[Coco] Re: Any reason to put a 6309 in a Coco2?

[jdaggett at gate.net]

Alex

I am not trying to be negative or the like. I just find it very hard to believe that on IC 
design, the designer just puts extra circuitry on for the heck of it. I do know a little 
about the process to make ICs. I can tell you a product manager will catch all heck 
if a wafer lot of 20 to 30 wafers bombs because the designer sticks some extra 
circuits in "just " to take up space. There was an intent and for some unknown 
reason, Hitiachi would not official recognize the new features. All we can do at best 
now is specualte as to why!

It is very much possible that the HD63x09 series is CMOS though I more likely 
believe HCMOS than CMOS. HCMOS takes advantage of higher density and thus 
better speed improvements.Hitachi using microcode instead of discrete state 
machines does offer a savings in the die real estate. It also allows for more registers 
and instruction more easily. To add a new instruction you just expand the 
microprogram and rom. Registers are added by just making the microcode wider. 

I still hold to my belief that Hitachi was intending to "one up" Motorola but at some 
point decided not to. Maybe because after further review they would loose to much 
to do so. Thus the part was released and the interanls was kept secret. What I have 
not been able to determine is when did the HD63x09 series start to appear? That 
more than likely will help in determining if Hitachi used CMOS or HCMOS process. 

One other interesting tidbit. In the 1983 Databook Motorola pre announced an 
HCMOS version of the 68x09E. By then Hitachi was already making the 6301/3 
which was Hitachi's version in CMOS or HCMOS and were also producing the 
HD68x09 series probably in HMOS which is High density NMOS, A cousin to 
CMOS. I do know that corporations will cross license technology. It is a way of 
buying technology not with cash but with technology. It allows one to quickly get into 
a market without the long time frame of developeing technology and processes on 
your own. Kind of jumping over growing pains. That is Motorola licensed the 
instructions of several of its processors in exchange maybe for some technology? 
Maybe HCMOS? The HD63x09 was to be a product to introduce after the 
technology swap cease date was mutually agreed on. Then Hitachi would release 
the info and a improved version of the 6809. Along came another deal and they had 
to put a lid on it. By 1990 the 8 bit market for home PCs was dead and buried. 
Except for games and embeded use the 8 bit processors were relagated to history. 
The HC05 and HC11 were becoming favorites in the hobby and industrrial embeded 
market.

james 

On 31 Dec 2003 at 20:04, Alex wrote:

> On Dec 31, 2003, at 6:42 PM, jdaggett at gate.net wrote:
> 
> > Hitachi in my opinon would not put circuitry on a die unless they
> > intended to use it. Believe me real estate on a die is precisous and
> > expensive. No one in the industry puts circuits on a die to just put
> > them on. If my memory serves me correct I think that the number 8 in
> > the orient is not one of the lucky numbers. I vaguely remember that
> > the number 4 is real unlucky.
> 
> I am just passing on some of what was in the original posts about the
> H6309 on the coco list.  My understanding from it is that the H6309 is
> a microcode implementation of the 6809 in CMOS.  The implementation
> left some free space, and the additional instructions were added by
> the engineers either 1) without Hitachi's knowledge, or 2) with
> Hitachi's knowledge, but that they remained undocumented because such
> an upgrade was in violation of Hitachi's license from Motorola.  As a
> Chinese linguist I can tell you that eight is only less than lucky
> because it is a multiple of four.   Four is unlucky because it sounds
> like death.
> 
> 
> -- 
> Coco mailing list
> Coco at maltedmedia.com
> http://five.pairlist.net/mailman/listinfo/coco