Mystery Q-bus boards

On 11/22/06, J Blaser <oldcpu at rogerwilco.org> wrote: I don't have any docs on the CompuServe board, but I can give you a historical tidbit or two... Originally, the 36-bit hosts were interconnected with a network of 16-bit machines that routed your session to the right host (GO TRAVEL, GO GAMES...) Eventually, real DEC PDP-11s got too expensive to maintain. At some point (mid-to-late 1980s?), CompuServe started making "node" hardware themselves. I've seen an Intel 486-based Qbus board, but was unable to obtain one. I _do_ have a later enclosure for the X.25 nodes - it's 3 9-slot Qbus backplanes in one largish rack-mount box. 3 reset buttons, 3 run buttons, etc. I have to reverse-engineer the front-panel (it's designed for CompuServe's 486s), but once I do that, I hope to be able to run 3 OSes simultaneously in one box (I'm thinking RT-11, 2BSD, and some other OS). As for your board, I see the 8 sets of 1-9 pads, but they don't look like DE-9 spots for me (the pins aren't numbered the same way, and with "3" linked to "5", and "7" linked to "4", it doesn't have the right feel, but perhaps it's a DLV11J pinout, not a PC-AT pinout). Also, I don't see any RS-232 line drivers. I _do_ see a 68B09 CPU, and 8 2K 6116 SRAMs. I'm guessing that the 68B40 might be a ROM since I don't see any ROMs anywhere else, but perhaps it's soft-loaded from the host. It does have a somewhat conventional DC005/DC004/DC010 Qbus interface. Try here... http://www.rdrop.com/~jimw/j-boards.shtml ... for a picture of what looks to be a CompuServe Node serial card. Anyone else have any observations? -ethan