
Resistance Measurements
Source Current / Measure Voltage
Source Voltage / Measure Current
Low Resistance Measurements
High Resistance Measurements
2007
Guildline Instruments Limited
Source Current / Measure Voltage
Best for Low Resistance Measurements (< 1kΩ)

Voltage Sources Noisier Than Current Sources For Low
Impedance
The Johnson Voltage Noise At Room Temperature
(270ºK)

Simplifies to:
k = Boltzmann’s Constant, T = Absolute Temperature of Source (ºK)
B = Noise Bandwidth (Hz), and R = Resistance of the Source (Ω)
As DUT Resistance (R) Decreases Noise Voltage
Decreases
2007
Guildline Instruments Limited
Source Voltage / Measure Current
Best for High Resistance Measurements


> 10 kΩ
Voltage Sources More Stable When Driving High
Impedance
The Johnson Current Noise At Room
Temperature (270ºK)
B = Noise Bandwidth (Hz), and R = Resistance of Source (Ω)
As DUT Resistance (R) Increases Noise
Current Decreases
2007
Guildline Instruments Limited
Comparative Results Sourcing Current
Versus Sourcing Voltage
Summary of 50
Measurements
Made at Three
Resistance Values
Using a Guildline
DCC Bridge
Sourcing Both
Current and
Voltage
2007
Test (Ω)
Source
Current
Uncertainty
(ppm)
Source
Voltage
Uncertainty
(ppm)
1k-1k
0.005
0.206
10k-10k
0.011
0.003
100k-100k
0.217
0.003
Guildline Instruments Limited
LOW RESISTANCE MEASUREMENT
1k – 1k
Source Voltage
3V, 0.206 ppm Std. Dev.
Source Current
3.16mA, 0.005 ppm Std. Dev.
At 1kΩ and Lower, Sourcing Current Gives Much
Better Measurements
2007
Guildline Instruments Limited
MEDIUM RESISTANCE MEASUREMENT
10k – 10k
Source Voltage
10V, 0.003 ppm Std. Dev.
Source Current
1mA, 0.011ppm Std. Dev.
The 10 kΩ Resistance Level Is the Approximate
Transition Point At Which Both Voltage and Current
Methods Perform Equally Well With Respect to
Measurement Noise
2007
Guildline Instruments Limited
HIGH RESISTANCE MEASUREMENT
100k – 100k
Source Voltage
32V, 0.003 ppm Std. Dev.
Source Current
0.32mA, 0.217 ppm Std. Dev.
At 100 kΩ and Higher Sourcing Voltage Gives
Much Better Measurements
2007
Guildline Instruments Limited
Very Low Resistance Measurements
100 µΩ Resistance
Standard (Guildline
9334A)
Below 1 mΩ
Recommended to
Use Current Range
Extenders
Up to 3000A
Uncertainties of 10-8
ppm or Better
2007
Serial #
50A
75A
100A
68343
99.9871 99.9871 99.9888
69181
99.9803 99.9810 99.9826
Guildline Instruments Limited
Very Low Resistance Measurements
(cont)
May Need Low Currents




Saturation Current For Nanoscale Materials Often
Very Low
Self Heating Effects Create Measurement Errors and
Excessive Heat Can Damage DUT
Exception Is Super-Conducting Materials
Current Comparator (CCC) Bridges Can Measure
Down to 10-9 Ω With Low Currents
Thermal Stability Very Important


2007
For Both Resistance Standard and DUT
Stable Air Baths (0.001 °C)
Guildline Instruments Limited
Very High Resistance Measurements
DCC bridges measure up to 1 GΩ

Provide Better Uncertainties At and Below
100 MΩ
Best Uncertainties of 0.02 to 0.04 ppm For MultiRatio Bridges
Teraohmmeters (i.e. electrometer based)
Better Above 1G

