Texas Instruments tidu131 User manual

Weight-Scale and Body Composition Measurement
Reference Design
TI Reference Designs
TI Reference Designs are mixed-signal solutions created by TI’s experts. Verified Designs offer the theory, complete PCB
schematic & layout, bill of materials and measured performance of the overall system.
Circuit Description
This document discusses the analog circuit design of weight-scale and body composition measurement applications. Two
separate signal chains are discussed in this document: one for weight-scale (WS) measurements and the other for body
composition measurement (BCM) analysis.
Design Resources
Design Archive (ZIP File) All design files
AFE4300 Product Folder
1. Design Summary
This design takes a block level approach for designing a Weight Scale and Body Composition system.
1.1 Design Goal
Provide a Weight Scale and Body Composition system reference example.
Foot BCM
AFE4300

2. Theory of Operation
2.1 Background on weight measurements
Weight-scale and body composition measurements are two separate signal chains, so weight-scale application will be discussed
first. While there are multiple techniques for converting weight into an electrical signal, perhaps the most common is to use a
resistive load cell configured as a Wheatstone bridge.
Figure 1 Load Cell Bridge Sensor
Figure 1 shows a bridge configuration where one of the resistors in the bridge changes value as weight is applied. Depending on
the construction of the bridge, more resistors may change value as weight is applied (four resistors at most for a Wheatstone
bridge). Either way, an excitation voltage is applied across the top and bottom of the bridge. The output signal is measured as
the differential voltage across the middle nodes. The challenge in designing a weigh-scale is how to make accurate
measurements of the signal produced by the resistor bridge, because this signal is usually very small (mV magnitude).
Body composition is traditionally obtained by measuring the impedance across several points on the body and matching the
result in a table linking both the impedance measured and the body composition. Bio-impedance is the response of a living
organism to externally applied electric current. It is a measure of the opposition to the flow of that electric current through the
tissues, the opposite of the electrical conductivity.
Figure 2 Cell Bio-impedance model
The bio-impedance model of a cell is described in Figure 2, where Re is the extracellular resistance, Cm is the cell membrane
capacitor, Rm is the cell membrane resistor, Ri is the intracellular resistance. The bio-impedance of cell is decided by various
factors like cell size, shape, density, homogeneity etc.

Figure 3 Bio-impedance measurement model
To measure the bio-impedance, a high-frequency AC current is injected into the tissue through the drive electrodes (refer
Figure 3). The AC current causes a potential voltage difference of the two receive electrodes (refer Figure 3). This potential
voltage difference is related to the resistivity of the tissue between the voltage-sensing or receive electrodes. The equivalent
resistance is defined as the ratio of the voltage difference between the two receiver electrodes and the current that flows
through the tissue.
2.2 Hardware overview and circuit description
2.2.1 Weight-Scale Analog Front-End
Figure 4 Weight - Scale Front-End
2.2.1.1 Overview and Gain Calculation
The weight-scale front-end has two stages of gain, with an offset correction DAC in the second gain stage. The reason for using
INA and PGA is to fit different input voltage from the sensor bridge. Though not shown in the diagram, an anti-aliasing network
is required in front of the INA to filter out electromagnetic interference (EMI) signals or any other anticipated interference
signals. A simple RC network should be sufficient, combined with of the attenuation provided by the on-chip decimation filter.

An internal reference source provides a constant voltage of 1.7V at the VLDO output to drive the external bridge. As the
Wheatstone bridge has high source impedance and the output common-mode voltage is relative high, so the output of the
bridge is connected to an INA (first stage). The good CMRR specification of the INA will reject the common-mode DC voltage at
the bridge output and amplifying the differential signal voltage, the difference in voltage between the two lines only. The high
input impedance and low bias current of the INA will reduce the error caused by measurement circuit to minimum.
The first-stage gain (A1) is set by the external resistor (Rg) and the 100-kΩ (±15%) internal feedback resistors (RFB1):
 󰇛
󰇜
The second-stage gain (A2) is controlled by feedback resistors RFB2, which have four possible values: 80kΩ, 160kΩ, 240kΩ, and
320kΩ. Because the gain is RF / 80kΩ, the gain setting can be 1, 2, 3, or 4.
2.2.1.2 Input Common Mode Range
The usable input common mode range of the weight-scale front-end depends on various parameters, including the maximum
differential input signal, supply voltage, and gain. The output of the first-stage amplifier must be within 250mV of the power
supply rails for linear operation. The allowed common-mode range is determined by the equation:

