A quick update on this. I built the clock described in this video over the weekend:
This went quite well, and since I have the 6502 kit, I could start on it. But Ben Eater has not completed the instructional series. Being able to replicate the procedure is important, but the best way to make sure you’ve really learned something is to change feature to something analogous and see if you can create the analogous situation.
I went on eBay and found 5 Motorola 6800 microprocessors for like $15. These are not the Motorola 68000 processors that turned up in stuff all through the 1980s, like the first generation of Mac, NeXT, and Sun boxes. These are 8-bit processors that were used in industrial applications and, interestingly, pinball machines. They came in the mail on Monday and I pulled up the datasheet. The question is, could I get it to start up and do a thing. So I snagged the datasheet and got a less than pleasant surprise. It needs two clocks.
Two clocks, alternating. It’s called a two-phase non-overlapping clock. That is, both clocks can never be on at the same time. They can both be off at the same time, though, without it being a problem. Okay, so I think for a minute and splitting the signal out and inverting it is not difficult. I have spare I/O pairs on the hex inverter. But something doesn’t feel right, and I am worried it does not guarantee that there is no overlap. After all, this stuff does not work instantaneously. So I Googled for “two-phase clock generator” and got this PDF:
Okay, so, yeah, look at the bottom one. It needs 2 ANDs and 5 NOTs. I have two gates on the AND free. And I have exactly 5 NOTs on the inverter free! So, I wired it up and lo-and-behold:
Well, if you watch through to end, you might have seen the thing skipped a phase at one point. I played with it a lot last night and I know this is crazy, but it seems to skip a phase more frequently the closer I am. In fact, if I can put my hand right next to the left-side blue LED and it will skip more than one phase! But I am probably imagining the correlation there.
Anyway, I am close, but there’s still some work to do.