Everything learned while compiling the code tables, written down — including the parts of hi-fi dating that do not work.
McIntosh numbered per model, so the number alone brackets the whole run. What the per-model tables cover, and what the run letters inside a number do not say.
2026-10-15 ↗Quad, the documented caseQuad is the one maker in this archive with an official serial-to-year table. How the 1953–1993 scale is read, and the five era thresholds that sit on top of it.
2026-10-15 ↗Marantz, the honest gapsMarantz tube icons have official years, the receivers have chassis stamps, and four numbering systems carry no year at all. What can and cannot be read.
2026-10-15 ↗Revox, tape machinesThe A77 was built in Switzerland and Germany in parallel. How the S and G prefixes select the range, and where the exact month and year actually survive.
2026-10-15 ↗Thorens TD-124The Mk I to Mk II changeover sits in a narrow serial window in spring 1965. Why the archive answers "Mk I or Mk II" instead of a single year.
2026-10-15 ↗Garrard, idler driveThe 301 serial is stamped on a plate under the platter. The two schedules, the community heuristic, and why nothing Garrard-specific reaches high confidence.
2026-10-15 ↗Kit equipmentKit equipment has no serial-to-year record. The catalogue number, the model year and the component codes date these units — the build year is not the shipping year.
2026-10-15 ↗UK component codesThree letters on a Hunts capacitor can be the most precise date in a British amplifier — and the one mark that dates a part, not the unit. How each code is read.
2026-10-15 ↗The honest refusalsScott, Leak, Radford, Sherwood, Harman Kardon and Dual published no serial-to-year table. Here is what each one left instead, and how the archive uses it.
2026-10-15 ↗ValuesThe archive publishes no invented figures: 49 models carry a sold-listing reading, 53 an estimate, and 127 no basis at all. How to read a sold aggregate honestly.
2026-10-15 ↗