AN-1302: Optimizing the ADuCM350 for 4-Wire, Bioisolated Impedance Measurement Applications

Introduction

The ADuCM350 is an ultra low power, integrated mixed-signal metering solution that includes a microcontroller subsystem for processing, control, and connectivity. The processor subsystem is based on a low power ARM^® Cortex^™-M3 processor, a collection of digital peripherals, embedded SRAM and flash memory, and an analog subsystem that provides clocking, reset, and power management capabilities.

The ADuCM350 has the ability to perform a 2048 point, single frequency, discrete Fourier transform (DFT). It takes the 16-bit analog-to-digital converter (ADC) output as an input and outputs the real and imaginary parts of the complex impedance.

The configurable switch matrix on the ADuCM350 allows the user to choose from a 2-wire, 3-wire, or 4-wire impedance measurement.

This application note details how to set up the ADuCM350 to optimally measure the impedance of a resistor capacitor (RC) type sensor, using 4-wire techniques while targeting IEC 60601 standards.

To target the IEC 60601 standard, the ADuCM350 is used in conjunction with an external instrumentation amplifier, in-amp, (AD8226) to complete high precision absolute measurements using a 4-wire measurement technique.

Configurations for Measuring the Impedence of a Sensor

The ADuCM350 offers three configurations for measuring the impedance of a sensor: 2-wire system, 4-wire system, and 4-wire bioisolated system.

2-Wire System

In the presence of varying access resistance to the unknown impedance, this configuration provides relative accuracy measurements for impedance magnitude and impedance phase. For further details on optimizing the ADuCM350 for 2-wire impedance measurements, refer to the AN-1271 Application Note, Optimizing the ADuCM350 for Impedance Conversion.

The 2-wire system measures the relative accuracy of impedance magnitude and phase.

4-Wire System

This configuration provides absolute accuracy for both impedance magnitude and impedance phase measurements because access resistances are calibrated out. This configuration does not operate where ac coupling capacitors are required to isolate a sensor from the device, that is, capacitors in series with the access resistances. Refer to the AN-1271 Application Note, Optimizing the ADuCM350 for Impedance Conversion, for information on optimizing the ADuCM350 for 4-wire measurements.

The 4-wire system measures the absolute accuracy of impedance magnitude and phase; however, isolation capacitors are not allowed.

4-Wire Bioisolated System

If isolation capacitors are required between a sensor and the device, an external instrumentation amplifier is required to measure the differential voltage across the sensor. It is not possible for the ADuCM350 to do this measurement as a single chip solution because the isolation capacitors cause instability when included on the sense (P channel and N channel) paths.

The 4-wire bioisolated system measures the absolute accuracy of impedance magnitude in the presence of isolation capacitors; however, this system is not targeted for accurate phase measurements.

Basic 4-Wire Impedance Measurement

To measure the impedance of an unknown sensor, Z, the following ratiometric measurement technique is employed using the ADuCM350:

1. Measure the impedance of a known precision resistor, R_CAL, as shown in Figure 2. An excitation voltage is applied at R_CAL 1 with associated D and P switches closed in the switch matrix. The resultant excitation current is measured through R_CAL 2, with associated T and N switches closed in the switch matrix. This current is converted to a voltage using the transimpedance amplifier (TIA), where the TIA resistor (RTIA) is optimized for the maximum current seen by the ADC and is converted to a voltage using the ADC. A 2048 point Hann sample is performed on the data to give real and imaginary components of the impedance.



2. Change the switch matrix configuration, as shown in Figure 3, and excite the sensor measuring the response current. The DFT engine now calculates the real and imaginary components of the unknown impedance, Z.



3. Calculate the unknown impedance magnitude on the core using the following equation:



4. Calculate the unknown impedance phase on the core using the following equation:

This 4-wire measurement approach to measuring impedances operates if there are no isolation requirements on the sensor. However, if an isolation capacitor, such as CISO, must be included in series with the access resistor, RACCESS, in a 4-wire measurement, then a single chip solution is not possible.

