There are actually more terminaorts than that ;-). I seem to recall seeing a combined terminator and Unbius jumper, I think that goes in an 11/40 Anyway... Unibus is done properly. What I mman by that is that both ends of the bus are terminated. So there should alway be 2 terminators, no more and no less, in the system. The original scheme was to use M930 cards. These are dual-height boards of short length Originally they used discrete resistors, later on, they used DIL resisotr packs. But they're much the same thing. In an 11/45, for example, one goes at hte front of the CPU backplane, another goes in the Unibus Out slot of the last backplane/device on the Unibus. FWIW, this terminator is also used on the RK05 The M9301 and M9312 are bootstrap/terminator boards. As well as containign the termination resisotrs, they also ahve boot ROMs on them. One of these goes at the CPU end of the bus in soemthing like an 11/34 The M9302 goes at the 'far end' fo the bus in such a machine. Again it carries the termination resisotrs, but alos a little logic to assert SACK f ia grant signal gets to it. I forget the reaosn for this, but it is the reason why machines with such a termintor will give bus errors if the grant chain is open. One trick i didn't metnion (and which the guy with the 11/04 might try) si to replace the M9302 with aan M930 ans ee if the bus error goes away. If it does, then you have a grant problem (not necessarily an open connection on the backplane). Actually, for small systems, just thje CPU backplane or the CPU and 1 device backplane, you cna often get away with not fitting that terminator, at least for testing. So if you have this bus error probklem, just pul lthe M9302 and see what happens. -tony