After prototyping the same circuit for an ICP/ IEPE sensor and testing it in the lab, I would like to emphasize that this circuit could potentially be terrible in DC blocking.
What happens if the load of this circuit is for example an opamp -which is quite often the case- implementing an anti-aliasing filter, etc?
Without determining a resistor from the decoupling capacitor to the Ground, two problems may arise:
- It is possible that no return path for the DC current exists.
This is explained in the following article, "Missing DC Bias Current Return Path When AC-Coupled": https://www.analog.com/en/analog-dialogue/articles/common-problems-when-designing-amplifier-circuits.html
"What actually happens is that the input bias currents will flow through the coupling capacitor, charging it, until the common-mode voltage rating of the amplifier’s input circuit is exceeded or the output is driven into limits. Depending on the polarity of the input bias current, the capacitor will charge up toward the positive supply voltage or down toward the negative supply. The bias voltage is amplified by the closed-loop dc gain of the amplifier. This process can take a long time, thus, a casual lab test (using an ac-coupled scope) might not detect this problem, and the circuit will not fail until hours later."
A malfunctional ac-coupled op-amp circuit.
In my case, I observed that the DC part of the signal wasn't blocked at all. However, when I used an oscilloscope, after the AC-coupling-capacitor, the DC part of the signal started slowly to decrease, which brings us to point 2:
- If there is a path for the DC current, the impedance of the load should obviously be taken into account, since together with the AC-coupling-capacitor they form a high-pass filter. When dealing with filters, most of us usually focus on the frequency domain. However in this case, the response of the filter in the time domain should also be carefully examined. In the scenario described above, the 20uF AC-coupling-capacitor formed a high-pass filter with the internal resistance of the oscilloscope that I used, which was 1M Ohm. As a simulation for the sensor, I used a sine wave with DC offset. The result is that the DC part of the signal was eventually blocked -only when the probe was attached to the circuit- but that happened after about 2 minutes, which could possibly be more than enough to fry the opamps. Using an online, high-pass filter calculator, the step response of the filter can be observed: (R=1MΩ C=20uF). This result is in line with the behaviour of the circuit that I empirically observed in the lab.
Concluding, it seems wise to determine and place a resistor and deliberately form the high-pass filter. Care should be taken when choosing the value of the components, in both the frequency and the time domain. Avoid too high values for the resistance, due to the noise it introduces in the signal (resistor noise model).