unwanted, nonstationary distortions that contaminate bio-signals due to movement of the subject or electrodes during recording
There are various methods to remove motion artifacts. In this project, we expect to use adaptive filtering techniques, providing the skin–electrode impedance signal as a reference to filter out motion artifacts.
In this project, we will design and tape out an application-specific integrated circuit (ASIC) to obtain the in-phase and quadrature (I/Q) components of the skin–electrode impedance signal, stimulated by a body signal, while using the EEG signal as the input. The ASIC consists of four fully differential operational amplifiers for implementing an active RC band-pass filter, along with four fully differential multipliers and an oscillator that generates a 1.2 kHz sinusoidal body stimulus signal.
In the first phase of the project, we will design a fully differential operational amplifier (op-amp).
This op-amp will serve as a fundamental building block that can be reused across all modules of the ASIC.
The design and simulations will be carried out using the Sky130 PDK.
In the second phase, we will design the complete ASIC for impedance measurement.
The chip will integrate:
- Four fully differential op-amps
- Four fully differential multipliers
- An oscillator to generate the stimulus signal
The design and layout will be implemented using the IHP PDK.
| Parameter | Value 1 | Value 2 | Value 3 |
|---|---|---|---|
| Supply Voltage (Design Input) | 1.7 V | 1.8 V | 1.9 V |
| Common Mode Voltage (Design Input) | 0.85 V | 0.9 V | 0.95 V |
| Common Mode Voltage (Design Output) | 0.85 V | 0.9 V | 0.95 V |
| Temperature (Design Input) | 20 °C | – | 50 °C |
| PSRR | 170 dB | 180 dB | 190 dB |
| CMRR | 230 dB | 250 dB | 270 dB |
| Phase Margin | 50° | 60° | 70° |
| Gain Bandwidth Product | 800 kHz | 1 MHz | 1.2 MHz |
| Open Loop (low-freq) DC Gain | 80 dB | 100 dB | 120 dB |
We expect to design a two-stage operational amplifier with a common-source stage as second stage for gain enhancement.
For each stage, a common-mode feedback (CMFB) circuit is used to ensure stability and maintain the desired common-mode voltage throughout the design.
We have employed a beta-multiplier based biasing circuit design with a startup circuit, as shown below:
For the operational amplifier stage, we have employed a folded cascode architecture.
Additionally, a single-ended differential amplifier based common-mode feedback (CMFB) circuit is used to maintain the common-mode voltage at a fixed reference level.
For the common source stage, we have used a resistor-based common-mode feedback (CMFB) topology to stabilize the output common-mode voltage.
For the simulations, we have used the following testbench setup:
The gain bandwidth product (GBW) of our design is shown below.
In this simulation, the design also achieves a 65° phase margin, ensuring stability.
We have achieved a high Common-Mode Rejection Ratio (CMRR) in our circuit, primarily due to its fully differential architecture.
This enhances noise immunity and improves the overall signal integrity.
Below are the test results obtained for the Power Supply Rejection Ratio (PSRR):
| Specification | Min | Typ | Max | Unit | Comments |
|---|---|---|---|---|---|
| Overall Design | |||||
| Supply voltage (Design Input) | 1.7 V | 1.8 V | 1.9 V | ||
| Common mode voltage (Design Input) | 0.85 V | 0.9 V | 0.95 V | ||
| Common mode voltage (Design Output) | 0.85 V | 0.9 V | 0.95 V | ||
| Temperature (Design Input) | 20 °C | 50 °C | |||
| Fully Differential Operational Amplifier | |||||
| PSRR | 170 dB | 180 dB | 190 dB | ||
| CMRR | 230 dB | 250 dB | 270 dB | ||
| Phase margin | 50° | 60° | 70° | ||
| Gain bandwidth product | 800 kHz | 1 MHz | 1.2 MHz | ||
| Open loop (low-freq) DC gain | 80 dB | 100 dB | 120 dB | ||
| Common mode voltage (Design Input) | 0.85 V | 0.9 V | 0.95 V | ||
| Multiplier | |||||
| Linear range | 50 mV | 200 mV | |||
| Common mode voltage (Design Input) | 0.85 V | 0.9 V | 0.95 V | ||
| Oscillator | |||||
| Frequency | 1 kHz | 1.2 kHz | 1.5 kHz |
The fully differential amplifier design is the same as described above.
For the fully differential multiplier design, we have employed a fully differential operational amplifier-based topology, as shown below:
The linearity of the circuit was tested, and the results are shown below:
Above is the linearity testbench for the multiplier
We implemented a multiple-feedback topology 2nd-order Butterworth bandpass filter.
The testbench setup for the filter is shown below:
The testing results for the bandpass filter are shown below:













