I was reading an article on a 6522 based parallel card for the Apple II

https://archive.org/details/JohnBellEng ... 1/mode/2up

The circuit diagram uses a 74LS05 to create a phase 2 signal since that signal isn't on the Apple II bus. The circuit has pull-ups on the output lines. If you look at the datasheets for the '05, you see that it's an open collector output and pull-ups are required.

I don't have any '05's but I have some 74LS04's. Which seem to be the exact same thing but not open collector outputs and the datasheets don't seem to mention any pull-ups on the output lines.

So I was wondering, can you see any reason to use an 05 vs an 04 in this design? By that, I mean using the 04 without any pull-ups on the outputs.

Furthermore, if that's the case, then what would be the point of the '05 in the first place?

Sorry for such a n00b question.

Thanks!

_________________
Cat; the other white meat.

It looks like the resistors R4 and R3 set the timing of the phase 2 signal, besides simply pulling up the outputs of the inverter gates. Gates D and E (the ones with pins 11,10 and 9,8) would work fine if you used a 74x04 and removed the pull-up resistors, but the other two gates being open-collector is integral to how the clock generation works.

There are ways to design an oscillator around a non open-collector inverter, though. This example refers to CMOS inverters but you might be able to tune it for 74LS with some experimentation. https://www.falstad.com/circuit/e-inv-osc.html

Anything you have on hand that can generate a square wave should work, though. 555 timers, crystal osc modules, Schmitt-trigger inverters, can be used for instance.

EDIT: Nevermind, I didn't look closely enough at the schematic. Basically disregard everything I said.

I think the phase shift would be difficult with totem-pole outputs. Basically they are trying to delay Phi-1 by ~500ns so it aligns with Phi-2. I've done stuff like this before with Apple II's and what I did was run a small bodge wire from Phi-2 on the processor to my custom card. A little dab of solder on the MB end and a little Dupont style single pin connector to the custom board end.

_________________
Bill

Oops, I didn't notice it was taking Phi-1 as an input. That's what I get for sneaking peeks on the forum while at work!
Running a wire from the MB might be the most practical idea. Would it be overkill to make it a two pin connection and use a twisted pair to create a return path?

Yeah, that would be perfect. I didn't do that before, but it would definitely be a better and more robust option.

_________________
Bill

Well, running a bodge wire wouldn't work for me as I would like to make an expansion card for other people.

I could go with the 05's if I have to.

Thanks for the information!

_________________
Cat; the other white meat.