Posted By
SukkoPera
on 2022-09-21
04:08:04
|  Re: External User Port for the C16: Let's make it!
@MCes: Thanks for your offer, we can surely work on that for the next version, but first I'd really like to make a bare-bones one that everyone can easily build. That is always one of my goals, I try to stick to the smallest subset of common components, which of course means avoiding CPLDs and the programming HW they require. GALs are already borderline, but at least they can be flashed with che cheap TL866 programmers, which every good retromaniac should already have so that they can tinker with ROMs.
So forgive me if I insist with the resistor mod, I'd much prefer going this way unless we find out it is really impossible to get it working. First let me mention that I had done some tests and they all worked very well, you can even see in the logic analyzer screenshot I put in my essay that the CPU can drive D2 to >4V even when U11 is outputting low. But now I am no longer sure I did those tests with an 8501 CPU! I had two C16 mainboards on my desk, one of which was running on your CPUC16 adapter. I moved things around quite a few times and I really cannot tell what setup I had in my final tests. That was really stupid on my part, I will redo the tests ASAP!
But in the meantime: do you think using a bigger resistor on U11's output (10k?) and/or adding a pull-up resistor on D2 could help having safe voltage levels in all conditions? We need to guarantee two things:
1. The CPU can read the correct level of CST_SENSE when there is no expansion board installed.
2. The CPU can safely read and write high or low from/to an expansion board responding at $FD1x regardless of what U11 is outputting. I guess problems can only arise when U11 and the CPU/board are outputting different levels, when they are outputting the same level, there should be no problem, right?
BTW, the datasheet you linked is for the 7501. Is it possible that the 8501 has better output driving capability?
About the 74654: I replicated that circuit from @dmantione's schematics linked earlier in this thread. He's selling boards with that circuit so I guess they work fine, unless I made mistakes in the circuit, of course. It is supposed to work with both 74654 and 76652. With the 654 the diodes must be replaced bith 0-ohm resistors as far as I understand. Does that look fine?
@gerliczer @MMS: No problem, we can also talk about mouse support here if this card can help with that. But let me say something: I have already designed an USB-to-Amiga mouse adapter (don't look for it, I haven't released it). I can easily recycle some of that code and output whatever you want. So if you can define a suitable mouse protocol, be it serial or whatever you want, I will implement that and we can have USB mice on the C16/+4! And of course I will release all of it openly, in case!
PS: I have now removed the serial signals from the edge connector and pushed the updated board to GitHub.
|
