This is the workbench computer (which I call Bench-1) with the LCD, with a board-edge connector socket wired to my PIC programmer (above and to the right), and a breadboard to simulate part of a product under development (to the left of the PIC programmer). The PIC microcontroller can be programmed in the ZIF socket or in-circuit. Bench-1 is on a 4.5" x 6.5" board with a 72-contact board-edge connector. All three boards are from Radio Shack. When the Bench-1 computer is used to control the PIC programmer, the A/D and D/A converters on Bench-1 are used for making sure Vpp is correct (usually 13V) and then setting Vdd=5V for programming and anywhere from 2V to 6V for verifying, so it's useful for production programming, not just prototyping. The larger ribbon cable goes to the printer, and the smaller ribbon cable is the link to the PC (or for that matter, to any computer with a text editor and that can send plain text out an RS-232 port). There's a diagram of the PIC programmer further down.

The computer has parts under the LCD as well. The blue trimmer on the left side is for LCD viewing angle adjustment. The light blue barely-visible button under the middle of the LCD (almost covered) is the reset button. Barely to the right of that are four annunciator LEDs. They're mainly for debugging, but I've hardly used them. To the right of the LCD is the five-key keypad, with Enter/Yes/Continue, Cancel/Escape/No/Exit, Help/Menu/special, Up, and Down. Not visible near the LEDs and keypad is the thin piezoelectric beeper soldered to the board so the whole board acts as a sounding board. (It is visible in the next picture.) The heat sink sticking up in the middle is for the 5V linear regulator.

Wire-wrap side. Generally, green wire is data bus, yellow is address bus, blue is I/O, white is control and select lines, red and black are power and ground, and orange and purple are data converter data and LEDs. You can't see it well in the picture, but power and ground are distributed in star systems from a point just left of the center. With 4MHz parts, it all seems to work fine up to just over 7MHz, so I run it at 5MHz for some margin. I've never had a single reliability problem with wire-wrap. The only hardware problems I've had with this computer were from faulty socket contacts, and were cleared up by pulling the corresponding IC out of the socket and plugging it back in so the contact surfaces would have a self-cleaning effect.

Here Bench-1 is plugged into a pair of breadboards for verifying the preliminary design of a front panel for a product with two gray-code rotary encoders and tons of buttons and LEDs. The computer has 3 65C22 VIAs, 3 65C51 ACIAs, and a fast A/D and D/A, but most of the I/O for this one comes through the serial port on one of the VIAs.