Any discussion of optimizing low-current and high-resistance measurements must include a definition of exactly what these terms mean, the instrumentation required, and appropriate techniques for connecting these instruments to the device under test (DUT).
For this discussion, let’s define low currents as less than roughly 10 nA. Typical low-current measurement applications include assessing FET gate leakage current, photodiode current, capacitor leakage current, ion-beam current, nanoelectronic device performance, and photomultiplier tube output. These are measurements for which the choice of the most appropriate instruments and measurement techniques are crucial to avoiding measurement errors.
We can define high-resistance measurements as being greater than 1 GO. These applications include measuring the resistivity of insulators; testing the insulation resistance testing of devices such as multi-conductor cables, connectors, and printed circuit boards; and measuring the resistance of high mega-ohm resistors.
The keys to ensuring optimal low current and high resistance measurements include:
Low-current measurement instrumentation
- Choosing an instrument with the current sensitivity the application requires
- Using suitable cables and connections to minimize offsets
- Verifying system offset in order to determine the noise floor
- Identifying potential error sources to minimize unwanted generated currents.
If the application requires measuring very low currents, choosing the correct measurement instrument is crucial. Most low current DC measurements are made using electrometers, picoammeters, or source-measurement units (SMU). An electrometer (see figure 1) is a highly refined DC multimeter that can measure low current, high resistance, voltage with a high input impedance, and charge. A picoammeter is an instrument that measures only low current and possibly high resistance if it has a built-in voltage source. SMUs can source current and voltage and also measure both current and voltage. Not all SMU offer low current capability, however, so review the specifications carefully.
Figure 1. Electrometers like this model 6517B high-resistance meter are highly refined DC multimeters.
Digital multimeters (DMMs) generally aren’t suitable for measuring very low current because they have excessive input offset current, insufficient current resolution, and a high input voltage drop.Current measurement methods
The two basic techniques for measuring current are the shunt ammeter method (used by virtually all DMMs so it offers limited current sensitivity) and the feedback ammeter method (used by instruments like electrometers, picoammeters, and SMUs). Although the ideal ammeter has zero resistance, all real ammeters suffer from some internal resistance. To minimize loading errors, the internal shunt resistance of the ammeter should be much less than the resistance of the DUT.
Figure 2a illustrates a shunt ammeter circuit, which is a basic voltage amplifier circuit with a shunt resistor on the input to form the ammeter. The input current flows through the shunt resistor and the voltage drop is measured and the current calculated. The output voltage Vout
is defined by:
Figure 2. Shunt ammeter circuit (a) vs. feedback ammeter circuit (b).
In instruments that use a feedback ammeter circuit (figure 2b), the input current flows through the feedback resistor (RF
). The low input bias current of the amplifier changes the current by a negligible amount. The amplifier output voltage is calculated as the inverse product of the input current multiplied by the feedback resistor:
In this case, the output voltage is a measure of the input current and the feedback resistor determines overall sensitivity. The low voltage burden and corresponding fast rise time are achieved by the high-gain op-amp, which forces the input voltage to be nearly zero.
When weighing the suitability of an instrument for low-current applications, it is important to take time to review its accuracy specifications. Instrument data sheets can use different formats to express accuracy, however, which can complicate making comparisons. Gain and offset errors may be combined into a specification that expresses accuracy in terms of a percentage and a number of counts of the least significant digit. Accuracy may also be expressed in parts per million.