Impedance Measurements on Grounded Components
The MFIA is a high-precision impedance analyzer. It measures voltage, current and phase to calculate impedance. As is the case for most impedance analyzers, the measurement circuit is designed so that the device under test (DUT) is operated under floating conditions, which means that none of the terminals of the DUT is connected to the ground. This configuration ensures that any current through the DUT will go through the current input terminal of the MFIA and that there is no additional spurious current path.
When measuring grounded components, the DUT is no more under floating conditions: this is problematic for most impedance analyzers. With the MFIA, we can modify the measurement setup to make it suitable to measure grounded components. Such measurements might be necessary when the DUT is integrated into a greater circuit (for example during diagnosis measurements) or, in a different scenario, when a large DUT has to be measured and cannot be easily removed from the (electrical) ground.
Building a "Current Sensor"
For measuring grounded components, we have to make some modifications to the current measurement setup. We use a shunt resistor, which is a resistor with four terminals and a well-defined resistance. One option is to use the 1 kΩ resistor on the MFITF carrier that comes with the MFITF Test Fixture, which is itself supplied with the MFIA. Another option is to solder a suitable resistor on an empty carrier (also included with the MFITF). Figure 1 shows the modified setup with the shunt resistor connected to the MFITF, and the description of the required connections is based on the labels of the MFITF. The current sensor works as such: by measuring the voltage drop over the shunt and dividing it by the resistance of the shunt, we obtain the current through the shunt as I = V/R.
The outer connectors of the MFITF, LCUR and HCUR, are used for the current path through the shunt, while the inner connectors, LPOT and HPOT, are used to measure the voltage drop over the shunt resistor. Typically, the shunt resistor has a smaller resistance than the DUT to ensure that most of the test signal voltage drops over the DUT and that the test voltage does not have to be increased significantly to reach the desired test signal amplitude on the DUT.
We could use the Auxiliary Input of the MFIA to measure that voltage. However, the voltage drop over the shunt resistor is supposed to be smaller than the voltage drop over the DUT, and the 'Aux Input' channel only has one measurement range, 10 V. To achieve the best results, the larger voltage should thus be measured on the 'Aux Input'. As shown in Figure 1, the voltage of the DUT (the grounded component) is measured through 'Aux Input', whereas the shunt resistor is monitored through the 'Signal Input' channel. Some additional connectors and cables are required, but those are all part of standard measurement equipment.
Beware that 'Aux Input' is single-ended and that the shield of the BNC socket is connected to the ground, as applies to all shields on the front panel of the MFIA. This means that the side of the MFITF that connects to the Auxiliary Input shield should correspond to the side of the DUT that is connected to the ground.
Figure 1: Schematic of the setup used to measure the impedance of grounded components with the MFIA. The numbers next to the connectors correspond to the numbers in brackets in the explanation of the setup found in the text.
Complete Measurement Setup
The DUT is integrated into the circuit as shown in Figure 1. By default, one side of the DUT is connected to the ground (connection 1 in Figure 1). To measure the voltage drop over the DUT, the other side is connected to 'Aux Input 1' (2). We further connect that side to the MFITF to measure the current (3). The two middle connectors of the MFITF are connected to LPOT and HPOT, respectively (4 and 5), to measure the voltage drop over the measurement shunt. Finally, the remaining connector of the MFITF is connected to 'Signal Output' on the MFIA (6). In this configuration, 'Signal Input' remains unconnected. One could argue that it should be connected to the right side of the DUT to close the circuit, but this is the connection that goes straight to the ground anyway (1) and an additional connection to the MFIA will only lead to errors. Tests have shown that the setup described here is also less prone to picking up noise from external sources. We saw minor improvements in the data quality if the ground (1) is connected to the MFIA's ground on the back panel (with a banana clip), but this depends on the test environment and the quality of the ground to which the DUT is connected. In principle, connecting all parts with one reliable ground is beneficial, but the more cables and connections are used, the higher the risk of creating a ground loop that might pick up noise from the line or other magnetic fields. It is thus quite difficult to distill a general rule here. One possibility is to test the quality of the signals with the LabOne® Scope module: in the FFT of 'Signal Input', only the test frequency should show up as a peak. Additional peaks at a frequency that is a multiple of the test frequency indicate a ground loop or another grounding issue.
Figure 2: Screenshot of LabOne where the modifications compared to standard impedance measurements are highlighted. The numbers correspond to the numbers in brackets in the explanation of the settings found in the text.
