The C128 and Z80 CPU ...or Intel 8088?

[psxphill]

I've never heard that story from either Bil or Dave.

From On The Edge:

To fix the problem, Herd required the C128 to start at memory address zero, but the 8502 started elsewhere. "One night, everybody left and it was broken," says Herd. "During the night, I said, 'I have no way to fix this, unless we startup by not starting at that address.' I said, 'Hey, Von. The Z80 chip starts from zero, doesn't it?' He said, 'Yup.' I said, 'Cool. I need somebody wire wrapping tonight.'"
The hour was too late to purchase a Z80 chip, so Herd looked elsewhere. "Everybody had doorstops that were actually Sinclairs," he recalls. "I went and tore open my doorstop because we didn't own a Z80 chip in the place."


I don't believe they re-used the Z80's in production models though. After hearing about how the C64 was put into production for Christmas however, nothing would surprise me

Putting in an 8088 would likely have not helped much as making the computer MSDOS compatible to any real degree would have been impossible. An NEC V20 would have been better as it could run CPM in it's 8080 mode, but you can also run 8086 software. It lacks the extra Z80 instructions but most CPM software is compatible with 8080 anyway (Z80 is an 8080 clone with added instructions). It might not add much, but you also wouldn't really lose anything.

The C128 was like the A3000. Both were based on custom chips from a previous design, but with extra glue around them because there was no time/money/inclination to improve them. If the C128 had an 80 column VIC that could work when the CPU was running at 2mhz then it would have been a worthwhile upgrade & much closer to the C65 but still C64 compatible.

The C65 was designed during the time when AAA was in development hell, I can understand why in that environment the C65 project would have made some sort of sense and it probably would have sold if it was cheap enough. Another couple of engineers also went off and pushed through the Pandora project, which turned into AGA.

[psxphill]

Thanks for quoting that.  I gave away my copy of On The Edge some time ago and didn't remember exactly what was written.

I still have mine, but I forgot where it was written anyway. I only found it by googling:

c128 z80 doorstop

[psxphill]

This is the way I remember it too. Using an 8088 in the C-128 makes little to no sense at all.

It might have made sense to design it in at the start, although an NEC V20 would make more sense as you could still run CPM on it. However you would want either want to increase the MMU address space or connect it on the other side of the MMU so you could address all the memory from 8088 mode using segments.

The magic voice startup problem was solved because the Z80 doesn't start by fetching the same address as the 6502, but neither does the 8088/V20. So it should have been an equally good workround, except I doubt anyone at commodore was using an 8088 computer as a doorstop at the time.

[psxphill]

Heh, I will reconfirm with Bil.  As Bil and Leonard Tramiel said at CommVEx, the Brian Bagnall book(s) is/are not entirely accurate.

If you watch all of Bil's talks back to back you should detect some inconsistencies. It's no wonder the book is not accurate.

[psxphill]

I was just thinking of what if the 8088 was used instead and what different path MAY have unfolded - had they perhaps made the C128 into some PC compatible too.

It would have been impossible to make it PC compatible. Using the 8088 would also not have solved the issue why the Z80 was put in there to start with. In most of Bil's videos he says it was because the CP/M cartridge exceeded the amp budget and the decision to start the Z80 first instead of the 6502, to fix magic voice autostart, came later.

Bil learnt what C64 compatibility meant when they built the C128. IBM learnt what PC compatibility meant when they built the IBM PCjr. A C128 with an 8088 would have been less compatible than the PCjr, which itself was a failure.

PC compatibility never helped the Amiga either.

CP/M was more suited because of hardware fragmentation, each manufacturer made hardware to their own design and so applications had to go through the OS. While there was some fragmentation in the MS-DOS market to start with (Chuck Peddles Victor/Sirrius for example) by 1985 the market demanded that every piece of CGA/DMA/floppy/hard disk/sound hitting software would work.

An NEC V20 on the other hand would have allowed CPM-80 and CPM-86 applications to run & we could even have seen a port of GEM rather than GEOS.

[psxphill]

Impossible? Nonsense.
My firm made the PT68K4 PC compatible with a V20 based board.
The Amiga could be made PC compatible with a series of boards that topped out with a '386 based board (and those have a much better resale value than PCs based on the same processors).

In 1985 you would need register level CGA/floppy/dma/irq/serial etc. You couldn't reuse anything from the C64, so it would essentially be a PC and a C64 in the same box (like https://en.wikipedia.org/wiki/Amstrad_Mega_PC).

The Amiga at least could display the video from a PC using it's graphics chip, so a combined hardware and software approach allowed it to work. But to make it compatible with games you had to use an actual ISA graphics card. So it was just an Amiga and PC in the same box, it didn't really save any money and you could get better PC's.

Sure you could stick an 8088 in the C128 and it would have access to 64k at a time. No software would know how to display anything on the screen or access more than 64k. The original PC came with 64k so you could run software that worked on that as long as it only used BIOS and DOS calls to access hardware (which a lot of software didn't).

The Z80 came quite late in the C128 design, so there was never a time when it would have made sense to design something radically different with an 8088 that allowed more than 64k at a time.

Of course everything is possible given enough time and money, but keeping to the C128 selling price and spending very little in chip design and it was impossible.

[psxphill]

The C= 128 should never have existed as late as it was (less than 1 year before the Amiga in early 1985). C= should have put a SID chip in the Amiga and paid to have a highly optimized C64+6502 emulator written in 68000 assembler. Perfect upgrade path to the Amiga plus the SID adds new sound synthesizing functionality to the Amiga.

I love SID tunes, but I'm glad that the Amiga had Paula. A full speed C64 emulator with decent compatibility on a 68000 Amiga is impossible..

The C64 lived on for another 7 years, so there was definitely a market for an 8 bit computer. The C128 just turned out too expensive and not enough people cared about C128 or CPM mode.

He and Dave must have felt a constant sense of butt hurt dealing with Commodore's management and their perpetual penny pinching.

I'm not sure that Bil was that bothered about the penny pinching, he seems to quite enjoy the challenge. The problem at that time was marketing, the 8563 chip designer and his boss who cancelled the C128 because it was taking too much time. If Jack had stayed then these problems would have resolved themselves with much less impact to the C128 schedule.

Without Bil you would have got this instead http://www.floodgap.com/retrobits/ckb/secret/d128.html An evolution of the P500 that was dropped because commodore needed as many VIC and SID chips as they could make to stuff in C64's.

The C128 with just a z80 would have been a disaster. After the C128 shipped someone worked on an upgraded design that could run the Z80 faster. I don't know if that ever survived and whether it would be practical as a retro-fit.

[psxphill]

I would have continued to manufacture the C64 until demand dropped too much. Developing and marketing the C128 was a waste of resources which could have been used for the Amiga.

According to google, in 1985 the C128 was $300 and the A1000 was $1295.

There was room in the market. If anything they should have stopped selling the C64 and cost reduced the C128 (which they started doing but then cancelled it and kept selling cost reduced C64's instead).

[psxphill]

I am not sure I would not have favored a 2MHz 6509 with no Z-80.

Accessing memory with the 6509 was a nightmare, the C128 can run the 8502 at 2mhz so you didn't lose anything. Ideally they should have made an 80 column VIC. It would need to fetch multiple bytes per clock so that the CPU didn't get starved of bus activity. But at least then you could run faster than 1mhz without turning off the screen.

A 6809 would have at least doubled the price. I like the 68000 but it's the only Motorola CPU that is really any good, if it hadn't been so good then commodore wouldn't have paid the premium.