stable, tested 6502 core

I'm looking for a stable, tested 6502 core

does one exist?

I'm looking for freebies or pay-to-gets

Rich

Reply to
aiiadict
Loading thread data ...

see previous post

formatting link

The 6502 core there is not bad - it is cycle and operation accurate for most stuff -and it's free.

Reply to
MikeJ

This unit for sure contains a 6502 hdl core: (by Jeri_Ellsworth)

formatting link

Same girl, another product:

formatting link

She proberbly have a 6502 core somewhere.

Here's another:

formatting link

Try 6520 vhdl in google..

Reply to
pbdelete

I have googled and found many broken, inaccurate cores.

R
Reply to
aiiadict

Hi,

Both the T65 from Opencores and Free6502 from the Free-Ip Project (its been a while that the guy took it offline, but you can still get it from fpgaarcade) are stable and rather small in terms of logic resources (30% of xc3s200). We have used them in a good five projects. Each one has their own pros and cons. The T65 has a well defined synchronous interface but tries to mimic the real 6502, signals are inverted and so on. The free6502 on the other hand, has a good interface too but lacks a ready input signal. For people that wants to run it at maximum speed (around 40 MHZ, 10 mips), its tricky to make it go into wait states (to handle slower peripherals like an RTL8019AS chip), while the T65 has that mechanism and will allow itself to be idle...

In all, both are very usable and well-defined soft processors suitable for FPGA implementation.

Jacques

formatting link

snipped-for-privacy@gmail.com wrote:

Reply to
jaxato

Daniel Wallner's 6502 core at opencores.org

I concur with your findings.

As my first boss used to say, starting projects is easy - finishing them is not.

If you have a project like building a vintage car or aeroplane, other people are only impressed when you can turn the key and take off. Or if they can see you making good progress in the project. If it is started briskly then left for ages, then chances are you got stuck and don't have the skills to complete it.

Likewise with reproductions of old CPU cores. We can all start one, but until it is fully working it is nothing to brag about.

In contrast, Daniel does seem to see the job through. So much so that I notice places that used to sell Z80 and 6502 cores as IP no longer bother.

I would like to use his CPU designs in my own projects, but mine now use more of my FPGA than his CPU designs do. So for now I just stick with external (and still cheap and readily available) CPU chips.

Reply to
Kryten

FYI - the T65 from opencores has known bugs. For example ADC (),Y is broken. You would be better off getting the T65 from fpgaarcade

formatting link
where this and some other issues have been addressed. I don't know why the opencores version hasn't been fixed yet. Maybe the author lost interest in it.

I'm pretty sure that the free6502 (now only available at

formatting link
- maybe you were thinking of the patched T65 at fpgaarcade?) has problems. I didn't track down the exact issue, but if you whip together an Apple 1 like system (we're talking a few hours work here if you substitute a UART for the video/keyboard interface) and run integer basic then try running a basic program like Hammurabi from the Replica1 site, and you'll see that it has problems. Just too much trouble to track down the problem. It's an interesting core, so would be nice to see it fixed.

Another possibility is the core at

formatting link
which seemed to work well in my limited testing. (The BusCycles feature is a little different - I guess that this came from the VIC/C64 world.)

The main problem is that none of these cores include documentation or verification suites. I would plan on being able to switch different cores in and out if you run into problems. Kind of a pain since they are all a little different, but might save pain during development.

Regards, Keith

Reply to
Keith

It might be a fun project to run two cores in lock step running the same program and see how far they get before they get a different result.

--
The suespammers.org mail server is located in California.  So are all my
other mailboxes.  Please do not send unsolicited bulk e-mail or unsolicited
 Click to see the full signature
Reply to
Hal Murray

I maintain the cores on fpgaarcade. That trick is how I found most of the problems in the T80 core, and is the reason it is more accurate than the T65 core which I have never got around to doing the same thing. I think there are some pictures on the pacman page on the website showing the real Z80 being run alongside.

When running the pacman code, the real Z80 and the T80 core stay in lock-step indefinately, which I am quite happy with.

formatting link

/MikeJ

Reply to
MikeJ

ElectronDepot website is not affiliated with any of the manufacturers or service providers discussed here. All logos and trade names are the property of their respective owners.