Microcomputers As A Space Spinoff 3838

That might have been the case in the earliest DRAMs, but by the time they were showing up in "personal computers" (and the Altair had a dynamic RAM board early on, though many said it didn't work), you only had to refresh a column or row.

Yes, but they were often pretty bulky. You had a set of address lines on the DRAMs, and you'd have to multiplex the upper address lines, and then the lower address lines (the sequence might have been reversed), to address the memory. When it came time to refresh, you'd have to multiplex a counter of the right number of bits onto those address bit pins of the DRAM. So apart from miscellaneous glue, in order to use DRAMs, you needed a three way multiplexer to drive the address lines of the DRAM (selecting between the lower address bits of the CPU, the higher address bits of the CPU and the counter), and an actual counter to do the refresh.

It wasn't a particularly complicated process, just bulky doing it with discretes. But the ICs that came along to do the refresh had large numbers of pins to them, so it is debateable how much they simplified the process.

Some designs avoided some of the complication by doing an interrupt, and then refreshing by software, merely reading the needed locations so they at least avoided the need for a refresh counter.

And of course, the Apple II used the video circuitry to refresh the DRAM.

Whether or not DRAM came before static RAM, in "personal computer" circles static came first. And DRAM appeared complicated (it did need that extra circuitry) and maybe not really understood at the beginning. Obviously there are tricks, but there was a trickle of articles in the hobby magazines about DRAM, and they'd all start with caveats about dynamic RAM. That sort of thing doesn't move the art forward, it basically says "hey, avoid DRAM if you can". But obviously DRAM can't be that much of a problem, since for decades it's been the standard for small comptuers.

1980 might have been when the construction articles started appearing, I seem to recall at least one article before that, but DRAM had taken over pretty much by that point. The Apple II and most of the other all in ones that showed up in 1977 used DRAM. I have a Processor Technology DRAM board that came out that year.

I suspect one thing that happened was that the construction articles lagged on the matter of DRAM, until the time where interest in building computers was fading, since by then there was a good selection of commercially available home computers. By 1980, there were fewer and fewer articles about building a computer, and more about accessories.

Michael