The 4-Wire Bioisolated Method

Basic 4-Wire Theory

In a classic 4-wire/4-terminal sensing system, a differential current source is used to force a known current into the sensor. This forced current generates a potential difference across the Unknown Z, which is measured according to Ohm’s law where:

V = I × R

When the current is forced, the access wires to Z also lead to a drop in voltage, which causes inaccuracies in measurement. To remove this loss from the actual measurement of Z, a differential pair of sense lines are connected to Z at Point A and Point B as shown in Figure 2.

The differential sense lines are designed with high input impedance stages so that no current flows through the lines and there is no voltage drop across the lines. The impedance, Z, is then measured using the following equation:

Z = V_METER/I_AC

where:

V_METER is the voltage across Z or the sensor

I_AC is the ac current through Z or the sensor

4-Wire Bioisolated Theory In Application

An alternative approach is to use a high precision excitation voltage source as the force signal. Apply this voltage to Z and measure the response current using a high accuracy current meter (see Figure 3). The unknown impedance, Z, is then measured by the following equation:

Z = V_METER/I_METER

where:

I_METER is the current through the sensor.

Referring back to Figure 3, it is possible to measure 4-wire impedance using the ADuCM350. The excitation stage excites the sensor with a known voltage, which is accurately differentially sensed using the internal instrumentation loop. The current response is measured through the TIA channel and converted to a voltage.

In a real-world application, such as those governed by the IEC 60601 standards, the Z (or sensor) allows a limited dc voltage across Z or the sensor. The restrictions on the ac current forced on the sensor are more relaxed. The ac voltage source is selected for the force connection to the sensor to utilize the ADuCM350 DFT capability.

In Figure 5, C_ISO1 and C_ISO2 are discrete isolation capacitors that ensure that no dc voltage appears across the sensor. RA_CCESS1 and R_ACCESS2 are access or lead resistance inherent in the connections to the sensor. R_LIMIT is an extra level of security to guarantee the maximum allowable excitation current seen by the sensor in a scenario where the R_ACCESS resistance is removed from the measurement.

A 4-Wire Bioisolated Solution

Referring to Figure 5, the following is required:

• A precision ac voltage source
• A high precision current meter
• A precision differential voltage meter

Precision AC Voltage Source

The ADuCM350 has a high precision excitation control loop, which drives a precision ac voltage to the sensor. An internal differential sense configuration guarantees the accuracy of the voltage source (see Figure 6). The positive sense, P, is tied to the drive terminal, D, in the configurable switch matrix. A direct digital synthesizer (DDS)-based sine wave generator is used to generate the ac stimulus through a 12-bit DAC. For more information regarding the transmit stage, refer to the ADuCM350 Hardware Reference Manual.

High Precision Current Meter

The ADuCM350 utilizes a TIA amplifier for current to voltage conversion for measurement by the high precision ADC, the gain of which is set by an external resistor, R_TIA. The TIA channel sinks the sensor excitation current, and the channel is precisely biased on a common-mode voltage of 1.1 V. Significant analog and digital filtering is performed on measurement for rejection of interferers and noise. The T and N channels are tied together using the switch matrix for accurate sense capability on the current measured (see Figure 7).

The ADC converts the current measurement with a 160 kSPS ADC. A 2048 sample point DFT is performed on the data; resulting real and imaginary components for the current measurement are calculated.

Precision Differential Voltage Meter

To differentially sense the voltage across the sensor, a low power in-amp with excellent noise and common-mode rejection is required (see Figure 8). The AD8226 is selected for this application. The AD8226 is referenced off the common mode of the system set by the V_BIAS voltage on the TIA channel. The output of the in-amp is fed back into the ADuCM350 through one of the auxiliary channels, for example, AN_A.

The ADC converts the auxiliary voltage measurement with a 160 kSPS ADC. A 2048 sample point DFT is performed on the data and resulting real and imaginary components for the voltage measurement are calculated.

