Diagnosing D/A converter output errors from scope display Given a D/A output that, under a code sweep, exhibits steps that occasionally reverse direction (drop instead of rise), what fault best describes this behavior?

Introduction / Context:
A D/A converter should be monotonic: as the digital code increases, the analog output must never decrease. Nonmonotonic behavior breaks control loops and causes audible/visible artifacts in many systems.

Given Data / Assumptions:

• Digital input code is swept upward sequentially.
• Oscilloscope shows occasional downward steps in the analog output.
• Connections and stimulus assumed correct unless otherwise noted.

Concept / Approach:
Nonmonotonic output error occurs when differential nonlinearity (DNL) is worse than −1 LSB at some transition, causing an actual decrease in output for an increasing code. This is distinct from missing codes (DNL > +1 LSB) and gain/offset errors (linear deviations).

Step-by-Step Solution:
Observe stair steps as code increases.Identify any downward (negative) steps.Downward steps imply nonmonotonicity → select the corresponding diagnosis.

Verification / Alternative check:
Datasheets specify guaranteed monotonicity up to a certain number of bits or over temperature; violation matches the observed reversal of steps.

Why Other Options Are Wrong:

• Noisy input: Noise causes jitter, not consistent negative steps at specific codes.
• Scope nonlinearity: Would distort all levels, not selectively invert steps.
• Incorrect coding: Would reorder levels or produce missing segments, but the hallmark of nonmonotonicity is a local decrease with increasing code.

Common Pitfalls:

• Blaming cables or probes; while they can attenuate or ring, they do not cause code-order reversals.

Final Answer:
The converter has a nonmonotonic output error.