2007
Measure From 1 MΩ up to 10 PΩ (1016) With
Direct Measurement Uncertainty Ranging
From 0.015% to 5% Across This Range
Guildline Instruments Limited
Very High Resistance Measurements
(cont)
Teraohmmeter With
Multi-Ratio Direct
Transfer Provides
Best Uncertainties [1]
Transfers (25) To Known
1G, 10G and 100G
Standards Using
Known 100M
Standard (Ratios Up
to 1:1000)
Current Research to
1017Ω Using 1014Ω
Standard.
2007
Resistor
Charted
Direct
Transfer
Nominal Uncertainty Reading Uncertainty
Value
(ppm)
(ppm)
(ppm)
(Ω)
100M
18
150
30.9
1G
41
200
33.7
10G
106
600
32.7
100G
94
800
46.6
Guildline Instruments Limited
Low Current Measurements
Generate or Measure Accurate and
Traceable Low Value Currents



2007
Use Commercial Voltage Standard and
Accurate High Value Resistance Standards
Traceable Reference Currents Down to 50 fA
(10-15 A)
Can be Verified Using a Teraohmeter [2]
Guildline Instruments Limited
Low Current Measurements
(cont)
Guildline 6520
Teraohmmeter
With Guildline
9336/9337
Resistance
Standards [2]
Uncertainties Can Be
Improved by the
Substitution
Method [1]
2007
Resistor
9336/9337
Teraohmmeter
Test Voltage
Effective
Current
Uncertainty
100k
1V
10 µA
0.025 %
1M
1V
1 µA
0.025 %
10M
10 V
1 µA
0.025 %
100M
10 V
100 nA
0.015 %
1G
10 V
10 nA
0.02 %
10G
10 V
1 nA
0.06 %
100G
10 V
100 pA
0.08 %
1T
10 V
10 pA
0.1 %
10T
10 V
1 pA
0.2 %
100T
10 V
100 fA
0.3 %
1P
10 V
10 fA
1 %
10P
10 V
1 fA
5 %
Guildline Instruments Limited
Low Voltage Measurements
In Order to Prevent Damage

Unless Material Is Super-Conducting
Nanovolt Meters Can Measure in the
Picovolt (10-12) Range
Johnson Noise (i.e. Motion of Charged
Particles Due to Thermal Energy) Limits
Accuracy of Low Voltage Measurements
2007
Guildline Instruments Limited
MEASUREMENT TECHNIQUES
AND TIPS
Temperature Effects
Digital Filtering
DC Reversal Techniques
Humidity Effects
Electromagnetic Interference (EMI)
Connectors and Leads
Guarding
Grounding
Settling Times
Direct Measurement With No Amplification
2007
Guildline Instruments Limited
Temperature Effects
1.0 µΩ Resistance
Standard (Guildline
9334A)
Serial #
21°C
23°C
25°C
t/c of 8.5 ppm/°C
µΩ
µΩ
µΩ
(8.5-12Ω or 8.5 pΩ)
68559 1.000048 1.000065 1.000083
Best Thermometry
Bridges < 0.025 ppm 68560 0.999997 1.000015 1.000032
Ruthenium Oxide Probe
68561 1.00033 1.00034 1.00035
(RTD) For < 1 ºK
needs 75 kΩ
Stable Air Baths At
<
1 °mK
2007
Guildline Instruments Limited
Digital Filtering
Order of Magnitude of Additional Accuracy
Large Number of Tests


Reduces the Bandwidth of the Noise
Ex: Remove ‘Outlier’ Measurements > k3
( i.e. > 3 x standard deviation)
Dynamically Alter the Sampling Times

Increase If Measurement Stable
If Periodic, Synchronize To a Clock

Telecommunications Industry
Analyze Total Set of Test Results

2007
Post Experiment Analysis With PC
Guildline Instruments Limited
Digital Filtering
(cont)
Sophisticated
Techniques Include
Profiling Noise,
Excitation Effects,
Systematic Errors,
and Other Effects
With a Suitable
Mathematical Model
Use Weighted
Coefficients
Ex: Closure Error For a
Multi-Ratio Guildline
DCC bridge [3]
2007
Correction
Method
Relative
Improvement
(ppm)
Uncorrected
(Baseline
Measurement)
0.000
Rounding
0.050
Linear
Interpolation
0.061
Logarithmic
Weighting
0.084
Guildline Instruments Limited
DC Reversal Techniques
Polarity Reversal