4-Wire, Bioisolated Measurement System Block Diagram

Figure 9 shows the combination of the following:

• A precision ac voltage source (ADuCM350 excitation stage)
• A high precision current meter (ADuCM350 TIA channel stage)
• A precision differential voltage meter (AD8226 instrumentation amplifier)

How to Calculate the Unknown Z

After obtaining the current and voltage DFT measurements, the device can exit the AFE sequencer and calculate the impedance of the sensor using the following equations:

where r and i are the real and imaginary components from the voltage and current DFT measurements, respectively.

To calculate the Impedance Z, use Ohm’s law by dividing the voltage magnitude by the current magnitude while taking into account the gains of the signal chain as follows:

The current measurement value is converted to a voltage, using the R_TIA, for measurement purposes. This gain must be taken into account.

The 1.5 gain in the equation is the ratio between the gain of the ADuCM350 current measurement channel, which is 1.5, vs. the gain of the ADuCM350 voltage measurement channel, which is 1. The gain of the in-amp is determined by the selection of RG. For the AD8226, this is determined by

R_G = (49.4 kΩ)/(G − 1)

By choosing RG = 100 kΩ, the gain is 1.494.

Note that these equations are taken into account in the example provided in the ADuCM350 Software Development Kit.

Example of a 4-Wire Bioisolated System

Sensor Configuration

In the example described in this application note, measure the impedance of an RC type sensor, with the configuration shown in Figure 10, for a 30 kHz excitation signal. Note that TOL indicates tolerance.

The sensor details are as follows:

• C_S = 220 pF
• R_S = 20 kΩ
• R_P = 100 kΩ

The total impedance of the sensor must be calculated to verify the system accuracy.

1. Calculate the complex sum of

R_S + C_S = Z_S =
34962 ∠ −55.11

2. Calculate Z_S || with

R_P = Z_T = 28337.15 ∠ −41.66

which is the total impedance of the RC sensor to be measured.

4-Wire Bioisolated Network

For this 4-wire example, select the following components:

• A lead access resistor, R_ACCESSx = 4.99 kΩ
• An isolation capacitor, C_ISOx, of 47 nF

If Z is close to or less then R_ACCESSx, a potential divider effect occurs that limits the bandwidth of the ADuCM350 thus degrading accuracy (see Figure 11).

AFE Optimization

Optimizing the ADuCM350 consists of the following steps:

• Calculate the R_LIMIT resistor.
• Calculate R_TIA.
• Calculate the R_G of the AD8226.
• Calculate R_CAL.

Calculate the R_LIMIT Resistor

When calculating the R_LIMIT resistor, note that the maximum output voltage from the ADuCM350 = 600 mV peak.

The maximum allowed ac current at 30 kHz is the following:

300 μA rms (Targeting IEC 60601) = 424 μA peak

Being conservative, set the maximum allowable ac current to 200 μA peak (<50%).

R_LIMIT ~= 600 mV peak/200 μA peak = 3 kΩ.

This calculation ignores CISOx due to its small size.

Calculate R_TIA

R_TIA is the feedback resistor on the TIA to convert the current to a voltage.

Minimum impedance/maximum current seen by the TIA is

Assume 20 kΩ is the minimum impedance of Z_UNKNOWN.

Note the following:

• The maximum voltage swing is 600 mV peak.
• The highest signal current into TIA = 600 mV peak/ 32.98 kΩ = 18.19 μA peak.
• The peak voltage at output of TIA (maximum allowed by the ADuCM350) = 750 mV peak.
• R_TIA resistor to give peak 750 mV voltage for peak signal current is R_TIA = 750 mV/18.19 μA and RTIA = 41.2 kΩ.
• To prevent overranging of the ADC, add a safety factor of 1.2, that is, the minimum impedance is 1.2 times less than the specified minimum impedance of 32.985 kΩ = 27 kΩ.
• The R_TIA with a safety factor included = 41.2 kΩ/1.2 and R_TIA = 34.3 kΩ.

Note that 33 kΩ is used for this example.

Calculate R_G of the AD8226

The maximum impedance of the sensor is the following:

Z_UNKNOWNMAX = 28.337 kΩ

A safety factor is incorporated on the R_TIA to prevent the ADC from overranging. The same must be done here thus the maximum peak current is divided across differential inputs of AD8226 by a factor of 1.2.

