EE 351 · Week 7 · Instructor record
Reference and Engineering Accuracy Record
What the Week 7 materials rely on, what was checked and how, where sources disagree, and what remains unverified.
← Lecture slidesStudy articlePractice sheet
1. Sources
| Source | Use | Notes |
|---|---|---|
| Reference A — DataGenetics, “How analog TV worked”, April 2018 (datagenetics.com/blog/april12018) | Raster scanning, blanking, sync and porch durations, interlacing, half lines, vertical blanking, why 525/625 lines, compatible colour history. | Read from the start through the compatible-colour section (the later material on shadow masks and displays was not used). The site’s TLS certificate had expired when fetched, so the text was retrieved with certificate checking disabled (read-only). The article’s animations were not used; all figures here are original. |
| Reference B — Chapter 23, Television, supplied PDF (37 pages, textbook chapter; author and edition not stated on the pages read) | Section 23-1 (TV signal), Figs 23-1 to 23-13, Examples 23-1 and 23-2, 4.2-MHz resolution case, I/Q, colour burst, receiver description. | Pages 1–14 were read and the figures inspected. Pages 15–37 (cable, satellite, digital and HDTV sections) were not used. This chapter is a supplementary reading; the official course textbook on the course page is Lathi, Modern Digital and Analog Communication Systems, 5th ed. |
Official curriculum — program-data/ee_curriculum.json | EE 351 CLO wording and mapping. | Read only; not modified. |
| General engineering knowledge | Items marked “general knowledge” below. | Not checked against a supplied reference in this work; see section 3. |
The supplied image OldTV.png was dropped at the instructor’s request; no photograph is used. Textbook figures were redrawn as original diagrams and cited; no textbook page or figure is reproduced, and the PDF is not distributed with the site.
2. Numerical specifications
| Item | Value used | Basis | Status |
|---|---|---|---|
| Lines per frame (NTSC) | 525 (2 × 262½) | Ref. A; Ref. B p. 6 | verified |
| Field / frame rate (colour) | 59.94 / 29.97 Hz | \(2f_H/525\), \(f_H/525\) computed; Ref. B p. 6 | verified |
| Line frequency (colour) | 15 734.27 Hz (≈ 15.734 kHz) | \(4.5\text{ MHz}/286\); Ref. B p. 6 gives 15 734 | verified |
| Line period | 63.556 µs (≈ 63.56 µs) | \(1/f_H\); Ref. B rounds to 63.6 µs | verified |
| Horizontal blanking / active | 10.9 / 52.66 µs (front porch 1.5, sync 4.7, back porch 4.7) | Ref. A gives sync 4.7 µs and front porch 1.5 µs; the back porch value 4.7 µs and the 10.9-µs total are general knowledge consistent with the 63.56 − 52.66 balance | rounded in textbook |
| Textbook blanking / active | ≈ 10 / 53.5 µs | Ref. B p. 7. Used only in the textbook worked case; labelled as such | rounded |
| Colour subcarrier | 3.579545 MHz (= 227.5 \(f_H\) = 3 579 545.45 Hz) | Ref. B pp. 9–11 and Fig. 23-9; computed | verified |
| Picture-to-sound carrier spacing | 4.5 MHz (= 286 \(f_H\) exactly) | Ref. B pp. 2, 4; computed | verified |
| Channel width, System M | 6 MHz | Ref. B Fig. 23-1 | verified |
| Picture carrier | Lower edge + 1.25 MHz | Ref. B p. 4 and Example 23-1 (channel 39: 621.25 / 625.75 MHz recomputed) | verified |
| Video bandwidth | ≈ 4.2 MHz | Ref. B p. 7 | verified |
| Resolution estimate, 4.2 MHz | \(N=2T_{\text{act}}B\): 449 (53.5 µs); 443 (52.66 µs); ≈ 337 TVL per picture height | Ref. B p. 7 gives 449.5 ≈ 450; the per-height conversion (× ¾) is added here | verified |
