_________________
x86?  We ain't got no x86.  We don't NEED no stinking x86!

When you report about "the bus timings" -- and even center one out for special attention ("especially the all-important tBH") -- readers are left with the impression that tBH was just one of a list of values included in your analysis. Aren't you maybe kinda stretching things? What's up with that? You have a nice prose style when you write, but knowing when to stop is a valuable skill, too. I'm trying to say that in a nice way, even though this is via email (sort of).

Be that as it may, hats off to you for actually performing some sort of experiment. It's something I wanted to do myself and haven't gotten around to.

cheers,
Jeff

_________________
In 1988 my 65C02 got six new registers and 44 new full-speed instructions!
https://laughtonelectronics.com/Arcana/ ... mmary.html

_________________
x86?  We ain't got no x86.  We don't NEED no stinking x86!

Well, it's an interesting subject. If you do collect some detailed information I hope you'll share it with us.
Easily hand-soldered? Heck, you're doing better than I am if "easy" is the word that comes to mind! But the .050" stuff is do-able.

Although it's slightly off-topic, I will quickly mention that proto-board using .050" grid is available. Unfortunately the selection is very limited; I know of only two examples, the Vector 8021 and the 8028. The 8028 is actually a combo board with areas of .100", .050" and 2mm grid.

cheers,
Jeff

_________________
In 1988 my 65C02 got six new registers and 44 new full-speed instructions!
https://laughtonelectronics.com/Arcana/ ... mmary.html

_________________
http://WilsonMinesCo.com/ lots of 6502 resources
The "second front page" is http://wilsonminesco.com/links.html .
What's an additional VIA among friends, anyhow?

I have used the WDC schematic for bank latching using an ACT573 and the ACT245 and had stable operation with a 10mhz clock. The rest of the glue logic was HCT based and the circuit was wire wrapped. Just keep the wires as short as possible and use plenty of bypass caps. (soldered in if possible) as close to the chips as you can get.

Hi Clockpulse - good to hear about your success. Could you comment please on how short your wires are and close the caps are to the chips? I prefer to see "a few inches" and "less than half an inch" to "as close as possible", because it's much more helpful to anyone without the experience.

Thanks!
Ed

It's not easy to keep the leads on the caps. short when using wire wrap sockets but I try to keep them right next to the socket or between sockets with a distance less than half an inch. Supply and Ground rails help. In some cases I have soldered the caps on the bottom straight between the Vcc and gnd pins of the chip (best for buffer drivers like the 245).

I 'try' to keep the wire wrap wires in the 2 to 3 inch range. Of course when you're running the data and address busses the total length might add up to 6 or 7 inches. A board with a ground plane helps, and keep the wires against the board.

Thanks!

According to WDC's Programming the 65816 it's a very similar die, but with differences in the metal layer(s):

I stand corrected - thanks!

_________________
x86?  We ain't got no x86.  We don't NEED no stinking x86!

Indeed. The same document I linked to makes it clear that the '802 is circuit compatible, and that it doesn't have the bank address when phi2 is low (limiting it to 64KB):

Interesting point, well worth mentioning. Still, do we conclude that "something had to have been done," or, "it would have been nice if something had been done"?

If the '816 and '802 are the same basic die, differing only in some of the metalization, it doesn't seem plausible that invalid-address cycles would appear (as we know they do) on the '816 but not on the '802. What seems more likely to me is that the '802 also exhibits the invalid-address cycles, and, since VPA and VDA outputs are absent, this is a hazard with no good solution. IOW, people using the '802 would need to be on guard against the possibility that an invalid address would read from an IO device, which, on some devices, can erroneously alter status.

Odd (and disappointing) as this seems, it doesn't contradict the definition of the '802 as a drop-in replacement for an NMOS 6502, since the latter exhibits the same hazardous behavior. (Or was the '802 touted as a drop-in replacement for the 'C02 as well?)

cheers,
Jeff

_________________
In 1988 my 65C02 got six new registers and 44 new full-speed instructions!
https://laughtonelectronics.com/Arcana/ ... mmary.html

Back in the mid 80's I did swap the 65c02 in an apple IIe for the 802 and it worked fine. The Merlin 8/16 assembler I used at that time had an option that would actually use the 802 instructions to enhance Merlin's performance.