On Mon, Feb 8, 2016 at 1:33 PM, tony duell <[email protected]> wrote: >> > I heard that when Bristol University physics department got its first VAX >> > (an 11/750, >> > somewhat before my time), it was cheaper to buy 256K memory boards full of >> > 16K RAM >> > chips, clip them out, clean out the holes and solder in 64K RAMs rather >> > than to buy 1MByte >> > boards full of 64K RAMs from DEC. And that is what they did.... > >> machine? I doubt DEC would be real happy with that. Also, >> the 64K RAM chips need an extra address pin, were the boards >> laid out with that signal already in place? > > Yes. IIRC there were some jumpers to re-set, but the boards were the same.
Yep. The boards were used in the 11/70, the 11/730, and the 11/750. I don't know if you could use the 256K boards (populated with 4116s) in the 11/730 due to the tri-voltage 4116s, but even if they worked, you wouldn't want to - 5 of them just isn't that much RAM. In the case of the 11/750, one of mine, BT000354 (early S/N) shipped from DEC with 512KB as two M8728 (256K) boards that I later upgraded to 8MB by removing the old memory boards and memory controller board, adding eight M8750 boards (it was not worth clipping and upgrading the actual boards at the time - 1MB boards were under $300 each by then), and L0016 memory controller board (8MB max) and adding the additional multiplexed address line to the memory slots. We ran it that way for many years. -ethan
