I received an e-mail from someone who noticed that I was obtaining a lot of PDP-8 documentation on e-bay. He inquired if I might be interested in some PDP-8 gear. There was bunch of equipement that we talked about but he never mentioned an 8/L. Once I got to his place though, tucked on a corner shelf was an 8/L. It was a bit sorry looking (most of the switch handles had been "scavanged"). I decided that it needed a new home...mine!
I really didn't know what I had, because he hadn't done anything with it and it's condition was pretty much unknown (other than one of the modules had a tag on it -- not usually a good sign). When I got it home (along with all of the other stuff), I put it in a rack. I powered it up and noticed that all of the switches are a bit "unreliable". Many (like the load address and start) required more than a couple of tries before it would be recognized. I'm sure some contact cleaner should work ok.
I tried loading some patterns in to memory to see how bad things were. To my suprize most things worked except that bit 3 (remember PDP-8's number their bits like IBM does...most significant bit is 0) was stuck on. A little bit of investigation noted that the bit could be zero, but when read a second time, the bit was on. This is a familiar problem from a previous system. It's a failed inhibit driver. Looking up in the maintenance manual discovered that it was a G228 in position A23. Swapping it with a G228 in position B23, saw the stuck bit move from bit 3 to bit 7 (as predicted). Swapping out the G228 for another one yielded success. I entered a small little program to see if the "L" would execute instructions and success!
Here's the front panel.
Here's the backplane with gobs of flip-chip modules.