[s]Could you use a 65c02 in a C64 with some tricks?[/s]

Edit: nm, I forgot I allready asked the same question: viewtopic.php?f=4&t=2085&start=15#p61776

[s]And does any1 know where the rights for VIC-II and SID are?[/s]

What would be more interesting; how could you interface from a 65c02 to a SPI screen?

_________________
MOS

Can I ask, What is the main issue here? Do you have a C-64 that you want to get working again? or Maybe modify one that does work already?

Not directly answering your question here but there was someone out there years ago that made a SID 'clone' called the SwinSID.

_________________
65Org16:https://github.com/ElEctric-EyE/verilog-6502

Technically yes, and all the hardware tricks we happen to know already went listed\described in that other thread of yours.

Problem is, that the 65C02 doesn't have the NMOS 6510 "illegal" instructions,
also some subtle details like stray reads are different for 65C02 and 6510.

Means, after you successfully replaced a 6510 with a 65C02 in a C64,
the C64 Basic + Kernal will work, a lot of the C64 software will work too,
but expect that some of the programs and games who make use of the "illegal" 6510 instructions
and\or the 6510 stray reads etc. will hang/crash.


The patents for the chips expired, the original plans for the chips went lost,
and I can't recall the name of the company that had bought the company that had bought Commodore...
Unless you want to re_start commercial production of the original chips based on the original plans,
copyright isn't a problem for the hardware.


#define 'SPI screen'.
If you are out to interface a C64 to a VGA monitor or such, you need a scan doubler,
because horizontal frequency of VGA is about twice the horizontal frequency of the video signal generated by the VIC-II.
If the scan doubler comes with a DVI\HDMI converter, it can drive a DVI\HDMI monitor.

Hope, this helps.

I want to learn in a way that is practically useful, which means in a way where I could repair a C64 without requiring old components while still not emulating it in a FPGA or in software. Also having lot's of software to test/make the system useful immediately.


SwinSID looks great, thx!


That's ok, software that uses illegal instructions will fail.


I want to understand the VIC enough to build my own GPU that does not have the legacy of PAL/NTSC but instead uses SPI directly, the first step in this direction is to build a VIC that uses SPI to a small IPS LED screen.

I'll contact the creator of SwinSID and see if he can make a SwinVIC!

_________________
MOS

The displays page of the 6502 primer has lots of links that will interest you, including ones to VGA and similar displays that have a serial input, mostly through a 6551 or other UART, and some through SPI and other interfaces. There's a short demo video of a small monochrome SPI-interfaced one I have on the workbench. I intend to eventually get a bigger one with more resolution, although I myself don't really have any need for color. At the bottom of the page there are lots of links to video topics on this forum too. The 6502 primer addresses many aspects of making your own 6502 computer. Take advantage.

You used the [s]...[/s] tag, which is meaningless on this forum. What did you intend?

_________________
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?

Some more C64 related reading material that might be (or not might be) interesting.
We have quite a collection:

FPGA64 > VHDL 6567 > sort of a "TTL styled schematic"
TTL VIC-II, some ideas
Understanding the SID
C64 PLA, and TTL address decoding
C64 clock generation
TTL 6502 Here I come

Are we sure a 65C02 can be dropped in place of the 6510? Even with modifications? The 6510 also had a bunch of I/O pins added, IIRC. I can't imagine they added all of that for something that wasn't needed.

Perhaps a small board and CPLD could be designed to replace all of that. But then you still have the illegal op codes like already mentioned.

_________________
Cat; the other white meat.

On the "hardware side", it requires a little PCB with some aditional chips, Ruud Baltissen had replaced a 6510 with a 65C816.
The problem is more at the "software side"...


Not to forget stray reads etc. and instruction cycle count.
If you are aiming for 99.99% compatibility, better use a FPGA containing the CPU too.

The C64 depends very strongly on its specialised video hardware, the VIC-II. There is no really good way to replicate this if the display is on the far side of an SPI bus.

I feel the same way. Many of the early games for the C64 were very simple. Basic sprites, drawing chars, etc. But it didn't take long for programmers to find exploits and flaws in the VIC-II. You would have to reproduce those flaws if you are shooting for high accuracy.

In fact, I don't think it's ever been 100% reproduced. Even in FPGA's. I could be wrong about this one. But I think people have tried to reproduce the flaws by emulating them. Which isn't the same thing.

In my opinion, a drop-in replacement of the VIC-II using a small FPGA board is the Holy Grail of retro goodness. A close second would be the VIC (1) in the VIC-20. The SID has been emulated pretty well over the years.

Shoot...the TED in the Plus/4 would be nice too. lol

_________________
Cat; the other white meat.

There is Visual6502 with a netlist extracted from the NMOS 6502 silicon,
there are transistor level schematics for the NMOS 6502,
and we have reasons to believe that the 6510 core isn't different from the NMOS 6502.

There is a transistor level dissection for all revisions of the SID chip.

There is a transistor level dissection of the address decoder PLA silicon.

But IMHO the microscopic pictures of the VIC-II are not good enough for starting a transistor level dissection of the VIC-II.
Somebody please prove me wrong by posting a link to a polygonized picture of all layers in the VIC-II silicon.

...Maybe we would need a transistor level dissection of the 6526, too.

OK, this is where my inexperience shows, but if we have a "transistor level dissection" of an IC like the SID, then what are the next steps for someone to run with that and start producing new ones?

I assume cost is the big one as a chip-fab is not cheap. What else?

_________________
Cat; the other white meat.

This is a question (or set of questions) which keeps coming up.

Some answers in this recent blog post:
https://chips4makers.io/blog/startup-co ... uring.html

(By Fatsie, here, I think, judging by this thread:
viewtopic.php?f=4&t=5261
)

What makes things worse is, that the SID is a mixed signal chip (analog + digital stuff).

We don't have any info about the MOS manufacturing process (yet),
and of course the new chips would have to be done with a new and different manufacturing process,
because the old machines and the MOS fab are gone for good.

Different manufacturing process means, the transistors in the new chips will have different electrical parameters,
transfer curves etc. like the transistors in the old chips.

To make it work, you need to figure out the parameters of the old transistors, the parameters of the new transistors,
and then to make a complete redesign of the analog part of the chip just to make it work like in the old chips.

This probably won't work well at first try, and every new try would cost quite some money...
By the time you eventually got it working, the fab probably would have switched to a newer and more different
manufacturing process already.

Clear text: if you want your mixed signal chip to stay in production for more than a few months,
_constantly_ adapting it for the next process to come would be a full time job, probably for more than just one engineer.

SID is certainly the difficult one, because it's mixed-signal, and it's audio, which people are very sensitive to. Indeed, the fact that there are different generations of SID, which sound different, to the point that some projects allow for switching between the two, shows that even Commodore gave up on the idea of making something identical.

If it takes a crowd-funding of $20k to do a custom chip on the chip, you're going to need a lot of die-hard customers who reject FPGAs to make it worthwhile. It's like making luxury handbags or something - it's well past the utilitarian and well into something else.

FPGAs are hardware too, and they have so many advantages. Custom chips are a solution to two kinds of market: where cost is no concern, and where volume is really high.