This system was heavily customized by its former owner, including a home-built serial board (wire-wrapped). The other boards are a Cromemco Z80 CPU board, Fulcrum Omniram 64KB Static RAM, Morrow Disk Jockey 2D/B controller, Solid State Music IO-4 serial/parallel board, Scitronics real-time clock board and a Compupro Active Terminator board. The system was fairly dirty on the outside, but cleaned up nicely.

This system hadn't been powered up for 20 years, so I took precautions with the huge electrolytic capacitors in the linear power supply. I connected a 6.8K power resistor in series with each capacitor and a variable DC power supply, then ramped up the voltage in steps while monitoring the current. Once I got them up to their rated voltage, I let them sit for several hours. I don't know if it was necessary but when I powered up the system, nothing exploded.

To check out the system, I installed only the active terminator card, the 64KB RAM and the ZPU. After I disabled the Power On Jump (POJ) on the ZPU card, I was able to use the Deposit and Examine switches on the front panel to enter a short program. It does nothing more than scan an LED across the front panel output buffer (Cylon style), but shows that the system generally works.

I attempted to boot the system from the 8" floppy drives, but the Disk Jockey 2D/B controller appeared unable to talk to the drives. I eventually found that the A: drive had a dead short across the 24V line, and was dragging down the supply. I changed the other drive to DS1 and installed the terminator pack, and the system appeared to load the system tracks. Nothing displayed on the terminal connected to the serial port on the 2D/B card though, and my guess was that the diskette was configured for either another port or one of the two video boards that came with the system. After some more experimenting, it turned out the the boot diskettes were set up for the Dynabyte Naked Terminal board. It produced some garbled text on boot, in which the word BIOS and "version" appeared. I don't have the ASCII keyboard used with this card, or any information about switch settings or pinouts.

I was eventually able to extract software from one of the system disks which happened to contain the CBIOS source code for the 2D/B and the Dynabyte. Through a convoluted process, I managed to rewrite the CBIOS for the 2D/B serial port, reassemble it and put it back onto a disk. The modified disk boots perfectly! For the nuts and bolts description, read the IMSAI 8080 Boot Disk Recovery page. I also located the problem with the other floppy drive (shorted tantalum capacitor), replaced it and now have two working drives.