Setting up LabOne
We now turn to LabOne, where we need to make some adjustments to the default conditions to correctly measure our grounded component in the described setup. For that, we open the Advanced tab of the Impedance Analyzer (IA). First, we need to assign the signals correctly. Beware that the options in the 'Current Input' and 'Voltage Input' menus (1 and 2 in Figure 2) only allow for a configuration that will calculate the admittance, which is the inverse of the impedance. This means that the measurement results displayed in the IA tab do not show the actual impedance of the DUT, but the real and imaginary parts of its admittance plus its magnitude and phase angle. To display the real and imaginary parts of the impedance directly, we need to change the impedance representation to 'G Conductance / B Susceptance' (see 3 in Figure 2). Indeed, conductance is the real part of the admittance and susceptance is its imaginary part; given that we swapped impedance and admittance with the considered setup and settings, the conductance actually corresponds to the real part of the impedance of the DUT and the susceptance corresponds to the imaginary part of the impedance of the DUT. The numeric values displayed for these quantities will be correct even though the units will be Siemens, which strictly speaking is not correct. For example, a value of 1.00314 kS (4) means that the real part of the impedance of the measured DUT is 1.00314 kΩ. It is also possible to set up a Nyquist plot in the XY representation by using 'Impedance 1 Rep Param 1' as 'X Signal' and 'Impedance 1 Rep Param 2' as 'Y Signal'.
To set up the user interface for our purpose, we assign 'Aux In 1' (1) to the current signal and keep the voltage signal as 'Voltage Input' (2), as in the default settings. The shunt resistor's value can be set as the factor underneath the signal assignment. The set value here will be the inverse of the resistance of our shunt resistor because we have to divide the voltage signal by the resistance.
Figure 3: Measurement results obtained from grounded components (2 resistors), measured with a 10 Ω shunt resistor with and without applying user compensation. The frequency range is 1 kHz to 1 MHz for all measurements. The red trace shows the measurement result for a 1 kΩ resistor without additional user compensation. The green trace illustrates the measurement result after user compensation with the 1 kΩ resistor. The yellow trace is the result for the 350 Ω resistor measured with the user compensation performed with the 1 kΩ resistor. As we can see, user compensation significantly improves the result at higher frequencies.
Results
The increased complexity of the measurement circuit and the additional cables make the setup more prone to picking up mutual inductances and phase delays of the cables. Additionally, the shunt resistor itself has a small spurious inductivity. The red curve in Figure 3 shows a measurement with a DUT of 1 kΩ and a shunt resistor of 10 Ω without user compensation. As can be seen, above 50 kHz the measurement result shows inductive artifacts. To avoid this, we can take advantage of user compensation (UC) and perform a "Load Compensation" with the exact same setup. The representation in the menu of user compensation is based on the resistance and capacitance. Given that we measure the admittance, the value for the resistance is the inverse of the resistance of our DUT, in our case 1 mS (see Figure 4). The capacitance can be left at the default value. After compensation, the result is flat and shows the right value from 1 kHz to 1 MHz (green curve). Testing the setup and user compensation for robustness with a resistor of a different value (here 350 Ω) shows that user compensation provides us with a reliable admittance up to 100 kHz.
If a larger variety of resistances is expected, we suggest performing a "Load-Load-Load" compensation with three DUTs that cover the whole range of those impedance values. User compensation will then use a polynomial interpolation for values within the range of the compensated values, but it might lead to less accurate results for DUTs outside this range because interpolation is always more reliable than extrapolation.
Figure 4: Screenshot of the menu used to perform user compensation for measurements on grounded components. Beware that the value for 'Load R (Ω)' entered here needs to be the admittance of the measured resistor (for a 1 kΩ resistor, this is 1 mS). The impedance of the grounded component can then be measured correctly, as detailed in the text.
Other Types of Grounded Components
In this blog post, a grounded resistor was measured as the simplest example. Other grounded components with a higher complexity in their impedance can be measured too. For measuring capacitors that are grounded on one side, it is beneficial to not use a shunt resistor but a test capacitor with a well-defined capacity instead. If only the absolute capacitance is required, a fast and easy way to determine it is to measure the ratio of the two voltages and multiply it by the capacitance of the test capacitor. In this case, make sure that the capacitors and the settings are chosen so that 'Aux Input' can resolve well the sine wave at its input.
Conclusion
In this blog post, we showed how to set up the MFIA for measurements on grounded components and presented the results for a resistor. This example may have brought to your attention yet another interesting feature of the MFIA as well as the instrument's versatility. The LabOne modules such as the Plotter, Data Acquisition and Sweeper allow you to acquire and display the measured impedance according to your requirements and application. Using the auxiliary inputs of the MFIA opens up a whole new range of possible measurement configurations. Get in touch to learn more about these possibilities!