6502

That was partly what I was thinking, but looking at this model, it has ROM and no external interfaces.....

I think TI has a Z80 cored bigger model, that has some user access, and I also see this

formatting link
which is 'not ready yet', and whilst it is very impressive, it may struggle to hit critical mass... Might be better if they did a software with optional keypad for the merchant PDAs that are out there.

-jg

Reply to
Jim Granville
Loading thread data ...

Not quite. The Hitachi HD6309 was the best 8-bit CPU ever made, because it was a functionally enhanced CMOS version of the 6809.

Reply to
Eric Smith

Well, you don't really need TWO whole index registers, do you? If I absolutely had to pare down an 6502, I'd get rid of the X index register. The "(FOO),Y" addressing mode is much more useful than the "(FOO,X)" mode.

Reply to
Eric Smith

Why do you need X AND Y ? Sheer frippery, one will suffice.

Reply to
Lewin A.R.W. Edwards

Nobody seems to have yet mentioned the Renesas 740 series (

formatting link
), which is instruction compatible (bugs and all!) with the 6502.

There's also a whole range of peripheral options, but I don't know what the availability would be for hobbyists.

Ed

Reply to
Ed Beroset

formatting link

Renesas availability sucks for the 740 core..... They are pushing their H8 and other 16-bit + processors.

I used the M38049 for a data logger project..... sweet.... 6502 compatable and flash.....

Tony

Reply to
The6502man

formatting link

Reply to
bogax

Why do you need index registers at all ?

A few bytes of free RAM for a self modifying table access routine should be enough.

When we start eliminating registers, why not eliminating stack pointers at the same time. There has been a few processors without a stack pointer, so this should not be a problem.

Paul

Reply to
Paul Keinanen

You don't need addressable registers at all. A stack machine gets along quite nicely with a SP and IP alone.

--
A: Because it fouls the order in which people normally read text.
Q: Why is top-posting such a bad thing?
 Click to see the full signature
Reply to
CBFalconer

Aaargh! I was afraid someone would say this. The enhancements don't change the architecture, and the technology (CMOS instead of NMOS) even less so. The design is pure Motorola, so that's where the kudos goes.

Reply to
steven

I seem to recall that the 6309 had a few more instructions and some other extras either by setting a bit or something. Memory it the first thing .... oh I forget. ;-)

I love the 6809, it was the first assembly language that looked like C (dare I call it high level?). Then I got hit with the 68000 which I hated (more because of all the extra support chips than anything else).

--
Linux Home Automation         Neil Cherry        ncherry@comcast.net
http://home.comcast.net/~ncherry/               (Text only)
 Click to see the full signature
Reply to
Neil Cherry

And C looked like PDP-11 assembly language.

--
Jim McGinnis
Reply to
Jim McGinnis

I wrote:

They most certainly do! They add new registers and new addressing modes. If it wasn't a new architecture, you'd be able to execute HD6309 code on a 6809, which you usually can't. The DEFINITION of computer architecture (as established by Gerrit Blauuw and Fred Brooks in the early 1960s) is the programmer-visible characteristics of the system. The 6809 and HD6309 definitely does NOT have an identical set of programmer-visibile characteristics.

True. In the Blauuw and Brooks taxonomy, that would be a matter of "realization" (if the architecture and logic-level implementation were otherwise identical).

The base design is by Motorola, but if you're trying to give them exclusive credit for the HD6309 design I would say you're way off base. The Hitachi designers did a non trivial amount of work extending the architecture.

Reply to
Eric Smith

If Real Men were required to give up one index register, they would prefer to give up the X register, as the Y register has better addressing modes.

Reply to
Eric Smith

That's the whole point. It's a tradeoff. At some point deleting the hardware for a register costs you more (due to increased RAM or ROM size) than it saves.

Removing the stack pointer from the 6502 would probably be much more expensive (in increased RAM and ROM size) than removing one of the index registers.

It may well be the case that deleting one index register from the 6502 architecture is cost effective for some applications. However, this would depend on the applications and the implementation technology. If you made this decision today, you might find out that it is no longer a win next year.

I doubt that anyone in this newsgroup has done a detailed analysis of the cost-effectiveness of removing an index register from the 6502, and most of us don't have access to nearly enough data to be able to do so. Sunplus may well have done such an analysis.

Reply to
Eric Smith

formatting link

Availability of Mitsubishi parts is terrible even for small to medium sized businesses. I worked for a company that used the 740 series parts in multiple products, and dealing with Mitsubishi was a nightmare. It was difficult to get them to honor the terms of the contracts their own lawyers wrote.

On the other hand, the same company had no trouble dealing with Hitachi on H8/300 family parts. Since Renesas is the merger of the Hitachi and Mitsubishi semiconductor divisions, I have no idea what to expect from dealings with them now.

Apple used 740 series parts in several products, and I never heard of them having any trouble, but of course Apple was buying them in much larger quantities than my employer.

Reply to
Eric Smith

Loss of the Y register was a problem as soon as Sunplus made a chip with more than 256 bytes of RAM. I wrote a speech recognition application for one such chip, and it required "creative" programming to use the higher-addressed RAM.

--
Jim McGinnis
Reply to
Jim McGinnis

No kidding. About a year ago I was looking at the m16c parts. You couldn't pry information out of what-used-to-be Mitsubishi if your life depended on it. Toolchain availability was awful, documentation was a nightmare, and there didn't seem to be any user community at all.

Same here. Documention is orders of magnatude better, there were all sorts of free development tools, the local Renesas FAE still doesn't return e-mails or phone calls, but there's enough other info and support out there that it doesn't matter.

Lets just hope the Misubishi division doesn't pull the Hitachi division down to their level.

--
Grant Edwards                   grante             Yow!  I was in a HOT
                                  at               TUB! I was NORMAL! I was
 Click to see the full signature
Reply to
Grant Edwards

... snip ...

As they did in extending the Z80 to the 64180, which they sold back to Zilog later and became the Z180. If that had existed in

1979 the 8086/8 architecture and the IBM PC would probably never have surfaced, since the 64180 could handle 1 Meg of memory. Not as conveniently as the 8086 architecture, though.
--
Some similarities between GWB and Mussolini:
a) The strut;  b) Making war until brought up short:
 Click to see the full signature
Reply to
CBFalconer

Not a problem. Most of those addressing modes have been ripped out as well. When you are shipping 100,000 toys per day, cutting the cost by a penny or two is more important than a few registers, adressing modes, and instructions. :)

Reply to
Guy Macon

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.