PCB Case Studies & Practical Application 2 — Questions and Answers
Question 1: A high-speed DDR4 memory interface is experiencing intermittent data errors. The trace lengths between the controller and DIMMs vary by 800 mils. What is the most likely cause?
- Insufficient bypass capacitors
- Length mismatch causing timing skew violations (Correct answer)
- Incorrect layer stackup impedance
- Missing termination resistors on address lines
Correct answer: Length mismatch causing timing skew violations
DDR4 requires tight length matching (typically within 25 mils for data groups) to prevent setup/hold timing violations from propagation delay differences.
Question 2: A power supply PCB fails thermal inspection. The 5A trace connecting the inductor to output capacitors is 10 mils wide on 1oz copper. What corrective action is most appropriate?
- Add thermal vias near the trace
- Widen the trace to at least 80 mils or add copper pours (Correct answer)
- Increase copper weight to 2oz on all layers
- Route the trace on an internal layer
Correct answer: Widen the trace to at least 80 mils or add copper pours
A 10-mil, 1oz copper trace can only safely carry about 0.5A; a 5A trace requires roughly 80 mils width on 1oz copper using the IPC-2221 standard.
Question 3: During EMC pre-compliance testing, a PCB radiates excessively at 120 MHz. The clock frequency is 60 MHz. Which layout change is most likely to reduce this emission?
- Add ferrite bead on power rail
- Reduce the clock trace loop area by routing return path directly beneath the signal trace (Correct answer)
- Increase board thickness
- Add ground pour on component side only
Correct answer: Reduce the clock trace loop area by routing return path directly beneath the signal trace
The 120 MHz emission is the second harmonic of the 60 MHz clock; minimizing the loop area of the clock signal and its return path reduces the antenna effect causing radiation.
Question 4: A USB 2.0 full-speed device fails eye diagram testing. The differential pair has 3 vias and a 2-inch stub. What is the primary issue?
- Differential impedance mismatch at the connector
- The via stubs create resonances that distort the signal at USB frequencies (Correct answer)
- Common-mode choke is missing
- Trace width is too wide for 90-ohm differential impedance
Correct answer: The via stubs create resonances that distort the signal at USB frequencies
Via stubs act as transmission line stubs that cause resonant notches in the frequency response, degrading signal integrity even at USB 2.0's relatively modest 480 Mbps.
Question 5: A medical device PCB must meet IPC Class 3 standards. The manufacturer reports 4-mil annular rings on 10-mil drill holes. Does this meet Class 3?
- Yes, 4-mil annular ring exceeds Class 3 minimum
- No, Class 3 requires minimum 2-mil annular ring which is satisfied
- No, Class 3 requires minimum 5-mil annular ring for supported holes (Correct answer)
- Yes, but only if the board is double-sided
Correct answer: No, Class 3 requires minimum 5-mil annular ring for supported holes
IPC-6012 Class 3 requires a minimum 5-mil (0.127mm) annular ring for supported holes, so 4-mil rings would be a defect in a Class 3 medical device.
Question 6: An automotive ECU PCB experiences intermittent resets during load dump events. The main power rail has 100µF bulk capacitance. What additional protection should be added?
- Additional bulk capacitors in parallel
- TVS diode rated for automotive load dump (typically 40V clamp) on the power input (Correct answer)
- Ferrite bead on the power trace
- Increase trace width to power connector
Correct answer: TVS diode rated for automotive load dump (typically 40V clamp) on the power input
Automotive load dump transients can reach 60-80V; a TVS diode clamps these spikes before they reach sensitive circuitry, whereas bulk capacitors alone cannot absorb the energy fast enough.
Question 7: A PCB layout review reveals that an analog sensor signal trace runs parallel to a PWM motor control trace for 3 inches on adjacent layers. What problem does this create?
- Increased trace resistance
- Capacitive and inductive coupling causing PWM noise on the sensor signal (Correct answer)
- Impedance discontinuity in the analog trace
- Increased manufacturing cost
Correct answer: Capacitive and inductive coupling causing PWM noise on the sensor signal
Running analog and high-frequency PWM traces in parallel for long distances creates strong capacitive and inductive crosstalk, injecting switching noise into the sensitive analog measurement.
PCB Case Studies & Practical Application 2
A high-speed DDR4 memory interface is experiencing intermittent data errors.
The trace lengths between the controller and DIMMs vary by 800 mils.
What is the most likely cause?