Analysis system optimizes low current measurements

I Analysis System Optimizing Low Current Measurements - Introduction
Many critical applications require the ability to measure low currents – in the pA range or less. These applications include determining the gate leakage current of FETs, testing sensitive nanoelectronic devices, and measuring leakage currents of insulators or capacitors.

The Model 4200-SCS Semiconductor Characterization System, when equipped with the optional Model 4200-PA Remote Preamplifier, provides exceptional low current measurement capability with resolutions down to 1E–16A. Successfully measuring low currents depends not only on using a very sensitive ammeter, such as the Model 4200-SCS, but also on properly setting up the system’s Interactive Test Environment (KITE) software, using low-noise fixtures and cabling, allowing adequate settling time, and employing techniques that prevent unwanted currents from degrading measurement accuracy. This article describes the best solutions for optimizing low current measurements with the Keithley Model 4200-SCS. Offset

Current in a Measurement System
One of the first steps in configuring a system for ultra-low current measurements is to determine the offset and leakage currents of the entire measurement system, including the Model 4200-SCS itself, the connecting cables, the switch matrix, the test fixtures, and the probes. This determines the noise floor limit of the entire system and sets a starting point to improve if possible. Start by measuring the offset of the source measure unit (SMU) and continue to add measurement circuit components until all components except the device under test (DUT) are connected. Measure directly from the 4200-SMU with a 4200-PA remote preamplifier using KITE software.

II Analysis System Optimization of Small Current Measurements - Internal Offset
For an ideal ammeter, its reading should be zero when its input terminals are left open. However, real ammeters do have a small current when the input is open. This current is called input offset current and is due to the bias current of the active devices and the leakage current flowing through the insulators in the instrument. The offset current generated within the SMU is included in the technical specifications of Keithley's Model 4200-SCS. As shown in Figure 1, the input offset current is added to the current being measured, so the meter measures the sum of the two currents.

Figure 1. SMU input offset current.

The offset of each 4200-SMU with a 4200-PA preamplifier is measured with nothing connected to the Force HI and Sense HI terminals except the metal caps. These three-pin metal caps are included with the system. The SMU should be warmed up for at least one hour with the metal caps connected to the preamplifier's Force HI and Sense HI terminals before all measurements are made. If the system has KTEI version 7.1 or later installed, the offset current can be measured using the project named "LowCurrent" in the following directory: C:S4200kiuserProjectsLowCurrentOpen
this project and select SMU1offset ITM. Click the Graph tab and run the test. The results should be similar to the graph shown in Figure 2. It may be necessary to use the Auto Scale feature to scale the curve appropriately. The Auto Scale feature can be found by right clicking on the graph. With the 4200-PA preamplifier connected to the SMU, the offset current should be in the femtoamp range. The current offset can be positive or negative. Verify these results against the published specifications for the 4200-SCS ammeter.

Repeat this test for each SMU in the system using a separate ITM. The LowCurrent project has an ITM that can measure the offset current for the 4 SMUs with preamplifiers.

Systems running KTEI software prior to version 7.1 can also easily measure the offset current. Follow these steps to create a test to measure SMU1:

1. In the project that was created, open a new Device Plan for a generic 2-terminal device.
Create a new ITM called SMU1Offset. Select SMU1 for Terminal A and GNDU for Terminal B.

Figure 2. Offset current measurement of SMU1.

1. In the Definition tab, set the following:
SMU source measurement configuration : Voltage bias 0V, 10pA fixed current range.
Timing menu : Mute speed, Sample mode, 0s interval, 20 samples, 1s hold time, check Enable time stamp.
Formula calculator : Create a formula to measure noise using standard deviation, NOISE=STDDEV(A1).
Create another formula to measure average offset current: AVGCURRENT=AVG(A1).

2. In the Graph tab, set the following (right-click on the graph):
Define graph : X-axis: Time
Y1-axis : Current (A1)

Data variables: Select NOISE to display on the graph. Select AVGCURRENT to display on the graph.

Once configured, save the test and run it. The results should be similar to the graph shown in Figure 2. Repeat this test for all SMUs in the system.

Performing an auto-calibration procedure in KITE can optimize the input offset current specification. To perform an SMU auto-calibration, click “SMU Auto Calibration” in the KITE Tools menu. Allow the system to warm up for at least 60 minutes after power-up before performing an autocalibration. Nothing should be connected to the Force HI and Sense HI terminals of the SMU except the metal caps. The autocalibration procedure adjusts the current and voltage offsets for all sourcing and measuring functions of all SMUs in the system. Do not confuse this with a full system calibration, which should be performed annually at the Keithley factory.

After completing the SMU autocalibration, you can repeat the offset current measurement.

III Analysis System Optimization for Low Current Measurements - External Offset
Once the offset current of the ammeter has been determined, the rest of the system is added to the test circuit in stages, verifying the offset of the rest of the system by repeating the current (0V) vs. time plot (using the "Append Run" button shown in Figure 3). Finally, measure the end of the probe tip or the test fixture with no device connected in the "up" position. This process will help identify any trouble spots, such as shorted cables or instabilities in the measurement circuit. However, be aware that connecting and disconnecting cables will induce current in the circuit. For ultra-low current measurements, it may be necessary to wait several minutes to hours after changing the connections in the test circuit to allow stray currents to decay. The graph in Figure 4 shows the offset under the following conditions: 1) a metal cap is worn on the Force HI terminal of the SMU; 2) only a triaxial cable is connected to the preamplifier; and 3) there is a probe in the "up" position through the Keithley Model 7174A Low Current Switch Matrix to the probe station. [page]

Figure 3. Append button

Figure 4. Offset current measurement of the entire test system

Repeat the test to determine the leakage current in the measurement circuit by applying a test voltage while generating a current-time graph. In the actual measurement of the DUT, the test voltage is used instead of zero bias. Any leakage in the test fixture and cables will now be measured and plotted. If the leakage is too high, the measurement circuit can be adjusted to reduce it. For information on how to reduce leakage, see the "Leakage and Protection" section of this article.

