This design amplifies a 1.3-mV signal between the
frequency band of 10Hz to 10kHz. The output of the charge amplifier signal is
filtered from the resonance frequency of the sensor and amplified for the expected
stable output of ±500mV (1G shock). The common-mode voltage is half of the power
supply, noted as VREF. The design implements a charge sensor amplifier intended to
detect excessive force or shock, such as what occurs when an object is dropped or
exposed to mechanical shock. The output of the detector can be acquired with an ADC
and sampled by a controller to initiate the necessary protective steps.
Use the op amp in a linear
operating region. Linear output swing is usually specified under the
AOL test conditions.
Sensor sensitivity increases with
higher capacitance and in turn increases the op amp voltage noise gain. This
trade-off must be kept in mind when selecting a sensor. The Murata PKGS-00GXP1-R sensor has the following
characteristics: 0.35 pC/G sensor sensitivity, 390-pF capacitance, 31-kHz
An amplifier with low input
voltage noise and low input current noise is preferred to minimize signal to
noise ratio (SNR).
Low input bias current and high
input impedance will minimize offset error and provide a suitable sensor
interface. The output should swing from rail-to-rail to allow simple biasing and
large output voltage swing with a low single supply voltage.
Select the highest value resistor
for Rf to minimize noise impact:
Calculate Cf for low
frequency cutoff (flow) of 10Hz.
Calculate Rin for high
frequency cutoff (fhigh) of 10kHz with sensor
Compute the expected output of
the charge amplifier:
To avoid resonance from the
sensor, which has a resonance frequency of 31kHz, place a Twin-T Notch filter with a stop band at 31kHz. For simplicity,
assume all three capacitors are equal 1nF (C1, C2, C3).
Use a second op amp in a
noninverting configuration to scale Vout to ±500mV. Select
Add a high-pass filter on the
output with a center frequency of 10 Hz, select a low capacitor value, C4 =