Microcomputers As A Space Spinoff 3837
Surely you're thinking of The Digital Group, out of Denver. I had the impression (I don't know from what, other than the name) that they might have started like...
On Wed, 05 Jul 2006 12:02:41 -0500
That might have been the case in the earliest DRAMs, but by the time they were showing up...
Just each row, hitting a row refreshed the entire row. Early DRAMS it was every 2ms, later ones pushed it to 4ms spec. Once I populated a board designed for "pin 1" refresh with chips that did not have that feature and discovered that the actual retention was several seconds before data began to fade away long enough that the board pbutted with the test ROM loaded but failed with the CP-N boot ROM loaded.
Yes there were many variations on the special refresh cycle each with it's own peculiar timing requirements and signal cycling. Most chips supported more than one option but none supported all of them.
Yes there were, another technique was to use a DMA controller for the job.
Yes there was that, also the Z80 refresh counter was only 8 bits wide and some of the later chips had 9 bits of row address which made it useless.
I think it is more accurate to say that by then most of the "wrong" solutions had been discarded and a number of workable solutions were known. I fiddled with the refresh mechanism on the Torch Z80 card several times before finalising it (basically trying to get the cost down as far as possible) and then changed it completely for the 6MHz board because the timings were so different that the tradeoffs changed.
box one I liked the looks, and specs, for the OSI's but I never worked with one at...
-- The computer obeys and wins. A better way to focus the sun You lose and Bill collects. licences available see
Microcomputers As A Space Spinoff 3838
Alt Folklore Computers from Newsgroups