| Vertical resolution factor | \(R_V=0.7N_L\) (NTSC); \(0.8N_L\) (textbook PAL example) | Ref. B pp. 8–9; called the Kell factor in the wider literature (general knowledge) | verified |
| PAL example | 51.2 µs, 0.1 µs/element, 5 MHz, \(R_V=464\) | Ref. B Example 23-2; recomputed | verified |
| Y′ weights | 0.299 / 0.587 / 0.114 | Ref. B p. 9 gives 30 / 59 / 11 % | verified |
| I, Q coefficients | I = 0.596 R′ − 0.274 G′ − 0.322 B′; Q = 0.211 R′ − 0.523 G′ + 0.312 B′ | Ref. B p. 9 gives rounded 60/28/32 and 21/52/31 %. Three-decimal values are standard; reproduced to 4 d.p. from the (B′−Y′, R′−Y′) rotation by 33° (script check) | verified |
| Chrominance bandwidths | I ≈ 1.5 MHz, Q ≈ 0.5 MHz | Ref. B Fig. 23-9a / 23-10. This is the historical design; later receivers often use narrower, equal bandwidths (stated on the slide) | historical |
| Burst | 8–12 cycles (nominally 9) on the back porch, at the −(B−Y) phase | Ref. B pp. 10–11 and Fig. 23-9b; nominal 9 cycles is general knowledge | verified |
| Sound | FM, ±25 kHz peak deviation, 50 Hz–15 kHz audio | Ref. B p. 2 | verified |
| Receiver IFs | Picture 45.75 MHz, sound 41.25 MHz; channel 4 LO 113 MHz | Ref. B p. 14, recomputed | verified |
| PAL line / channel data | 625 lines, 15 625 Hz, 64 µs, 4.43 MHz, 5 MHz video, 7/8-MHz channels (B/G) | Line numbers: Ref. A and Ref. B. The 4.43-MHz subcarrier, 5-MHz video band and 7/8-MHz channel widths are general knowledge | general knowledge |
3. Discrepancies and judgement calls
- 3.85 MHz versus 3.579545 MHz. Ref. B p. 2 says the colour subcarrier is 3.85 MHz; pp. 9–11 and Fig. 23-9 give 3.579545 MHz. The lesson uses 3.579545 MHz everywhere and the article names the discrepancy.
- Which textbook case is NTSC and which is PAL. The 4.2-MHz resolution derivation is in the body text (p. 7); Example 23-2, which many readers assume is the NTSC case, is PAL (5 MHz). The lesson covers both and labels them.
- Textbook rounding. 60 Hz / 30 Hz / 63.6 µs / 10 µs / 53.5 µs are rounded; the lesson shows both rounded and exact values and notes the roughly 2 % effect.
- “Interlacing halves the bandwidth.” Ref. A states this. Strictly, interlacing gives the flicker rate of 60 fields/s at the detail and bandwidth of 30 frames/s, so the saving is relative to sending 60 full frames per second. The slides use the precise wording.
- Polarity. The lesson draws baseband video with white high. The textbook’s Fig. 23-6 draws black above white, which corresponds to the negatively modulated RF envelope (sync tip = maximum carrier). The slide “Why the receiver needs sync pulses” states both.
- Horizontal resolution wording. The textbook’s \(R_H\approx450\) is across the full width; the conversion to ≈ 337 TV lines per picture height is added and flagged.
- Vestige. The textbook figure shows 1.25 MHz between the lower channel edge and the carrier and a “partial lower sideband”. The demo draws a 0.75-MHz vestige with an odd-symmetric slope (sliderable 0.25–1.25 MHz) as an idealisation; the actual vestige roll-off is set by the receiver filter and is deliberately not drawn as a specified mask.
4. Historical analog versus present practice
- Everything in the lesson is analog NTSC System M unless labelled otherwise.
- Full-power US analog broadcasting ended in 2009; ATSC digital television uses 8-VSB in 6-MHz channels (general knowledge, stated on the slide and in the article). Frequency chart and channel numbers are the textbook’s historical US plan; the textbook’s remark about future auctions of upper VHF/UHF channels is not repeated as current fact.
- The textbook’s analog receiver and cable/satellite/digital discussions are not used.