IV Measurement Error Sources and Methods to Reduce Errors
Once current offsets, leakage currents, and any instabilities have been identified, taking steps to reduce measurement errors will help improve measurement accuracy. These error sources include insufficient settling time, electrostatic interference, leakage currents, friction effects, piezoelectric effects, contamination, humidity, ground loops, and source impedance. Figure 5 summarizes the magnitudes of some of the currents discussed in this section.

Figure 5. Typical amplitude of generated current.

Settling Time and Timing Menu Settings
The settling time of the measurement circuit is especially important when measuring low currents and high resistances. Settling time is the time it takes for the measurement to stabilize after applying or changing a current or voltage. Factors that affect the settling time of the measurement circuit include shunt capacitance (CSHUNT) and source resistance (RS). Shunt capacitance is due to connecting cables, test fixtures, switches, and probes. The higher the source resistance of the DUT, the longer the settling time. The shunt capacitance and source resistance are marked in the measurement circuit of Figure 6.

Figure 6. SMU measurement circuit including CSHUNT and RS.

The settling time is the result of the RC time constant, τ, where:
τ = _RSCSHUNT
Below is an example of calculating the settling time, assuming CSHUNT = _10pF and RS ₌ 1TΩ, then:
τ = 10pF × 1TΩ = 10s
Therefore, the settling time required for the reading to settle to 1% of the final value is 5 times τ, or 50 seconds. Figure 7 shows the exponential response of an RC circuit to a step voltage. After one time constant (τ = RC), the voltage rises to 63% of its final value.

Figure 7. Exponential response of an RC circuit to a step voltage.

To successfully measure low currents, it is important to allow enough time for each measurement, especially when sweeping voltages. Settling time can be added in the Timing menu in the Sweep Delay field for sweep mode and in the Interval time field for sample mode. To determine how much interval time to add, measure the settling time for the DUT to settle to a step voltage by plotting the current vs. time. The step voltage should be the bias voltage used in the actual measurement of the DUT. Settling time can be measured using the ITM in the LowCurrent project. The #Samples in the Timing menu should be increased appropriately to ensure that the settled readings are displayed in the graph. When measuring low currents, use Quiet Speed ​​Mode or add additional filtering in the Timing menu. Note that this is a trade-off between noise and speed. More filtering and delay will reduce noise, but will also reduce measurement speed.

V Electromagnetic Interference and Shielding
Electrostatic coupling or interference occurs when a charged object is close to the circuit being measured. At low impedance, the effect of interference is not significant because the charge disappears quickly. However, high resistance materials do not decay the charge quickly, which can result in unstable and noisy measurements. Typically, electrostatic interference becomes a problem when the measured current is ≤1nA or the measured resistance is ≥1GΩ.

To reduce the effects of electrostatic fields, the circuit being measured can be enclosed in an electrostatic shield. Figure 8 shows the huge difference between unshielded and shielded measurements of a 100GΩ resistor. Unshielded measurements are much noisier than shielded measurements.

Figure 8. Comparison of shielded and unshielded measurements of a 100GΩ resistor.

The shield can be just a simple metal box or mesh that surrounds the circuit under test. Commercial probe stations often enclose sensitive circuits in an electrostatic shield. The shield is connected to the measurement circuit LO terminal, which is not necessarily grounded. For the 4200-SCS, the shield is connected to the Force LO terminal, as shown in Figure 9.

Figure 9. Shielding high-resistance devices

Take the following steps to minimize erroneous differential currents caused by electrostatic coupling:
• Shield the DUT and electrically connect the shield to the test circuit common, the Force LO terminal of the 4200-SCS.
• Keep all charged objects (including people) and conductors away from sensitive areas of the circuit.
• Avoid movement and vibration near the test area.

VI Leakage and GuardI
Leakage is the error current through a (leakage) resistor when voltage is applied. This error current becomes a problem when the impedance of the DUT is comparable to the impedance of the insulators in the test circuit. To reduce leakage, use high-quality insulators in the test circuit, reduce humidity in the test lab, and use guards.

A guard is a conductor driven by a low-resistance source with an output at or near the potential of a high-resistance terminal. Guard terminals are used to protect the test fixture and cable insulation resistance and capacitance. The guard is the core shield of the triaxial connector/cable, as shown in Figure 10.

Figure 10. Conductors of the 4200 triaxial connector/cable

[page]

Do not confuse guarding with shielding. Shielding usually means using a metal fence to prevent electrostatic interference from affecting a high impedance circuit. Guarding means using an added low resistance conductor, maintained at the same potential as the high impedance circuit, which will intercept any interfering voltage or current. Guarding does not necessarily provide shielding. The following figure

shows two examples of guarding: 1) using guarding to reduce leakage caused by the test fixture, and 2) using guarding to reduce leakage caused by cable connections.

Figure 11 shows the principle that guarding can eliminate leakage through the isolation insulator in the test fixture. In Figure 11a, leakage current (IL) passes through the isolation insulator (RL). This leakage current is added to the current coming from the DUT (IDUT) and then measured by the SMU ammeter (IM), which can adversely affect the accuracy of low current measurements.

Figure 11. Using guarding to reduce leakage in a test fixture.

In Figure 11b, the metal mounting plate is connected to the SMU's guard terminal. The voltage on the top and bottom of the isolation insulator is close to the same potential (0V drop), so there will be no leakage current in the isolation insulator to affect the measurement accuracy. Since the metal mounting plate will be at guard potential, the metal shield must be connected to ground for safety reasons.