There are many considerations to take into account when planning for a low level capacitance measurement. The following information addresses the most common considerations.
One option for the capacitance measurement device is the NI PXI-4072 (NI 4072). The NI 4072 is a 6½ digit digital multimeter (DMM) that has a built-in LCR meter to measure inductance and capacitance. It has capacitance ranges from 300 pF to 10,000 uF and can measure with a resolution as low as 0.05 pF. For more information on how the NI 4072 performs capacitance and inductance measurements, refer to Capacitance/Inductance Measurements at ni.com/zone.
After selecting a capacitance measurement device, decide how to route the test signals to this device. A cost-effective and standardized approach is to incorporate switching on the front end of your system.
Nearly every ATS application uses some type of switching. Switching allows for customizable configurations, fast connection speeds, and the ability to programmatically connect a high number of channels to a minimal number of test instruments, thereby improving system performance while reducing the cost of an automated test system. For example, with high-density NI PXI and SCXI switch modules, you can automate the connection of one NI 4072 to hundreds of DUTs. The connection speed can be altered to maximize throughput or accuracy, and the configuration can be changed using software to run subsequent tests.
National Instruments provides a variety of flexible, cost-effective switch modules. To see our complete switch offering, visit ni.com/switches.
A switch topology is an organizational representation of the channels and relays in a switch module. One or more switch modules configured in a variety of topologies is necessary to provide your system’s required connection points and functionality. For additional information on available topologies and their respective configurations, refer to Understanding Switch Topologies at ni.com/zone or the NI Switches Help.
There are three common topologies used for switching capacitance measurements: MUX, Single-Pole Double-Throw (SPDT), and Matrix.
The MUX topology is often selected for multiple channel measurement applications, where numerous channels connect one at a time to a measurement device. This topology can be used for larger capacitance measurements. However, with mid- and low-level capacitance measurements, parasitic offsets can significantly affect results, and the MUX topology may no longer be the best option.
SPDT and Matrix topologies provide system configurations that assist in limiting parasitic offsets when performing mid to low-level capacitance measurements. These topologies and their useful configurations are discussed in the Advanced Considerations section of this tutorial.
In most ATS applications the DMM will be connected to the DUT through cables, switches, terminal blocks, and fixtures. These additional components add parasitic capacitance which can cause undesirable error in the measurement. Figure 1 shows a model for the other residuals that may exist in your system. For capacitance measurements, we are mostly concerned with the parasitic capacitance, Cp, but the other residuals may need to be considered, depending on your test system. Cp can easily reach the uF range in a large switching system if certain precautions are not taken into account.
Figure 1: Model of the Test Fixture Residuals
Compensation can be performed to reduce the error between the DMM and the DUT. Two commonly used methods of compensation are offset nulling and open/short compensation.
Offset nulling, or Open compensation, is useful if you assume that Rs and Ls are negligible. Offset nulling involves taking two measurements and subtracting them to remove the effect of parallel capacitances (Cp in Figure 1) that exist in your test system. An initial “offset” measurement is taken with the DUT disconnected (open) and a second measurement is done with the DUT connected. Subtracting the initial offset from your DUT measurement produces the more accurate result.
Open/Short compensation is useful if you assume that your system error is a combination of many different test fixture residuals. Open/short compensation incorporates offset nulling, but also measures when the connections next to the DUT are shorted. This short circuit allows Rs and Ls to be measured. For more information on how to perform open and short compensation please refer to Capacitance/Inductance Measurements at ni.com/zone or the NI-DMM Help File.
Note: The NI-DMM driver that is used with the NI 4072 can automatically compensate for all subsequent measurements after taking the open and short measurements.
Noise and Averaging
When taking low capacitance measurements, noise and other environmental factors can result in measurement error. Therefore, every effort should be made to remove or protect your system from noise sources. The NI 4072 has a preset aperture time that it uses during capacitance measurements. If a greater effective resolution is desired, multiple samples can be averaged by modifying the “Number of LC Measurements to Average” property. By averaging more samples together in the driver, you can reduce the effect of noise on your signal.
Use Relays with Low Minimum Current and Low Path Resistance Specifications
The NI 4072 DMM sources a small test current to measure capacitance. Switch modules that have low minimum current specifications should be used to maintain consistent signal integrity.
Switch modules with low path resistance specifications should be used to keep the impedance between the DMM and DUTs as low as possible. However, keep in mind that higher path resistance switches can be used as long as a short compensation is performed.
Consider Terminal Block Connectivity
Verify that the switches you choose can be connected cleanly to your test fixture. NI switches have several connectivity options, including front-mounting terminal blocks and cables. NI also provides mating connector part numbers so customers and third-party companies can build their own interconnects.
Isolation and cabling
Arrange switch channels in a manner that reduces the amount of capacitance that needs to be compensated by the DMM. Positioning switch channels in parallel is a simple and often-used configuration, but it results in the summation of parasitic capacitances and a higher probability of measurement error. Instead of placing switch channels in parallel, another configuration option is to arrange the switches in a tree structure to prevent the parasitic capacitances from adding up. This is discussed further in the Advanced Considerations section of this tutorial.
Isolating the switch channels into different sections also reduces the amount of capacitance that needs to be compensated by the DMM. Since the total capacitance is lower, it could allow you to use a lower range on the DMM, which will increase the resolution of the measurement. When using this technique, it is important to shield different sections of channels from one another.
Shielded cables can be used for all channels to reduce the effect of noise and channel-to-channel capacitance. However, shielded cables have some capacitance between the conductor and the shielding which would increase the amount of channel-to-ground capacitance. This increase in channel-to-ground capacitance creates a larger offset capacitance for the DMM. This trade-off between noise and offset capacitance should be taken into account before using shielded cables for every channel. Shielded cables are also typically larger, which can create a bulkier test fixture.
When cables are stressed, moved around, or rubbed against one another, piezoelectric and triboelectric effects can create potentials that can lead to erroneous readings. Make sure cables are securely positioned using tie-wraps so that they do not move around.
Note: If individual cable diameter is an issue, micro coax cables may be an option for interfacing between the instruments and the test fixture, as it would still provide shielding.