When reviewing circuit designs, many approaches can be followed without any approach being obviously superior in all cases. In this course, we will use a relaxed approach to structured reviewing by following these steps:
- General presentation of the circuit design by the designer:
- Target specifications & explanation where needed
- Benchmarking against state of the art: what performance has been achieved in literature and how does that compare to the target specifications?
- Analysis of the target specification: what are the most critical (=important & difficult) parameters in the specification?
- Selection of topology & explanation of method used for this selection
- DC Biasing of the circuit
- Dimensioning of the cicuit components
- Floorplan and modeling of main parasitics of the circuit
- AC analysis of the circuit (noise, gain, phase, stability)
- Transient/Harmonic Balance analysis of the circuit (distortion, blocking, SFDR)
- Circuit simulations to verify DC/AC/Transien analysis
- Layout design
- Verification (DRC/LVS)
- Simulation of extracted layout
- General questions by the reviewers:
- Items that were not clear from the presentation
- Items that were not included in the presentation
- Items of the design method and/or analysis that were questionable
- Specific questions by the reviewers:
- Are the specifications complete, clear & realistic?
- Checking for typical/usual learning opportunities using a structured checklist
- For each of the subcircuits, if present
- For the circuit in its application (PCB, external components etc.)
- For the packaged circuit
- For the complete circuit
- Checking for industrialization:
- Design for fabrication yield (robustness against process variations)
- Design for application robustness (temperature and supply voltage variations, ESD, overvoltage protection, load mismatch, unconditional stability etc.)
- Design for reliability (lifetime effects such as electromigration)
- Design for test (can the circuit be easily debugged, evaluated, and characterized in the lab, and can it be easily and cheaply/quickly tested in production
- Any potential issues with the design:
- Will the design self-destruct?
- Breakdown
- ESD
- Temperature
- E-Migration
- Latch-up
- Will the designed circuit be functional?
- Does the signal path exist?
- Is the biasing correct?
- Will the circuit be stable?
- Will the designed circuit meet specifcations?
- Is it unconditionally stable?
- Does it achieve sufficient immunity against supply variations?
- Will it meet all specifications (gain, noise figure, linearity, bandwidth etc.) also across process variations as well as voltage and temperature changes?
- Will the design self-destruct?
- List of issues to be addressed and questions to be answered by the designers before the next design review (if any)
- Sign-off if all issuess and questions have been adequately addressed