Diode circuits and model consistency
Choose an ideal, constant-drop, or exponential model. An ideal conducting diode has zero voltage and nonnegative anode-to-cathode current; an off diode has zero current and nonpositive anode-to-cathode voltage. Solve the assumed state, then verify its inequalities without mixing models.
Check yourself: Does the computed diode voltage or current support the assumed state?
BJT and MOSFET amplifiers
Verify DC bias and operating region before linearization. Small-signal gain relates increments, not total voltages. For incremental analysis, constant independent voltage sources become shorts and constant independent current sources become opens; dependent sources remain. Frequency-dependent impedances require a stated AC regime.
Check yourself: Could the predicted signal excursion leave the region used for linearization?
Current mirrors and differential amplifiers
Mirror ratios depend on device matching, geometry, and operating region; insufficient output voltage can invalidate them. A differential pair responds to input-voltage differences around its bias point. Common-mode shifts move both inputs together and can change device operating regions.
Check yourself: Are the mirror output and pair inputs within the model's allowed voltage range?
Ideal op-amp circuits and rail limits
Ideal input currents vanish. Equality v+ = v− requires stable negative feedback and unsaturated operation. Solve under that equality, then compare the output with specified rails. During saturation, use the rail voltage and recompute input nodes; equality can fail.
Check yourself: Is the proposed linear output strictly between the stated rails?
Feedback and oscillator conditions
With forward gain A, feedback factor β, and error e = input − β output, linear closed-loop gain is A/(1 + Aβ). Frequency-dependent phase affects stability. For oscillation, the signed round-trip gain must reach unity at a candidate frequency; this alone guarantees neither startup nor stable amplitude.
Check yourself: Did you establish the feedback sign before applying a closed-loop formula?