[coco] 6309 speed

[jdaggett at gate.net]

Gene 

First off you are determining the speed of the processor by observation of the rise 
and fall times of the CMOS output pair of transistors. That tells you nothing of the 
switching characteristics of the internal CMOS transistors nor propogation of a 
signal through the many levels of N and P  MOS transistors that make up logic 
gates within the IC. 

All CMOS processes are not equal. Switcing characteristics are determined by load 
capacitance and on resistances (Rds) of the CMOS output devices. Generally 
speaking the geometries of the CMOS output transistors are slightly larger than 
those used internal to the die. This allows the outputs to sink and source more 
current and handle larger capacitive loads. 

What limited the old 4000 series of CMOS in their speed was that rather large gate 
capacitance and the large ON resistances of the Drain/Source channel (Rds) of the 
NMOS and PMOS transistors internal the the devices. This was the early years of 
CMOS. And they were compared to later CMOS processes that improved 
geometrries and well construction that allowed CMOS to move incredibly faster. 

Lastly what I also caution is that just because one 6309 runs at say 20 or 30 MHz 
does not necessarily mean all will. There will be a frequency where each individual 
IC will cease to function properly. Two ICs from the same wafer, one could crap out 
at 5MHz and the other at 12MHz. There is no way to tell unless you test each 
individual part. By specing the 63C09, Hitachi is guaranteeing that the IC will 
operate to 3 MHz. That is all they test to. Any higher and that is a bonus. ICs vary 
from their position on the wafer and from wafer lot to wafer lot. Statistically I would 
say there is a frequency around 5 or 6 MHz where say 70 to 80% of all will work to. 
As one pushes the frequency higher, less and less will work properly. FInally you will 
get to a frequency where less than 1% of all ICs  will work properly.

james

On 3 Aug 2005 at 22:23, Gene Heskett wrote:

Date sent:      	Wed, 03 Aug 2005 22:23:35 -0400
From:           	Gene Heskett <gene.heskett at verizon.net>
Subject:        	Re: [coco] 6309 speed
To:             	CoCoList for Color Computer Enthusiasts 
<coco at maltedmedia.com>
Organization:   	None, usuallly detectable by casual observers
Send reply to:  	CoCoList for Color Computer Enthusiasts 
<coco at maltedmedia.com>
	<mailto:coco-
request at maltedmedia.com?subject=unsubscribe>
	<mailto:coco-
request at maltedmedia.com?subject=subscribe>

> The 63C09's that I have can muster up edges on most pins as fast as my
> scope can see, 10 ns as its only a 100mhz dual trace.  That sort of
> speed isn't old time plain cmos by quite a long row of apple trees.