[Coco] GIME chip data collection

[jdaggett at gate.net]

Nick

I need a bit more info to help you:

On all the chips the coding following the 6xx/7xx/8xx remain the same?

"V CO30"


Also the second line do they remain the same across all variants? If they are different 
how many variants are there?


This is my best guess as to what info is placed on the chip. 

the 6xx/7xx/8xx more likely is not a date code. I doubt that it is a fab code either. Three 
digit date code followed the convention of first digit of the date code is the last digit of the 
year and thenext two digits would be the week. An example would be 745. That would 
corresspond to xxx7 week 45. The weakness of this is the decade and century 
information is lost unless that is stamped elsewhere. Then that would be duplicate info 
and space is limited on the top surface.

The next five digits "V CO30", could be fabrication facility and/or mask set. each time 
you create a mask set for a particular IC you have to identify the many mask layers. An 
IC can have upwards to 25 or more mask layers to fabricate a single wafer. Any change  
to one layer would generate a new mask set ID. These two were often three to five digits 
with a combination of numbers and letters. 

The actual part number: TCC1014 any improvements to it or revisions for problems or 
even a die shrink would change this part number by addign a revision letter after it. Thus 
you get

TCC1014
TCC1014A
TCC1014x

Then you have the coustomer's name and copyright symbol and year made followed by 
country of origin and the last letter is anyone's guess. 

Interesting that you mention Brother Jeremy's Prototype Coco3 and the markings of that 
GIME chip. Robert Gault and I have had a conversation about the GIME chip off line and 
one point I made to him about the prototype board is that by the date codes I see on the 
chips, the earliest that particular board could have been made was late June/ July time 
frame of 1985!! The ltest date code I see, with a quick glance, on any chip on the 
prototype board was 8523. Twenty-third week of 1985. This would lead me to believe that 
a first engineering run of the GIME chip could have been late 1985 or very very early 
1986, say January/February time frame. The fact that you mentioned the part number as 
TCC1014-Tequila makes me believe that Br Jeremy's GIME is a preproduction chip. 
Very well that chip I mentioned about to Robert. 

Saying this is not in passing. The prototype board has two 74LS189 chips. These are 16 
bit by 4 bot static RAMs. Two chips would make a 16 byte SRAM.exactly the size of the 
color palette RAM. Now here is the interesting part. There is also a TBP28S42 chip on 
the protoype board. This is a 512 byte SRAM. Yes a 512 byte SRAM. Looking back at the 
Tandy R&D memo at items #5, this could be that realization. Furthermore Br. Jeremy's 
Coco3 may be the only one out here in the world that can do the "256 color" mode. If so 
then it would be displayed in the 160x200 resolution mode max. 



hope this helps some

james


On 19 Jun 2009 at 19:41, Nick Marentes wrote:

> Thanks to those who have submitted their GIME chip data so far.
> 
> I've been adding the info into a spreadsheet and I can see one trend
> so far.
> 
> The first number, possibly a batch number has come in three variants.
> 
> A 6xx number which appears to be batches of Korean '86 chips.
> 
> A 7xx number which appears to be batches of Mexican '86 chips.
> 
> A 8xx number which appears to be batches of Korean '87 chips.
> 
> Does this first digit signify a batch? Maybe a version number?
> 
> Someone once told me that there were at least 7 revisions of the GIME
> and that the '86 and '87 were the last and final 2 releases. We all
> know that the '86 GIME has more bugs than the '87 so could there be
> various versions of the '86 with different fixes?
> 
> Also the chip "model number" is TCC1014 on the Korean '86 chips
> whereas it is TCC1014A on the Mexican '86 and Korean '87 chips. Maybe
> the Korean '86 chips are the buggy ones while the other '86 and the
> '87 are fixed?
> 
> Interestingly, Brother Jeremy's prototype production CoCo3 so far has
> the lowest 6xx number (614) and the "model number" seems to have a
> name .... TCC1014-TEQUILA.
> 
> Frank thinks that first number is a batch/date code. Can anyone shed
> more light on these GIME chip markings?
> 
> A typical chip has these numbers...
> 
> 629 V C030
> 2645A0001
> TCC1014
> (c) TANDY 1986
> KOREA A
> 
> There also seems to be two manufacturer logos... VTi and VLSI. Can
> anyone explain what they stand for or even who and where these
> companies are?
> 
> I will post the completed list after a few weeks after more data is
> gathered so keep it coming and I'll keep the spreadsheet up to date.
> 
> Nick Marentes
> 
> --
> Coco mailing list
> Coco at maltedmedia.com
> http://five.pairlist.net/mailman/listinfo/coco