Technical Paper Model-Answer Audit and Completion Record¶

Audit date: 2026-07-26
Completion date: 2026-08-02 Current scope: all 55 notes under 03_technical/, the official NTC syllabus, and the available question transcriptions in technical-questions-c.md. Goal: maintain a bank of answers that can be reproduced in a 3-hour Paper II examination.

[!IMPORTANT] The rewrite is complete for the current question maps. Every mapped question now links to one mark-bounded, same-chapter model answer with an explicit practice target. The historical findings below explain why the rewrite was needed; they are not the current readiness state.

Completion Status¶

The answer contract is enforced by make audit (scripts/audit_notes.py). It checks question/link/answer ownership (a question may link to one or more same-file answers), unique IDs, mark consistency for single-link questions, multipart split metadata, explicit anchors, practice targets, and a figure or table for 10-mark answers.

1. Historical Audit Baseline¶

Readiness verdict¶

The immediate problem is not lack of words. It is the absence of bounded, examiner-facing answers and the fragmentation of genuine multi-part questions.

2. The Paper II Writing Contract¶

The official format is 4 questions x 5 marks and 8 questions x 10 marks in 3 hours. The preparation budget below is a practical target, not an official marking scheme.

What to write for 5 marks¶

Use the shortest complete scoring chain:

1. Definition or principle in 2–3 precise lines.
2. One scoring visual: labeled diagram, waveform, truth table, or two-column comparison.
3. Four to six technical points or ordered working steps.
4. The governing equation, with symbols defined, when applicable.
5. One closing line stating application, significance, or limitation.

A 5-mark answer must be complete by itself. It must not end halfway through a derivation or say “details as above.”

What to add for 10 marks¶

Retain the 5-mark core, then add only material that earns distinct marks:

1. Detailed construction, block-by-block operation, or full derivation.
2. A second diagram, waveform, characteristic, timing chart, or design example.
3. A comparison table, calculation, performance analysis, or assumptions.
4. Advantages, limitations, applications, and an interpreted conclusion.

Do not double the word count with repeated definitions. A 10-mark answer should add new scoring surfaces.

Multi-part 10-mark questions¶

Preserve the original split and divide time and page space proportionally. For [2+3+5=10], use roughly 20%, 30%, and 50% of the answer. A 2- or 3-mark limb is not a standalone canonical PYQ; it belongs to the original 10-mark question.

3. Canonical Question Rules¶

1. Every bank entry must be a definite 5 marks or 10 marks. Remove [5–10], [3–5], and similar ranges.
2. Preserve exact observed wording and internal splits for genuine PYQs.
3. A fragment may retain a provenance note, but it must not be presented as an independently observed 5-mark PYQ.
4. A predicted standalone variant is allowed only with a clear [likely] label.
5. Each canonical question links to one dedicated model-answer heading, not a collection of general chapter sections.
6. One chapter owns each answer family. Other chapters cross-link rather than duplicate it.
7. Each model answer visibly separates For 5 marks and Add for 10 marks when the topic reasonably scales.
8. Historical systems such as X.25, Frame Relay, PDH, ATM, and SS5 remain because they are in the syllabus, but must be labeled as legacy where appropriate.

4. Domain Readiness¶

Electronic Devices and Digital Logic¶

Electromagnetics and Control¶

Signals and Communication¶

Telecom, Optical, and Wireless¶

Basic Electricity and ICT¶

5. Exact NTC PYQs to Rebuild First¶

NTC 2079¶

• MOSFET V-I characteristics [5]
• PCM transmitter and receiver [5+5=10]
• Fourier/Laplace/Z comparison + impulse FT + convolution properties [4+4+2=10]
• VHF antennas, gain increase, and parabolic feasibility [2+6+2=10]
• System bus [1+4=5]
• SMPS + full-wave rectifier + electrical safety [10]
• OSI advantages/layers/examples + OSI/TCP-IP comparison [10]
• Satellite definition + Kepler + atmospheric loss [2+3+5=10]

NTC 2081¶

• Thyristor-controlled rectifier [5]
• Scalar/vector magnetic potential + magnetic boundary conditions [5+5=10]
• Root-locus construction and calculation [6+4=10]
• PLL + FSK + PCM [3+4+3=10]
• GEO uplink/downlink factors [5]
• NGN architecture/advantages/services + physical layer [4+3+3=10]
• TIR + optical Tx/Rx + satellite stabilization/tracking/earth station [3+2+2+3=10]
• Earthing/shielding + lightning measures [4+6=10]

NTC 2082¶

• 2-bit counter [5]
• Sampling theorem + A-law/mu-law [5]
• JK flip-flop from SR flip-flop [10]
• Dipole + polarization/matching + skin-depth calculation [2+3+2+3=10]
• Open/closed loop + closed-loop TF + poles/zeros [3+3+4=10]
• DS-CDMA synchronization + DS-SS/FH-SS + importance [4+4+2=10]
• Shock and lightning protection [5]
• Cache memory with example [5]
• NGN convergence + interoperability [3+3+4=10]
• TIR + fiber types/properties [5+5=10]
• Single/three phase + star/delta conversion [6+4=10]
• Cloud concept/types/apps + algorithm/flowchart [2+4+4=10]

Engineering Service questions remain valuable pattern evidence but are secondary to the exact NTC set.

