Biomedical Signal Processing

A: A: Auto-zero helps offset, not bias current. The chopping spikes show up as wideband trash in ECG. Not acceptable at the electrode. B: JFET input keeps bias current down, and DRL knocks down common-mode where mains lives. That's the balance. C: Bipolar bias current will polarize the electrodes and drift the baseline. You see it after a few minutes on the bench. D: Discrete INA looks fine on paper, but matching and CMRR collapse with motion and temperature.

A: A: There is a junction, but the gel dominates the chemistry, not the wire. B: Saline gel dries, chloride distribution shifts, and half-cell potential walks. That's your noise. C: No stim here, and EEG levels won't pit silver. D: PCB oxidation is slow and doesn't explain per-channel impedance drift overnight.

A: A: Nyquist moved. Without more analog attenuation, muscle content folds back. Fix it before the ADC. B: Amplitude isn't the issue; frequency folding is. Gain just makes the mess bigger. C: Resolution doesn't stop aliasing. Wrong lever. D: Baseline wander is unrelated to sampling rate change.

A: A: Those resistors are part of patient protection and fault current limiting. No symbol, no proof. Stop. B: Test points are nice, not a safety blocker. C: AA cap value matters, but it's not a patient hazard. D: Decoupling is assumed and audited elsewhere.