Applications Information
Bridge Measurement Conguration without
Instrumentation Amplier
The MAX44267’s low input offset voltage and low noise
make it ideal for biasing strain gauges (Figure 8). The
strain-gauge bridge is most commonly biased from the
reference source and the output from the bridge is then
at approximately half the reference voltage. In the case
of biasing around ground, a negative supply needs to be
available. Hence, the bridge is biased at twice the refer-
ence voltage and the center is at 0V (to within the offset
voltage of the amplifier X1). Doubling the voltage across
the bridge doubles its sensitivity, but of course, the down-
side is twice the current flows.
Since the bridge’s center is now forced to be at 0V, the
node “ip” must also be at 0V when no strain is applied, to
within the calibration of zero strain of the bridge. Having
controlled the zero bias point, the application can use
the second amplifier within the dual MAX44267 to take
a very large, direct-coupled gain without the usual risk of
common-mode errors causing the output to saturate.
A full bridge, as shown, typically produces a differential
output with a full scale of approximately 0.1% of the bias-
ing voltage. Doubling the biasing voltage yields a doubling
of sensitivity while also removing any common-mode
error for the high-gain amplifier. Assuming that RB3 and
RB2 are configured to decrease their resistance as the
strain increases while RB1 and RB4 increase their resis-
tance, then the full-scale output will be +819.2mV at the
output. Capacitor C1 can be sized to roll of any unwanted
bandwidth and its associated noise. Assuming the bridge
uses resistances of 10kΩ in each leg, the noise will be
about 24nV/√Hz, which is then amplified by 100x, along
with the signal to give 2.4µV. If the bandwidth is kept down
to 100Hz then this is only 24µVRMS or about 300µVP-P
yielding a signal to noise ratio of 68dB. This can of course
be improved by averaging the readings over a suitably
long period of time with an integral number of 60Hz
(50Hz) cycles usually offering both improved resolution
and reduced sensitivity to the 60Hz power system.
Layout Guidelines
The MAX44267 features ultra-low offset voltage and
noise, causing the Seebeck effect error to become sig-
nificant. Therefore, to get optimum performance follow the
following layout guidelines:
Avoid temperature gradients at the junction of two dis-
similar metals. The most common dissimilar metals used
on a PCB are solder to component lead and solder to
board trace. Dissimilar metals create a local thermo-
couple. A variation in temperature across the board can
cause an additional offset due to the Seebeck effect at the
solder junctions. To minimize the Seebeck effect, place
the amplifier away from potential heat sources on the
board, if possible. Orient the resistors such that both the
ends are heated equally. It is good practice to match the
input signal path to ensure that the type and number of
thermoelectric junctions remain the same. For example,
consider using dummy 0Ω resistors oriented such that
the thermoelectric sources due to the real resistors in the
signal path is cancelled. It is recommended to flood the
PCB with ground plane. The ground plane ensures that
heat is distributed uniformly reducing the potential offset
voltage degradation due to Seebeck effect.
Figure 8. Bridge Measurement Configuration without Instrumentation Amplifier
R1
18kΩ
½ MAX44267
V
CC
R2
2kΩ
R3
180kΩ
V
REF
4.096V
RB1
RB2
RB3
RB4
ip
X1
X2
MAX6070
V
OUT
C1
1nF
½ MAX44267
V
CC
V
IN
MAX44267 +15V Single-Supply, Dual Op Amp
with ±10V Output Range
www.maximintegrated.com Maxim Integrated
│
16