5. Curriculum alignment policy
- Slides carry the same machine-readable attributes as Week 6 (
data-clo,data-official-so,data-clo-alignment,data-alignment-status="suggested"). All are suggested teaching alignment, not official mappings or formal evidence. - Only two existing course learning outcomes are referenced: CLO 1.1 (explain modulation and demodulation techniques using time- and frequency-domain representations; official SO1) for the colour-quadrature and VSB slides, and CLO 2.2 (solve communication system problems from specified system parameters and requirements; official SO1) for the timing, bandwidth and channel calculations. The SO value shown is the existing mapping in
ee_curriculum.json; no PI mapping is displayed. - Scanning, interlacing and raster demonstrations are tagged foundational only; they are not claimed as direct CLO evidence.
- No new CLO–SO–PI relationships, attainment thresholds, QAC assessments or rubric results were created. The seven exercises are formative learning exercises, not a new Major Exam assessment.
6. Checks performed
- Arithmetic. All numbers in section 2 and in the exercises were recomputed with a Python script (line rates, \(N=2T_{\text{act}}B\), I/Q coefficients from the rotation, channel carriers, burst duration).
- MATLAB.
week07_tv_video_chain.mwas run in MATLAB R2024b batch mode; see the verification statement in the study article. - Demos. Browser checks of the five demonstrations are recorded in section 7 below.
7. Browser and layout testing
Carried out on 20 October 2026 on Windows 11 with the Claude desktop app’s built-in browser pane and headless Chrome 155, serving the repository over local HTTP. Three defects found during this testing were fixed and re-tested (a clipped display equation on the NTSC-numbers slide, grid columns that overflowed at phone width, and figure labels that overlapped).
- Slide fit at 1440 × 900. All 62 slides were driven one at a time by URL hash and measured for vertical overflow, horizontal overflow of tables, figures and widgets, and clipped display equations: none remaining. No slide exceeds 88 % fill. Not tested at 1280 × 720.
- Phone width (375 px). No horizontal overflow on any of the 62 slides, nor on the article, practice sheet, accuracy record or course homepage. Demos reflow to one column. Visual spot-checks of the raster, interlacing, chrominance and VSB demos used a 375-px iframe in headless Chrome.
- Demonstrations. Raster: image selection, line slider, play/pause, and “step one line” advances exactly one line. Interlacing: Field 1, Field 2 and Frame views, dropping field 2, and the moving object (local-brightness ripple about 80 % progressive versus 50 % interlaced). Bandwidth: 240 stripes gives 2.277 MHz and \(N_{\max}\)=443; 500 stripes at 4.2 MHz is reported as unresolved (uniform grey). Chrominance: (0.8, 0.4, 0.2) gives \(Y'\)=0.497, I=0.303, Q=0.022, 119°; pure red gives 103°; grey gives zero chroma; a 30° receiver phase error rotates the reconstructed hue. VSB and channel: bandwidths 8.40 / 4.20 / 4.95 MHz; carriers for channels 2, 6, 13, 14, 39 and 69 equal the textbook’s own examples where it gives them (channel 39: 621.25 / 625.75 MHz; channel 6: 83.25 / 87.75 MHz).
- Reading mode. All six widgets render with non-blank canvases.
- Quizzes. Seven quiz blocks, each with exactly one option marked correct; feedback verified on exercise 2.
- Links and markup. 188 links and anchors on the five new or edited pages resolve; no duplicate element IDs in the deck. No unrendered LaTeX delimiters and no MathJax errors on the deck, article, practice sheet or accuracy record.
Not tested: the browser pane suspends requestAnimationFrame when it is not in view, so the animations were stepped by calling each widget’s update and draw functions directly, and real-time smoothness was not measured. The reduced-motion behaviour (animations start paused) was checked in code, not in a browser. Keyboard use and screen-reader behaviour were not tested beyond using native buttons, radios, sliders and selects with labels. Firefox and Safari were not tested.
8. Known limitations
- The chrominance demonstration shows the colour-difference chroma as a baseband sinusoid with a few cycles and omits transmission gains, set-up and the chroma bandpass; amplitudes are indicative.
- The interlacing demo uses 21 lines and a slowed clock; the flicker plot is a simple phosphor-decay model, not a perceptual model.
- The VSB demo uses an ideal, odd-symmetric slope and an ideal 4.2-MHz video cut-off; it shows the principle, not the broadcast emission mask or a real SAW response.
- The textbook and DataGenetics were not compared against the original FCC/ITU standard documents.