Using Signal Conditioning ICs to Drive Strain Gauge Bridge Sensors

Strain gauges are available with a wide range of zero-strain resistors, and a wide range of sensor materials and related technologies, but a few values ​​(such as 120Ω and 350Ω) are used in a large number of applications. In the past, standard values ​​were easily connected to basic magnetic reflectometers, which contained matching input impedance networks to simplify strain measurements.

Types and Composition of Strain Gages
Metallic strain gauges are produced using a number of alloys, selected to minimize the difference in temperature coefficient between the strain gauge and the strain gauge material. Steel, stainless steel, and aluminum are the primary sensor materials. Beryllium copper, cast iron, and titanium are also available, and “mostly” alloys facilitate the mass, low-cost production of temperature-compatible strain gauges. 350Ω copper-nickel alloy strain gauges are the most commonly used. Thick and thin film strain gauges are used in the automotive industry for their reliability and ease of production, and are generally produced using a ceramic or metal substrate with an insulating material deposited on the surface. The strain gauge material is deposited on the surface of the insulating layer by a vapor deposition process. The sensing gauge and connecting wires are etched into the material using laser vaporization or photomasking and chemical etching techniques. Sometimes a protective insulating layer is added to protect the strain gauge and connecting wires. Strain gauge materials generally include specialized alloys to produce the desired strain gauge impedance, impedance pressure change, and (for temperature stability) the best temperature coefficient match between the sensor and the base metal. Strain gauges and bridge resistors with nominal 3 to 30 kΩ have been developed for this technology for the production of pressure and force sensors.

Bridge Excitation Technology
Strain gauges, thin film and thick film strain gauge sensors are generally constructed using a Wheatstone bridge. The Wheatstone bridge converts the resistance generated by the strain in the strain gauge into a differential voltage as shown in Figure 1. When the excitation voltage is applied to the +Exc and -Exc terminals, a differential voltage proportional to the strain appears at the +VOUT and -VOUT terminals.
In a semi-active Wheatstone bridge circuit (shown in Figure 2), the only two elements of the bridge are strain gauges, which respond to the strain in the material. The output signal of this configuration (typically 1mV/V at full-scale load) is half that of a fully active bridge.

Another fully active bridge circuit uses more than four active 350Ω strain gauges as shown in Figure 3. The characteristic bridge resistance is 350Ω, the output sensitivity is 2mV/V, and the strain gauges use strain materials over a large range.

Figure 1 Strain gauges connected in a Wheatstone bridge configuration

Figure 2 Strain gauges connected in a semi-active Wheatstone bridge configuration

Figure 3. A 16-gauge Wheatstone bridge configuration

Effects of Temperature on Sensor Performance
Temperature causes a shift in the zero-load output voltage (also called offset) and a change in sensitivity (also defined as full-scale output voltage) under load, which can adversely affect sensor performance. Sensor manufacturers introduce temperature-sensitive resistors into the circuit to compensate for the first-order effects of these changes, as shown in Figure 3. As temperature changes, resistors RFSOTC and RFSOTC_SHUNT modulate the bridge excitation voltage. In general, the RFSOTC material has a positive temperature coefficient, and the bridge excitation voltage decreases with increasing temperature. As temperature increases, the sensor output becomes increasingly sensitive to load. Reducing the bridge excitation voltage reduces the sensor output, effectively canceling out the inherent temperature effects. Resistor RSHUNT is insensitive to temperature or strain and is used to adjust the amount of TC compensation produced by RFSOTC. An RSHUNT of 0Ω cancels out all effects of RFSOTC, while an infinite value (open circuit) enables all effects of RFSOTC. This method works very well for compensating for first-order temperature sensitivity, but does not address more complex higher-order nonlinear effects. Temperature compensation for offset changes is accomplished by inserting a temperature-sensitive resistor in one arm of the bridge. These resistors are shown as ROTC_POS and ROTC_NEG in Figure 3. The shunt resistor (ROTC_SHUNT) adjusts the amount of temperature effect introduced by ROTC_POS or ROTC_NEG. Using ROTC_POS or ROTC_NEG depends on whether the offset has a positive or negative temperature coefficient.

