[Coco] MC6809 Inards

jdaggett at gate.net jdaggett at gate.net
Wed Oct 4 12:12:39 EDT 2006

On 4 Oct 2006 at 8:23, Andrew wrote:

> This really does sound interesting. I wonder if maybe Microware or 
> Motorola may be listed for the GIME patent? I know it was Microware 
> which had (and gave to "us" - I can't remember who originally obtained it)
> what we think is a CoCo 3 prototype. There wasn't a GIME on it, and it was
> surmised that the logic of the GIME was done on this prototype in more
> discrete parts. Maybe the prototype could yield clues as to the
> manufacturer/designer of the GIME?


I have searched patents for assignee of Motorola from 1976 to 1993 and can't 
really find any single patent that can be in part or the whole of the GIME chip. I 
have also searched Tandy and Radio Shack and found nothing either. 

I doubt that the GIME chip was designed by Motorola. If it were it would have 
some form of marking on the chip that would indicate so. At minimum the circled 
bat wings (M) emblem would be there. My guess is the GIME chip had no new 
invention but used prior art that they paid royalties for or were in public domain. 
This would result in nothing is a patent search. Haven't tried copyrights. 

>From that era many chips were breadboarded before anyone laid out the design 
on silicon. According to Terry Ritter, the MC6809 was first done on a breadboard 
using LS-TTL logic. Judging from the MC6801 patents, that is very doable.  
Looking at some of the pictures of the Coco 3 prototype, the GIME chip was 
definitely done in LS-TTL logic. Back then compu ter time to simulate logic design 
was very very expensive and very slow. You had to feed the chip design in 
chunks. Most computers were unable to handle the whole design at one time. 

Today a chip with 10 million transistors can be simulated in days or less. Comp 
uters are more powerful and able to take on very large designs and give the 
designer results without  having to build a breadboard. The chip designer then 
has a high degree of confidence that the chip will work with first silicon. In the late 
70's and ealy 80's, sometimes three or four passes were needed to get the chip in 
at least working and shippable state. 


More information about the Coco mailing list