   
Where:  = maximum differential input signal at the input of the first gain stage;
CM = Common-mode Range.
2.2.1.3 Input Differential Dynamic Range
The max differential (INP–INN) signal depends on the analog supply, reference used in the system. This range is shown in the
equation below:
󰇛󰇜
   
The gain in equation above is the product of the gains of the INA and the second-stage gain. The full-scale input from the bridge
signal typically consists of a differential dc offset from the load cell plus the actual weight signal.
Having a high gain in the first stage helps minimize the effect of the noise addition from the subsequent stages. However, make
sure to choose a gain that does not saturate the first stage with the full-scale signal. Also, the common-mode of the signal must
fall within the range.
2.2.1.4 Offset Correction DAC
One way to increase the dynamic range of the signal chain is by calibrating the inherent offset of the load cell during the initial
calibration cycle. The offset correction is implemented in the second stage with a 6-bit differential DAC, where each output is a
mirror of the other and can source or sink up to 6.5µA. The effect at the output of the second stage is an addition of up to
±6.5µA×2× RFB2. This is equivalent to a voltage at the input of the second stage (A+ / A–) of up to ±6.5µA×2×80kΩ=±1V, when
RFB2 = 80kΩ. Notice that this has no effect in avoiding the first-stage saturation. Because the offset correction DAC is a 6-
bitDAC, the offset compensation step is 2V/= 31.2mV when referred to the input of the second stage.

Typically, a high-resolution analog-to-digital converter (ADC) then follows the PGA to digitize the amplified voltage. Given that
the signals being measured usually are changing slowly, and that very high resolution is required, delta-sigma topologies are
often used to implement the ADC. As with the gain stage, the stability of the ADC overtime and temperature is very important
so as not to limit overall performance.
2.2.2 Body Composition Measurement Analog Front-End
The AFE4300 provides two options for body impedance measurement: ac rectification and I/Q demodulation. Both options
work by injecting a sinusoidal current into the body and measuring the voltage across the body. The portion of the circuit
injecting the current into the body is the same for each of those options. The difference, however, lies in how the measured
voltage across the impedance is processed to obtain the final result.
2.2.2.1 AC rectification
Figure 5 BCM in AC Rectifier Mode
The top portion of Figure 5 represents the current-injection circuit. A direct digital synthesizer (DDS) generates a sinusoidal
digital pattern with a frequency obtained by dividing a 1-MHz clock with a 10-bit counter. The digital pattern drives a 6-bit, 1-
MSPS DAC. The output of the DAC is filtered by a 150-kHz, second-order filter to remove the images, followed by a series
external capacitor to block the dc current and avoid any dc current injection into the body. The output of the filter (after the dc
blocking capacitor) drives a resistor setting the amplitude of the current to be injected in the body:
󰇛󰇜
  󰇛󰇜
Current flows into the body through an output analog multiplexer (MUX) that allows the selection of up to six different contact
points on the body. The same MUX allows the connection of four external impedances for calibration. The current crosses the
body impedance and a second MUX selects the return path (contact) on the body, closing the loop to the output of the
amplifier.

At the same time that the current is injected, a second set of multiplexers connects a differential amplifier across the same
body impedance in order to measure the voltage drop created by the injected current:
󰇛󰇜 󰇛󰇜
Where Z and θare the module and phase of the impedance at , respectively.
The output of the amplifier is routed to a pair of switches that implement the demodulation at the same frequency as the
excitation current source in order to drive the control of those switches. This circuit performs a full-wave rectification of the
differential amplifier output and a low-pass filter at its output, recovers the dc level, and finally routes it to the same 16-bit
digitizer used in the weight-scale chain.

