PDP8/S

DEC PDP8/S

Here is the 8/S after restoration. It's connected to a wire-wrapped serial card and is running 4K FOCAL.

Restoration started with the power supply. Actually, my 8/S did not come with it's original power supply but I had one that was in the base of my straight 8. It was most likely used to power an analog interface of some kind. The components are the same as used in the 8/S. It looked much worse than it actually was. All of the green stuff was generated by a mouse chewing away at a piece of plastic. Everything cleaned up fine, and after reforming the capacitors, the supply worked just fine.

All of the flip-chip cards were covered with a thin film of some sort of greasy material. I decided to pull every card and clean them. As I cleaned each one, I also tested each transistor and diode for basic shorts and opens. I found a few bad transistors and even more bad diodes that way.

After cleaning every board, I powered it up. It mostly worked, but not completely. I had to do a little debugging to get it fully running. It passed a bunch of diagnostic tests I toggled in. Next, I decided to build a serial card so I could load programs into the machine. Normally, the 8/S would have been bought with a Teletype machine and it's interface to the 8/S. The interface is a small backplane with a few flip-chip cards. My 8/S did not come with the interface of course. Here are some photos of the wire-wrapped card I built. It converts the negative 8/S bus to TTL and uses an UART chip for the serial interface.

I used a small backplane and built some jumper cables out of ribbon cable and some oddball flip-chip cards. There are six jumper cable cards, the UART interface, and a voltage regulator I built on a seventh card.

Here is the main interface card. The UART is the large chip in the upper right. DIP switches are used to set the BAUD rate and serial formats. Rates include 110, 134.5 and 300 - 9600. The transistor level shifters can be seen on the lower left.

After getting the serial card working, I loaded FOCAL. It worked, but had some strange bugs. It would not do math. Could not even add. Character operations worked fine. This took a long time to figure out. It turned out to be some diodes on the core array that had turned into 600 ohm resistors and messed up loading of FOCAL into memory. This failure developed after the system was running for awhile. Over time it got worse, which made it easier to find. Here's a photo of a logic analyzer set up to trace out memory operations.

Finally, success! Here is a screen shot of FOCAL.71 after start-up. I decided to add a point plot and ADC interface to my system. Both are supported in FOCAL and look easy to do. Stay tuned...