6. Critical Technical Corrections Before Model Answers¶

The following must be corrected before they are condensed into memorized answers:

• Distinguish controlled-rectifier formulas for half-wave R load, full-wave R load, and continuous-current RL bridge.
• Resolve the Wien derivation and Colpitts feedback-ratio conflict.
• State practical CLA delay as hierarchical, not constant for arbitrary word length.
• State that simple parity detects any odd number of bit errors and misses even counts.
• Correct JK-from-SR conversion logic and verify the two-bit counter clock-edge convention.
• Add scalar magnetic potential and actual pillbox/loop boundary-condition derivations.
• Solve the 2082 copper skin depth at 5 MHz: approximately 29.5 micrometres.
• Reconstruct the exact 2082 EMT and control composites rather than inflating their limbs.
• Correct the FFT 1024-point direct-operation comparison.
• Remove the claim that ordinary FM places all power in sidebands; power is Bessel-distributed and may include a carrier component.
• State bandwidth conventions for BFSK rather than presenting one expression as universal.
• Separate T1 (24 channels, 1.544 Mbps) from E1/PCM-30 (32 slots, 2.048 Mbps).
• Add NGN vendor-interoperability analysis and DS-CDMA acquisition/tracking detail.
• Correct MEO range and distinguish one satellite leg from full GEO end-to-end delay.
• Add cloud computing to the 2082 ICT composite and cache locality/example to its PYQ.
• Qualify protection resistance targets, electrode dimensions, and bonding practices by applicable standard and site design.

7. Model-Answer Placement¶

Keep the opening box as a compact question index. Put the only full model answer beside the relevant chapter section in a collapsed VitePress details card. This keeps one chapter as the owner, reuses its figures, works on mobile and print, and avoids turning the page top into thousands of words.

Exact 5-mark template¶

### Model Answer — MOSFET Drain Characteristics [5 marks]

::: details Show exam-ready answer
**Opening:** ...

**Diagram:** ...

**Scoring points:** ...

**Key equation:** ...

**Conclusion/application:** ...

*Practice target: 8–9 minutes; 1–1.5 handwritten pages.*
:::

Scalable topic template¶

### Model Answer — BJT Switching Characteristics

::: details Show 5-mark core and 10-mark extension
#### For 5 marks
A complete short answer with definition, one diagram, formula, and 4–6 points.

#### Add for 10 marks
Only new scoring content: detailed mechanism, second visual, derivation/design,
comparison, limitations, and application.
:::

Fixed multi-part 10-mark PYQ template¶

### Model Answer — Satellite Communication [2+3+5=10]

::: details Show exam-ready answer
#### (a) Definition [2]
...

#### (b) Kepler's laws [3]
...

#### (c) Atmospheric loss above 10 GHz [5]
...

*Practice target: 16–18 minutes; allocate space in the ratio 2:3:5.*
:::

8. Rollout¶

Phase 0 — Correctness and normalization¶

• Fix the technical errors listed in section 6.
• Recombine all genuine PYQ fragments.
• Convert every range/invalid mark to a fixed 5- or 10-mark unit.
• Deduplicate answer ownership.

Phase 1 — Exact NTC papers¶

Write and time-test all 28 visible NTC questions first. These are the highest-confidence answers and define the style standard.

Phase 2 — Syllabus-gap answers¶

Add missing syllabus answers with no confirmed PYQ: wireless propagation models, optical switching/networks, measurement/test topics, cloud, cells, interfaces, and other explicit omissions.

Phase 3 — High-confidence predictions¶

Write predicted 5- and 10-mark variants based on repeated command verbs and adjacent PYQs. Keep them labeled [likely].

Phase 4 — Mock validation¶

Generate papers with 4 x 5 and 8 x 10, write them under three hours, and trim answers that cannot be reproduced within their time budget.

9. Quality Gates¶

A chapter is Green only when:

1. Every canonical question is exactly 5 or 10 marks.
2. Exact PYQ wording and split match the source transcription.
3. Every question links to one model-answer heading.
4. A 5-mark answer contains a definition, scoring visual/formula, 4–6 points, and a closing consequence.
5. A 10-mark answer contains a complete derivation/operation plus a second scoring surface.
6. Every 10-mark diagram-heavy answer includes at least one exam-reproducible labeled figure.
7. Word/page/time targets are stated and pass a handwritten timing test.
8. Formulas state their assumptions and symbols.
9. Predictions are never labeled as observed PYQs.
10. make build and make audit pass with dead-link checking.

10. Acceptance Target¶

Do not optimize for “242 long answers.” Optimize for a smaller set of canonical answer families that cover every unique syllabus/PYQ demand without duplication. Completion means:

• all 28 visible NTC questions have tested model answers;
• every explicit syllabus topic has at least one 5- or 10-mark answer family;
• predicted variants cross-link to those families;
• no Red files remain;
• a full mock can be completed within three hours.