68008 Static Testers

Conlog — 1986

68008 static tester internals

Static tester for a 68008 from 1986.

The 68008 work was started in 1986 after joining Control Logic in Durban to finance a PhD (never completed). The 68000 processor selection would be made by management within a week, followed by a cross-compiler and databooks on our desks within another week. They were under a lot of pressure to deliver a sequential events recorder that had already passed the due date. My prior work on a segmented Z8000, multiple boards of Z80s for airline traffic conversion for travel agents, plus the sage advice from a friend during a visit to Silicon Valley about preferences of 68000 over NS32000 — made the 68000 selection an easy one for me. I prepared a short processor selection paper and presentation, but the decisions were quick and three of us were assigned to the project — Dave Ritten, Franco Pucci and me. Dave and Franco worked on the ICE while I worked on the static tester and the first 68008 wire wrap board.

My earlier (1981) Z80 breakpoint work, Z80 static testers and Z8000 static testers were the starting point, but we did not try to drive any particular bus as there was going to be an ICE. The static tester design was given to a lab technician to assemble, which is why it looked a lot neater than my prior rushed Z80 efforts.

Static testing our first 68008 board

68008 static tested and our first 68008 board

The first 68008 board used my previous wire wrap cards that were made while in the Navy for National Service. The regular memory board was also chopped up and wire wrapped into the card. The 68008 data bus was 8-bits and the address bus was 20-bits wide. The TIL311 hex displays on the board were useful for single stepping the processor, but in the early days we resorted to a logic analyzer at Conlog.

This was our first design that used a real-time kernel. The first pass of the kernel was developed while at Conlog, but the sequential events recorder was changed from being 68008 based back to 6809 based. When Dave, Franco an me were given the 68008 project, it was past a promised delivery date and development was prepaid fifteen months prior to our starting the design. We developed a board, a monitor, started on a real-timekernel and had the ICE running. The project termination was not my happiest day. The person asked for firm delivery dates on a 6809 version ended up overshooting the mark by many months. We were not far from delivery and only about six months into the sequential events recorder project.

Dave, Franco and I would go on to work on a video vending machine and retain the 68008 investment. Two previous versions of the video vending machine used mask programmed chips, and to “buy some time” the assembled units were sent by sea from Durban (South Africa) to Los Angeles in the USA. The chips would be flown later and married with the boards in LA. Guess what? They did not work, and after two mask production runs, the company was in a bit of financial trouble. We developed the prototype video vending machine, but the order was cancelled by the USA. There were massive retrenchments, and being young, we also left the sinking ship. The USA still used the mechanical assembly, but installed PC hardware without the development delays.

Wire wrap 68008 board

68008 board with single step and displays

The board shown on the left was a wire wrap card with single step and TIL311 hex displays on the address- and data bus. I am not sure when this board was developed, but it may have been 1987 or 1988 when we moved from Durban to Port Elizabeth. The earlier static tester belonged to Conlog, and by now we had a fair amount of confidence in the 68008. That was our last board that included so much debug assistance. In future, it would be software based as we developed an excellent monitor program that we could port within a few days to new hardware.