On 2018-Nov-20, at 11:18 AM, Paul Koning wrote: >> On Nov 20, 2018, at 2:07 PM, Brent Hilpert via cctalk >> <cctalk@classiccmp.org> wrote: >> >> ... >> The PCB-transformer key design is novel, but (in the TS-1) so is the >> scanning & communication. >> There is no microcontroller on the keyboard, scanning is performed by the >> main unit, it sends a key matrix address to the keyboard in serial, the >> keyboard sends back nothing but a binary key-pressed indication. > > I'm not positive, but my impression is that the same is true for the VT-100 > keyboard.
You may be right, rings a bell from working on one, Wasn't intending to suggest the falco was necessarily unique. IIRC, the VT-100 keyboard comm involved standard async UARTs. It's a little weirder on the TS-1, one edge of pulses on the address-line are a clock-pulse, some uS later - determined by a monostable - the state of the same line is sampled for an address bit. After 8-10 such pulses the matrix is strobed, involving another monostable, so there are some timing issues both sides have to independently account for.
