Accurate Impedance Measurements Despite High Parasitic Resistance

July 13, 2023 by Tim Ashworth

When measuring the impedance of a component or material, it is good practice to measure with cables of lowest possible impedance to ensure an accurate and precise measurement. However, when it is not possible to use low resistance cables or the measurement loop requires a high-resistance in series, how can we measure accurately? This blog post takes a look at an impedance measurement where an in-series resistor is used to simulate an unavoidable resistance in the measurement fixture, cables or sample stage. It shows that thanks to the high dynamic reserve of the MFIA (and MFLI with MF-IA option), coupled with the user compensation feature, accurate impedance measurements can be taken even in the presence of a high in-series resistance such as 1 MΩ.

Photo of the MFIA with additional parasitic resistance

Figure 1: MFIA Impedance Analyzer with two MFITF Fixtures; On the left, the DUT is inserted and on the right an in-series resistor is inserted to simulate a high-impedance measurement path

To simulate a high-impedance measurement path, we connect two mid-frequency impedance test fixtures (MFITF) to the MFIA as shown in figure 1. In the MFITF on the left, we insert the device under test (DUT), while the MFITF on the right has a 1 MΩ component (1206 SMD part number P1206Y1004BNT) inserted to simulate a high-impedance pathway.

Reference Measurement Without the In-series Resistor

Let’s first measure the DUT without any additional in-series resistor to give us an idea of the accuracy we can expect.

Figure 2: Screenshot of the LabOne Sweeper Module showing the reference measurement of the DUT (1kOhm resistor nominal, 999.93 Ohm DC value). No in-series resistor is used in this sweep from 1 kHz to 1 MHz. The measured value of the absolute impedance is 999.59 Ohm, which matches the expected DC resistance value to within 0.034%

Figure 2 shows a screenshot of the LabOne Sweeper Module showing the absolute impedance of the DUT (nominal 1kΩ SMD 0805 part number Y1624-1KCT-ND) to be 999.59 Ω when averaged from 1 kHz to 1 MHz, which matches the expected DC resistance of 999.93 Ω to within 0.034%. To quantify the precision of the measurement we measure the standard deviation with the Sweeper Module toolset, averaged over this 400-point measurement we find it to be just 20 mΩ for this measurement taken in application preset mode “high”.

Insert the In-series Resistor and Compensate

We now insert the in-series resistor of 1 MΩ to act as an unavoidable resistance in the measurement path. Before we retake the measurement, we carry out a “user compensation”. This simple step requires a well-known resistance and a short, giving us a short-load compensation. Please note, it is necessary to disable the “validation” setting in the user compensation advanced settings when the measurement fixture deviates from the typical short and load values. In this case, we use the same load as the DUT, as it is well defined and has low parasitic impedance. For the short, we use the short carrier card from the MFITF kit.

We run the compensation from 1 kHz to 1 MHz, and when complete this is automatically applied to the measurement, but can be toggled on and off and can also be saved and recalled for later measurements.

Once the compensation is applied, we retake the same measurement of the DUT with the same measurement parameters to allow for a fair comparison. Figure 3 shows the measurement taken of the same DUT with the 1 MΩ in-series resistance. The measurement values of 999.52 Ω matches the expected DC resistance of 999.93 Ω to within 0.041%. This shows that the MFIA can achieve excellent accuracy even when the measurement path contains an in-series resistance of 1 MΩ. However, we note a significantly higher standard deviation of 166 mΩ (compared to 20 mΩ without the in-series resistor). This corresponds to reduced measurement precision, and shows the trade-off required when measuring with an in-series resistor. The standard deviation of 166 mΩ still represents a high-precision measurement, and further averaging would allow for the precision to be improved.

Figure 3: This LabOne Sweeper Module measurement shows the same 1 kΩ (nominal) DUT measured from 1 kHz to 1 MHz with the same measurement parameters as in figure 2 to allow a fair comparison. The measurement values of 999.52 Ω matches the expected DC resistance of 999.93 Ω to within 0.041%.

Confirm the Results on Other DUTs

The results on the above 1kΩ DUT show high accuracy with decreased but still acceptable precision. We now turn to a DUT with even smaller resistance to confirm the performance. The 100 Ω DUT nominal is measured without the in-series resistor and has an absolute impedance of 99.99 Ω with 2 mΩ standard deviation when measured from 1 kHz to 1 MHz as shown in figure 4.

Figure 4: Screenshot of the LabOne Sweeper Module showing the reference measurement of the 100 Ω DUT with no in-series resistor used in this sweep from 1 kHz to 1 MHz. The measured value of the absolute impedance is 99.99 Ω, which matches the nominal value of this component.

Now measuring with the same setup as above with the 1 MΩ in-series resistance, we measure the absolute impedance to be 100.00 Ohm as shown in figure 5, but now the measurement has a standard deviation of 79 mΩ. Once again confirming the accuracy but showing the trade-off being the decreased precision.

Figure 5: This LabOne Sweeper Module measurement shows the same 100 Ω DUT with 1 MΩ in-series resistor measured from 1 kHz to 1 MHz with the same measurement parameters as in figure 4 to allow a fair comparison. The measurement values of 100.00 Ω matches the measured value without the in-series resistance.

To confirm the measurement on a reactive DUT, we measure a capacitor of value 99.0 pF. The resulting measurement of this DUT can be seen in figure 6, which shows the capacitance to be 98.9 pF, so within 0.1% of the expected value.

Figure 6: Screenshot of LabOne Sweeper module showing a swept measurement of a capacitor to be 98.8 pF, agreeing within 0.1% with the expected value of 99.0 pF

Conclusion

Measuring with high in-series parasitic resistance is not only possible with the MFIA, but the accuracy of the measurement is also maintained at the cost of a relatively small decrease in precision. This is thanks to the high dynamic reserve of the MFIA and its user compensation feature. The user compensation feature makes it easy to compensate for your fixture, cables and stage, even when they have a high intrinsic resistance. If you have questions about fixture compensation, please get in touch.