Peak Current Seen at V_IN (AD8226) =

(18.19 μA peak)/1.2 = 15.16 μA peak

V_IN (AD8226) =

15.16 μA peak × 28.337 kΩ = 439.6 mV peak

AD8226 G = 750 mV peak/(439.6 mV peak) = 1.706

If a further safety factor of 1.1 is used on the peak-to-peak of the voltage (this may be unnecessary for the application), then

AD8226 G = 750 mV peak/(1.1 × 439.6 mV peak) = 1.55

AD8226 G = 1 + (49.4 kΩ/RG) = 1.55

R_G = (49.4 kΩ)/(1.55 − 1) = 89.8 kΩ

Select an R_G of 100 kΩ because it is a standard value.

AD8226G = 1 + (49.4 kΩ/R_G) =

1 + (49.4 kΩ/100 kΩ) = 1.494

Note that the AD8226 has bandwidth limitations. For a frequency of 50 kHz, the gain is limited to 10 (see Figure 12).

Calculate R_CAL

The calibration of the auxiliary channel and the TIA channel must take into account the gain through the system.

For the voltage measurement channel, the auxiliary channel is calibrated.

For the current measurement channel, the temperature sensor is calibrated, and the results are loaded to the offset and gain registers of the TIA channel. This calibration and loading ensures that the difference between the voltage and current gain is exactly 1.5.

The calibration of the auxiliary channel and the TIA channel is done for the user in the 4-wire bioisolated example code in the ADuCM350 Software Development Kit.

4-Wire Bioisolated Measurements

Hardware Setup For 4-Wire Bioconfiguraton Board

When setting up the EVAL-ADuCM350EBZ motherboard, do the following:

• For the voltage measurement, insert LK1 (Auxiliary Channel A).
• Open LK6.

For the ADuCM350 4-wire bioconfiguration board, do the following:

• Insert LK7, LK8, LK9, and LK10.
• Insert LK16, LK17, LK18, and LK19 in Position A to measure an external sensor. In this example, the network shown in Figure 10 and Figure 11 is measured. The result appears as shown in Figure 13.

Firmware Example

Code available in the ADuCM350 Software Development Kit is designed to be used with the 4-wire bioconfiguration board to validate the solution discussed in this application note

The Readme.txt in the example folder provides more details on the measurement.

After downloading the ADuCM350 Software Development Kit, go to C:\Analog Devices\ADuCM350BBCZ\EVALADUCM350EBZ\examples.

1. Click the BioImpedanceMeasurement_4Wire folder.
2. Click the .eww file in IAR.
3. During the download and debug stage, open the Terminal I/O window to read the returned results.

Measurement Results

Impedance Magnitude

Measured Result = 28405 Ω

Theoretical Value Measured ZT = 28337 Ω

However, the CS of 220 pF used in the calculation had a tolerance of 1%

Upon analysis, the capacitor measured closer to 221 pF.

In theory, a C_S of 221 pF gives a Z_T of 28416 Ω vs. the measured result of 28405 Ω.

For more details, refer to the Sensor Configuration section.

Impedance Phase

The current 4-wire bioisolated configuration is not capable of measuring accurate phase measurements.

If an absolute phase measurement is required, use a single chip ADuCM350 4-wire measurement configuration. Note that this configuration does not have isolation capacitors (C_ISOx).

Schematics for the 4-Wire Bioconfiguration (Bio3z) Board

Limitations on Use and Liability

The application described in this application note is specific to the ADuCM350 for use with the EVAL-ADuCM350EBZ evaluation board. In addition to the terms of use contained in the evaluation board user guides, it is understood and agreed to that the evaluation board or design must not be used for diagnostic purposes and must not be connected to a human being or animal. This evaluation board is provided for evaluation and development purposes only. It is not intended for use or as part of an end product. Any use of the evaluation board or design in such applications is at your own risk and you shall fully indemnify Analog Devices, Inc., its subsidiaries, employees, directors, officers, servants and agents for all liability and expenses arising from such unauthorized usage. You are solely responsible for compliance with all legal and regulatory requirements connected to such use.