TSU Meeting MONDAY — March 4, 2024

David, Joe, Paul, Carl Miles, Ryan Gray, Adam Trionfo, Jeff Burrell, Gustavo, Jeremiah, Rich Sutherland, Mike

Joe walked through the workflow he developed for archiving his old ZX81 tapes after a JavaScript tape-converter kept rejecting the raw recordings. Using Audacity, he normalizes and inverts the signal, runs it through a 'rectivert' rectifier to push the noise below the zero line, removes the DC offset, and then hand-silences stray noise pulses by amplifying them to -50 dB so the spacing between bits is preserved. The end-state waveform reads cleanly as zero (four-pulse) and one (nine-pulse) bits, and the converter then produces P-files with no errors. The group compared notes on tape decks (the Realistic Mini-Set 9 as a cheap 2020 equivalent, the GE computer recorder, head azimuth alignment, the value of using TDK/3M cassette shells over budget tape) before Paul described a Sinclair-era program called Chromosoft (G. Russell Electronics / Bill Russell) that produces three or four perceived colors on a black-and-white TV by oscillating the image — a Newton-disc / persistence-of-vision trick rather than an actual color signal. Paul is mailing the tape to be preserved.

Carl gave a long update on the TS1510 cartridge-interface clone. Boards are shipping to community members in the US, Canada, and Portugal; he now has 35 PCBs in hand and has switched to 27C512 (64K x 8) flash so 16K cartridge images work, which the original SuperFo design did not support. He has been experimenting with converting TS1000/ZX81 programs into cartridge images: Crazy Kong (~7.3K of code) appears to work after disassembly-style conversion, but Kong's Revenge (13.5K of BASIC plus trailing display-file/variable data and an auto-start at line 9999) corrupts on load and is still being chased. Ryan is writing a utility that takes a P-file and emits a ready-to-burn cartridge ROM image, which will make wider distribution of cartridge software practical. Flight Simulation (16K) loads and runs from the 16K-capable build. Carl plans to verify the carts on a real TS1000 (his testbed is a TS1500-in-1000-case clone) since 1000 owners will need to type RAND USR 8192 to start.

Carl also showed his in-progress Grant Wilco (2009) ZX80/ZX81 hybrid clone — all discrete TTL, dome-switch keyboard with 3M adhesive rings, composite (no RF) video, USB-power, and a 64K ROM chip with a three-position DIP switch selecting among eight 8K ROMs including multiple ZX81 revisions, ZX80 ROMs, Forth, and the ASZMIC machine-code monitor. He has five board sets and parts in hand and is open to building extras. The build is partly motivated by getting onto Paul Farrow's Chroma 80 FPGA color board (and an eventual Chroma 81), which Carl demoed: SCART-RGB output, FPGA-based color generation, and games that have been either colorized or rewritten to use real-time color.

Adam introduced Baud Day (3/12/24 — 300/1200/2400) and his plan to host a BBS-calling session on real Atari and TS2068 hardware after interviewing two active sysops (The Basement, Southern Amis); roughly a thousand BBSes are still up, some reachable by telnet via Linux gateways. David is building hardware to match: a community-posted schematic that drops the modem half of the TS2050 and exposes the 8251 as a straight RS-232 port (boards on the way), plus a retro Wi-Fi modem (ESP32 + microSD speaker for handshake noises) housed in a Hammond case with a red lens. The WESPI Pi-based loader and Fred Nachbaur's CNX-Term-80 (high-res 80-column terminal for the 81/1000) came up as ways to actually use a 2050-style interface. Greg Hartner's high-res ZX81 work (including a Mandelbrot demo and a high-res BASIC extension co-written with Fred) and Johan Koelman's '81 Games for the 1K ZX81' book got plugs. Jeremiah shared an assembly-language Dragon-Quest-style test screen he is building for the TS2068 with HotZ; Gustavo explained that the TS-Pico exposes the standard tape save/load entry points to assembly with one extra command to set the TAP filename, and the group agreed the relevant section of the existing TS-Pico manual should be put on GitHub for developers. Mike pointed to Dan Roy's July/August 1983 SYNC article 'So you wished you had bought a TS2000', which attached a TMS9918 plus sound and joystick ports to a TS1000 and was almost shipped by CompUSA-NJ as 'Colors in 81'. David announced that David Ahl is donating his complete SYNC run and related Creative Computing titles for archiving, and that a 1000-specific edition of 'Katie and the Computer' is among them. The next meeting is Sunday March 17 at 2 PM Eastern; topic to continue: why the TS-Pico uses a rewritten-ROM command syntax instead of the syntax-error/special-character convention used by Larken/AERCO-style disk systems.

Highlights

Joe demoed his ZX81 tape-archiving workflow in Audacity — invert, rectivert, silence stray noise pulses at -50 dB to preserve bit spacing
TS1510 clone now supports 16K cartridges (27C512 flash); Crazy Kong converts, Kong's Revenge and a line-9999 auto-start still fight back
Ryan is writing a P-file → 1510 cartridge ROM converter so community software can be packaged as carts
Carl is building a Grant Wilco ZX80/ZX81 hybrid with a switchable 8-ROM chip (incl. ASZMIC and Forth) to pair with Paul Farrow's Chroma 80
David Ahl is donating his complete SYNC magazine run plus Creative Computing titles for preservation

Projects discussed: TS1510 Cartridge Interface Clone · TS-Pico · ZX-Wespi / Wespi-Nano · Library Tape Archiving Process Topics: TS1510 cartridge interface · Cartridges, EPROM & dumping · TS1000 / ZX81 / TS1500 · Tape & library archiving (TAP/TZX) · Modems, serial & RS-232 · Pico / modern interfaces (UnoDos, etc.)

Summary generated from the meeting recording; the full transcript is kept in the group's private archive.