 󰇛󰇜  

Ultimately, the dc output is proportional to the module of the impedance. The proportionality factor can be obtained through
calibration with the four external impedances. Although, with one single frequency or measurement, only the module of the
impedance can be obtained; two different frequencies could be used to obtain both the real and the imaginary parts.
2.2.2.2 I/Q Demodulation
Figure 6 BCM in I/Q Demodulator Mode
As explained previously, the portion of the circuit injecting the current into the body is the same for both configurations.
Therefore, the circuit is the same in Figure 5 and Figure 6. The difference between them is that an I/Q demodulator is used in
this second approach, as shown in Figure 6.
As with the case of the RMS detector, a differential amplifier measures the voltage drop across the impedance:
󰇛󰇜 󰇛󰇜

Where: Z= the module of the impedance at 
θ= phase of the impedance at 
The I/Q demodulator takes the v(t) signal and outputs two dc values. These two values are used to extract the impedance
module and phase with a single frequency measurement. Figure 6 shows the block diagram of the implementation. Using the
I/Q demodulator helps reduce power consumption while yielding excellent performance.
3. Verification and measured performance
3.1 Health Hub Demonstration Suite
The figure below shows up the Health Hub measurement setup hardware description. The app requires a PC, a BlueRadios USB
Serial Dongle along with the Foot BCM data acquisition system. The pictured device connects to the app via Bluetooth low
energy.
Figure 7 Health Hub Measurement Setup Hardware Description
3.2 Health Hub App
3.2.1 Interface Overview
Health Hub is designed to allow the control and display of many BLE enabled health monitoring devices on a single screen. The
screen is divided into multiple discrete areas, Figure 8, called Device Controls and each area allows a specific BLE enabled health
monitoring device control.
3.3 Demonstration usage
Terminology:
Advertising mode: the Bluetooth radio is broadcasting advertising data; this allows another device to initiate a connection to
the advertising device.
Device: a piece of hardware required for a demo.
BlueRadios USB
Serial Dongle
Foot BCM

Device control: The area on the screen of the app that controls a device.
Figure 8 Health Hub PC GUI Interface with highlighted Device Controls
3.3.1 Common Operations
Common operations apply to all of the demos with the Health Hub app.
3.3.1.1 Find Devices
The first step in initiating a connection to a Health Hub demonstration device is to have the PC finding the device. To this end,
the desired device must be in advertising mode. Generally the devices will advertise any time they are turned on and not
connected. When the Foot BCM is in advertising mode the display will be flashing. To find a device press the Discover Devices
button as shown in the figure below. The icon will change to discovering and will find devices for about ten seconds.
Figure 9 PC GUI front panel before and after searching for BLE enabled health monitoring devices
Discover Devices Button
Device Found

3.3.1.2 Connection
The second step is to form a connection. Before finding a device by using Discover Devices the device controls appear as in the
figure below.
Figure 10 BLE enabled device not found
After a device has been found the device control will appear as in the figure below.
Figure 11 Device found –connection can be established
When the Connect button is pressed a selector will be shown as below

Figure 12 Desired Device selection from the list
When the desired device is selected from the list, the app will form a connection to the device over Bluetooth low energy.
After connection the device control will open fully and control over the device can begin.
3.4 Measured Results
The device pictured below is the foot BCM device. The On/Off button is on the underside of the device covered by the device
serial number. A short press of the On/Off button will turn the device on. A longer press of about a second is required to turn
the device off. The On/Off status of the device can be verified by looking at the top of the device in the LCD window. The
backlight will flash periodically while the device is powered ON.
Figure 13 Measurement setup
The first step is the Calibration - press the “Calibrate” button. As in the figure below the button will change to “Scale Ready”
when the calibration is complete. Even though the front panel of the GUI shows up the weight, height and age options it’s not
necessary entering those values as the device is able to measure the weight.