How to implement current excitation drive
It is very difficult to use current to excite bridge sensors because the bridge resistance varies with load, and the current in the built-in sensitivity compensation network (RFSOTC and RFSOTC-SHUNT shown in Figure 2) is too large or the current is reversed. Various methods can be used to solve these problems and implement current excitation drive. A simple method is to use MAX1452 and configure it to achieve voltage drive. The circuit includes few external components, which can meet the high current requirements required for voltage excitation. The MAX1452 integrated signal conditioning IC can complete sensor excitation, signal filtering and amplification, temperature linearization of offset and sensitivity, etc. The MAX1452 includes a PRT current excitation circuit as shown in Figure 4. The circuit includes a current mirror (T1 and T2) that amplifies the small reference current by 14 times, which is enough to drive 2-5kΩ PRT sensors. The reference current can be obtained by applying voltage to RISRC and RSTC. This voltage is set by the 16-bit precision full-scale output D/A converter (FSO DAC) in the feedback loop of operational amplifier U1.

[page]The FSO DAC uses a sigma-delta architecture, with digital inputs derived from a temperature coefficient table in flash memory. A unique 16-bit coefficient is provided to the DAC every 4ms for every 1.5°C increment in temperature. The DAC output voltage drives the gate of p-channel MOSFET T1, which in turn drives enough current into RISRC and RSTC to produce a voltage equal to the FSO DAC voltage. The current through T1 is mirrored by T2 by a factor of 14 and becomes the bridge drive current. The resistor RSTC enables first-order modulation of the sensor excitation current, which is a function of temperature. For silicon PRT transducers, the temperature is obtained from the resulting sensor bridge voltage as current passes through the bridge. Such sensors have a very good transfer function between the bridge resistance and temperature. By exciting the bridge with current, you can adjust the resulting bridge voltage and use it to perform first-order compensation for offset and sensitivity. This can be achieved by connecting the bridge voltage (pin BDR) to the reference input of the full-scale output temperature compensation DAC (FSOTC DAC). Keep in mind, however, that current excitation is generally not suitable when using metal or thick film strain gauges.

Figure 4 PRT bridge excitation circuit diagram

Voltage drive circuit
The internal 75kΩ resistors of the MAX1452 can be used as RISRC and RSTC, or external resistors can be connected through switches SW1 and SW2, as shown in Figure 5. The operational amplifier is accessed through the ISRC pin to achieve voltage feedback for bridge driving. Figures 5, 6, and 7 introduce three different voltage drive circuits.

Figure 5: High impedance sensor circuit diagram, no external components are used

Figure 6 Low impedance sensor circuit diagram with npn transistor

Figure 7 Circuit using external RSUPP driver

For high impedance sensors above 2kΩ, the simple circuit in Figure 6 provides voltage drive excitation to the bridge. Opening SW1 and SW2 disables the FSOTC DAC modulation circuit. Connecting pins ISRC and BDR forms an op amp feedback loop to obtain bridge excitation voltage feedback. By sourcing current into the bridge, transistors T1 and T2 (in parallel) increase the bridge voltage to equal the FSO DAC voltage. Low impedance (120Ω to 2kΩ) strain gauges or thick film resistors connected in the Wheatstone bridge circuit cannot be driven directly by T2. This problem can be solved by using an external npn transistor in an emitter follower configuration as shown in Figure 6. The current flowing through the npn transistor comes directly from the collector VDD supply. Driving T1 and T2 enough to conduct turns on the npn transistor, allowing op amp U1 to increase the bridge voltage. To close the loop, the bridge voltage at ISRC is fed back to the op amp. The bridge voltage is regulated to match the FSO DAC output voltage, and a small 0.1μF capacitor is added to the bridge to maintain stability. The base-emitter voltage (VBE) of the npn transistor has a large temperature coefficient, and the effect of this term in the equation is eliminated by feedback from U1. At low temperatures, VBE is large, and the maximum bridge voltage is limited to
VBRIDGEMAX = VDD-VT2SAT-VBE.
Similar to the VBE temperature compensation, the control feedback loop eliminates the temperature component of the TNPN gain in the equation. Another way to provide sufficient drive current for a low-impedance bridge is to connect a small external resistor in parallel with T2 (RSUPP in Figure 7). RSUPP ensures that the bridge voltage is slightly less than the required value (3.0V for VDD = 5.0V). T2 provides more current, raising the bridge voltage to the required value. Since T2 provides the minimum current when it is in the OFF state, RSUPP should be adjusted for the worst-case small bridge voltage. Similarly, the maximum current capability of T2 (2mA at VBDR = 4.0V) determines the maximum bridge voltage modulation that can be used. This circuit can be used for bridge sensors with relatively low temperature coefficients of sensitivity (TCS) that do not require large bridge voltage modulation. U1 feedback eliminates the sensitivity effect caused by the RSUPP temperature coefficient. When designing the circuit, the maximum and minimum RSUPP power derating should be considered to ensure appropriate drive current margin.