DoPECC
back to Machine Log

Input Faults

Prev Entry

The machine was now able to maintain a display and this allowed more diagnostic microcode to be used, producing more diagnostic information:

ALU Test: pass
ALU Test: pass

ALU Test: pass

ALU Test: pass ALU Test: pass

With two tests not detecting keyboard input, the keyboard input hardware was the first place to go.

Keyboard Input Diagnosis

The 600 keyboard uses a keypress detection and key debouncing scheme that was inherited from the 700 series. It’s a complicated arrangement of TTL gates used as digital logic, TTL gates used with RC networks as delay elements and discrete transistor logic. It’s arranged as a series of latches within a circular self-latching ring and is rather complicated to debug.

Initially, it was worth testing the main gates with the HP545A Logic Probe to see if the key detection logic was properly initialized. This gave an immediate result when two gates in L4, a 7410 triple 3-input NAND, showed good input levels but undefined (illegal) output levels. Both gates were involved in the keyboard enable latch, and it was clear that this fault would block keyboard input.

The 7410 was replaced with a NOS IC. It’s interesting to see that the original IC is marked with the Signetics logo, 7338 (1973, week 38) and 376-0003 (Wang in-house part number). In the early digital logic era some manufacturers attempted to obscure their design information by ordering parts with in-house markings only. On removing this IC for replacement, it was found to be marked “7410” on the underside!

Keyboard Restored

The keyboard functions were restored by the above repair.

Calculator Microcode Test

Calculator microcode runs, generates a correct floating or scientific display and accepts numeric input.

However, there are anomalies, some functions work but others do not. More investigation is required.