BLITTER revisions ?

[exxos]

While looking at the recovered ASCI diagrams, there are 2 pdf called 4021 and 4051. It looks like the 4051 had the "fixes" for the issues Atari found on the MEGA ST (basically lots of double inverters on some lines to clamp noise). I won't go into all that again..

BUT, there is another doc 2 years later, 4082, which looks to have more logic blocks.

1.JPG (137.6 KiB) Viewed 3910 times

2.JPG (140.47 KiB) Viewed 3910 times

Also some logic seems to be optimised / missing. Even if this file was the "prototype" for the STE combo chip blitter, which could explain some logic differences, but the addition of what looks like the entire databus being latched for some reason...

So it begs the question, was more features added to the STE blitter, or even STE's combo which isn't present on the original blitter chips, or what other reason would there be for all this extra logic relating to the blitter ? Hard to think they are huge bug fixes.. but never know..

[DrF]

I got a question I think

If there are different revisions of the blitter chips out there in the wild and lets say X game or Y demo is programmed around these bugs/fixes what ever, then surely that means there's a small % of STs where said software would break or not work if Atari fixed/moved/bodged something?

[exxos]

I got a question I think

If there are different revisions of the blitter chips out there in the wild and lets say X game or Y demo is programmed around these bugs/fixes what ever, then surely that means there's a small % of STs where said software would break or not work if Atari fixed/moved/bodged something?

There are revisions, but these are just more workarounds to other problems (noise on the bus etc) Its why older blitters don't work good in the MEGA ST and needed to patch board etc. So theres no problem really as such.

But the "newest" blitter design, has 5 more pages of stuff.. which the original blitters didn't have... its almost like it has more functions.. but it could be a prototype for another machine totally...

[DrF]

I got a question I think

If there are different revisions of the blitter chips out there in the wild and lets say X game or Y demo is programmed around these bugs/fixes what ever, then surely that means there's a small % of STs where said software would break or not work if Atari fixed/moved/bodged something?

There are revisions, but these are just more workarounds to other problems (noise on the bus etc) Its why older blitters don't work good in the MEGA ST and needed to patch board etc. So theres no problem really as such.

But the "newest" blitter design, has 5 more pages of stuff.. which the original blitters didn't have... its almost like it has more functions.. but it could be a prototype for another machine totally...

I was wondering that, I guess the software sees the revisions as all the same, it don't really know, all it knows is its a blitter?
And these are hardware workarounds for all the other bodges on the thing

But the "newest" blitter design, has 5 more pages of stuff.. which the original blitters didn't have... its almost like it has more functions.. but it could be a prototype for another machine totally...

Or cheaper

[exxos]

Or cheaper

Could be, or just really re-vamped logic. It's hard to compare it page for page..

[exxos]

It looks like it might be the "half tone" RAM block. Because that seems to missing from the new version others. But there also looks like other stuff has extra latched buffers over the original.

EDIT:

I am going to make the assumption that as 4081 is called ST MCU / GLUE COMBO, and 4082 BLT CELL, That its the file which later went into the STE's COMBO for the integrated blitter combo. It could explain all the various latchings of data for read and write I guess as its all done internally on one chip.

EDIT2:

Oddly the 4081ORIG file mentions "ST GAME COMBO" 1989 so maybe its not even a STE chipset.

[czietz]

BUT, there is another doc 2 years later, 4082

The files are dated end of 1986 (4021), 1988 (4051) and 1989 (4082). Rather conspicuously the GE-manufactured Blitters are marked "(C) Atari 1987", the ST-manufactured Blitters are marked "(C) Atari 1988" and the IMP-manufactured Blitters are marked "(C) Atari 1989". Hmm?

Oddly the 4081ORIG file mentions "ST GAME COMBO" 1989 so maybe its not even a STE chipset.

It is known that Atari wanted to use the STE's combo chip in an ultimately cancelled game console codenamed "Robin". Remnants of that project are inside each STE, see the ominous "GAMECART" register that I found.

[exxos]

The files are dated end of 1986 (4021), 1988 (4051) and 1989 (4082). Rather conspicuously the GE-manufactured Blitters are marked "(C) Atari 1987", the ST-manufactured Blitters are marked "(C) Atari 1988" and the IMP-manufactured Blitters are marked "(C) Atari 1989". Hmm?

National seem to be 88 - 91 from what I can see. I think 86-88 was just "bug fixes" with noise issues, mostly relating to the MEGA ST. But also somewhere around 88 seems to be the turning point https://www.exxosforum.co.uk/forum/viewt ... 062#p22062

So we pretty much know the 4051 blitters were in production by various manufactures somewhere in 1988.

Oddly the 4081ORIG file mentions "ST GAME COMBO" 1989 so maybe its not even a STE chipset.

It is known that Atari wanted to use the STE's combo chip in an ultimately cancelled game console codenamed "Robin". Remnants of that project are inside each STE, see the ominous "GAMECART" register that I found.

Is that register part of the blitter though ? If that were the case, the later national chips would have something new in them.

I think as before though, that the 4082 is the BLIT CELL for the STE combo, or some other similar or future chipset. I don't think there is any new features in the 4082 blitter, but there seems to be a huge chunk of changes. Though I am assuming its optimisations for inclusion in some combo chip more than anything.

[czietz]

Is that register part of the blitter though ?

No. It isn't. It's in the GSTMCU (Glue & MCU) part of that chip.