Eliminate Thermal EMFs
Reduces the Effect of White Noise
Increases the Signal-To-Noise Ratio
Can Be Optimized


2007
Faster When Measured Parameter Is
Changing
Slower When Measured Parameter Is Stable
Guildline Instruments Limited
Humidity Effects
Make Measurements In a Controlled, Low
Humidity Environment

Essential If DUT Absorbs Water
Use High Quality Insulators

2007
Teflon, Polyethylene, Sapphire
Guildline Instruments Limited
Electromagnetic Interference
(EMI)
EMI Noise In Most Laboratories

Florescent Lights, Cell Phones, Fixed Point Temperature
Furnaces, Electric Motors, AC Electrical Power Lines
Ambient EMI Noise Often Higher Than Nanoscale
Electrical Measurements
Instruments Have Built-in EMI Noise

Display Screens, Microprocessors / Microcontrollers, Power
Supplies
EMI Shielding For Both Measurement Circuitry and DUT

High Quality Air Baths Provide Both EMI Shielding and
Temperature Stability
Power Line Filters
2007
Guildline Instruments Limited
Connectors and Leads
4 Terminal Mode



Most Accurate Method for Measuring Small Resistances
Corrects For Lead Resistance
Allows Longer Test Leads
Current Supply Compliance Important
Very Low Resistances May Have Greater Voltage Drop
Across Leads and Connectors Then Across Shunt
Condition of Connectors, Cleanliness Important


Poor Measurements From Cracked Terminals, Dirty Contacts,
Moisture Absorbed By Standards and DUTs
Errors As High As 10 ppm
High Resistance Needs Very Good Insulation
2007
Guildline Instruments Limited
Guarding
Conductor Connected To Low Impedance Point
In Circuit That Is At Nearly Same Potential As
High Impedance Lead Being Guarded
Reduces Leakage Currents and Noise In Test /
Measurement Circuits


Very Important For High Resistance Measurements
Measurement Instruments Should Provide Guarded
Connection Terminal
Reduces Effect Of Shunt Capacitance
2007
Guildline Instruments Limited
Grounding
Single Point Ground For All Components In Test Setup
Including DUT

Avoids Ground Loop Currents Between Measurement Circuit
and DUT, or Measurement Circuit and Test Fixture
Noisy Power Lines

Largest Contributor Is Typically PCs
NOT Good Measurement Practice To Connect Different
Components Of Test Setup To Different Power Outlets

Power Line Grounds May Not Be At Same Electrical Potential,
Thus Creating Spurious Currents
NOT Good To Connect Instrument’s Common Ground To
Chassis Ground (i.e. Power Line Ground)
2007
Guildline Instruments Limited
Settling Times
Needed To Overcome Capacitance
Effects, Self-Heating Effects, Dielectric
Absorption

Present In Measurement Instruments,
Standards, Cabling, DUT
Longer Settling Times Very Important For
Resistances > 100 kΩ
2007
Guildline Instruments Limited
Direct Measurement With No
Amplification
NOT Recommended To Use Operational
Amplifiers or Other Techniques To
Increase the Measured Signal


Will Proportionally Increase Noise
Operational Amplifiers Or Other Circuitry Will
Introduce Additional Noise
Need Instruments Capable Of Directly
Measuring Electrical Properties At Very
Low Values
2007
Guildline Instruments Limited
References
[1] Mark Evans and Nick Allen, Guildline Instruments
Limited, Evaluation of a Concept for High Ohms
Transfers at Ratios > 10:1, 2007 Conference
Proceedings of the NCSL International Annual
Workshop and Symposium.
[2] Mark Evans, Application of the Guildline Model
6520 Teraohmmeter for the Nuclear Power Industry,
White Paper, Guildline Instruments Limited.
[3] Mark Evans and Xiangxiao Qiu, P. Eng., Guildline
Instruments Limited, Application of Software Enhanced
DCC Bridge Measurement, 2005 Conference
Proceedings of the NCSL International Annual
Workshop and Symposium.
2007
Guildline Instruments Limited
Providing Precision
Measurement Solutions
Guildline Instruments Limited
2007
Guildline Instruments Limited