Google Groups no longer supports new Usenet posts or subscriptions. Historical content remains viewable.
Dismiss

8375 Agnus question (FAO: former MOS/CSG Commodore Semiconductor Group engineers)

530 views
Skip to first unread message

a1_noc...@hotmail.com

unread,
May 1, 2007, 9:37:39 AM5/1/07
to
I've been trying to find out the details about all of the different
8375 Agnus variations. It is hard going with almost no information out
there.

Does anyone have any information?

Particularly about the differences between the

8375 390544-01
8375 318069-10

They are both PAL and both are 2Mbyte Agnus.

Are they pin compatible?

Any information appreciated. Particularly from former MOS/CSG
engineers.

Alex

Daniele Gratteri

unread,
May 1, 2007, 9:39:59 AM5/1/07
to
a1_noc...@hotmail.com ha scritto:

> Particularly about the differences between the
>
> 8375 390544-01
> 8375 318069-10
>
> They are both PAL and both are 2Mbyte Agnus.
>
> Are they pin compatible?

Yes they are. 390544-01 often has "R0" written on it, while the
318069-10 should be a later revision. Both of them are PAL.

--
___ __
/ __|___ Daniele Gratteri, Italian Commodore-Amiga user... ///
| / |__/ Nickname: FIAT1100D - ICQ: 53943994 Ritmo S75 __ ///
| \__|__\ Home page: http://www.gratteri.tk forever! \\\///
\___| E-MAIL: dan...@gratteri.tk ...since 1990 \///

a1_noc...@hotmail.com

unread,
May 1, 2007, 11:20:57 AM5/1/07
to
> the 318069-10 should be a later revision.

I knew that the date codes on the chips confirm that the 318069 chips
were manufactured later than the 390544 chips.

HOWEVER

I've never seen the same chip on the same revision of motherboard
which made me wonder about pinout.

Are you sure they are pin compatible (before I give it a try)?

Alex

Daniele Gratteri

unread,
May 1, 2007, 12:49:05 PM5/1/07
to
a1_noc...@hotmail.com ha scritto:

> Are you sure they are pin compatible (before I give it a try)?

Sure, I tested 318069-10 on A500 Plus, which originally had 390544-01,
and works just fine.

Clockmeister

unread,
May 1, 2007, 8:54:44 PM5/1/07
to

<a1_noc...@hotmail.com> wrote in message
news:1178032857....@e65g2000hsc.googlegroups.com...

8375 should be pin compatible with eachother but they are not pin compatible
with any other 837x.

a1_noc...@hotmail.com

unread,
May 3, 2007, 6:10:25 AM5/3/07
to
On May 2, 1:54 am, "Clockmeister" <whowh...@andwhy.com> wrote:
> 8375 should be pin compatible with each other but they are not pin compatible
> with any other 837x.

I do not think that is true. Not all variants of the 8375 can possibly
be pin compatible.

8375 318069-16 and 318069-17 is a 1Mbyte Agnus' for A500/A2000/CDTV to
replace the 8372(A).

Alex

Clockmeister

unread,
May 3, 2007, 7:41:15 AM5/3/07
to

<a1_noc...@hotmail.com> wrote in message
news:1178187024....@e65g2000hsc.googlegroups.com...

The 8372A already is a 1MB Agnus.


a1_noc...@hotmail.com

unread,
May 3, 2007, 12:19:19 PM5/3/07
to
On May 3, 12:41 pm, "Clockmeister" <whowh...@andwhy.com> wrote:
> The 8372A already is a 1MB Agnus.

I'm sorry my wording confused you.

The 8375 318069-16 and 318069-17 Angus' are definitely 1Mbyte.

Unlike the other 8375 2Mbyte Agnus chips I think they are a pin
compatible alternative to the 8372A which went out of production.

I think there is at least one other different pinout for the 8375
(beside the 1Mbyte ones), possibly more.

The 318069-19 is a 2Mbyte NTSC agnus which appears to be pin
compatible with the 8372AB (A3000 2Mbyte Agnus)

However I have no proof / pinouts etc. Just a few "hints" collected
across the net. Which is why I wanted to get a CSG engineer who worked
on them to reply.

Alex


Daniel Mandic

unread,
May 3, 2007, 12:55:00 PM5/3/07
to
Clockmeister wrote:

> The 8372A already is a 1MB Agnus.


Where?


Best Regards,

Daniel Mandic

Clockmeister

unread,
May 3, 2007, 6:48:20 PM5/3/07
to

"Daniel Mandic" <daniel...@aon.at> wrote in message
news:463a13e4$1$2298$91ce...@newsreader01.highway.telekom.at...

> Clockmeister wrote:
>
>> The 8372A already is a 1MB Agnus.
>
>
> Where?
>

Everywhere, why?

http://www.amiga-hardware.com/showhardware.cgi?HARDID=1448


Clockmeister

unread,
May 3, 2007, 6:56:35 PM5/3/07
to

<a1_noc...@hotmail.com> wrote in message
news:1178209159.0...@y80g2000hsf.googlegroups.com...

> On May 3, 12:41 pm, "Clockmeister" <whowh...@andwhy.com> wrote:
>> The 8372A already is a 1MB Agnus.
>
> I'm sorry my wording confused you.
>
> The 8375 318069-16 and 318069-17 Angus' are definitely 1Mbyte.
>
> Unlike the other 8375 2Mbyte Agnus chips I think they are a pin
> compatible alternative to the 8372A which went out of production.

That would seem very likely but I haven't seen many, if any around so
perhaps they aren't very common.

> I think there is at least one other different pinout for the 8375
> (beside the 1Mbyte ones), possibly more.
>
> The 318069-19 is a 2Mbyte NTSC agnus which appears to be pin
> compatible with the 8372AB (A3000 2Mbyte Agnus)
>
> However I have no proof / pinouts etc. Just a few "hints" collected
> across the net. Which is why I wanted to get a CSG engineer who worked
> on them to reply.
>
> Alex
>
>

There does seem to be some conflicting info out there.

If you find out which Agnus came from which computer you can check for a
matching pinout from the schematic.

a1_noc...@hotmail.com

unread,
May 4, 2007, 8:14:11 AM5/4/07
to
On May 3, 11:56 pm, "Clockmeister" <whowh...@andwhy.com> wrote:
>
> There does seem to be some conflicting info out there.

I have tried to help improve the info out there but it is still poor.

http://amigahardware.mariomisic.de/cgi-bin/showhardware.cgi?HARDID=1448

> If you find out which Agnus came from which computer you can check for a
> matching pinout from the schematic.

I tried to capture as many photo's of Agnus chips in place taken from
websites and posted by users.

http://www.amiga.org/modules/newbb/viewtopic.php?topic_id=29535&forum=8
http://eab.abime.net/showthread.php?t=21961

I've been trying to get BBoAH to update their list based on my
findings.

Still lots of unanswered questions.

Alex

Daniel Mandic

unread,
May 4, 2007, 1:00:10 PM5/4/07
to
Clockmeister wrote:

Not in a std. AMIGA, AFAIK. You have to patch the hardware to real 1MB
Chipmem, like on an A500 with the later 8372, and even A2000 (rev.4.3).
Just the Rev.6 A2000 is with 1MB Chipmem, as far as I can remember. My
Rev4.3 A2000 (8372) has been patched...

By the way, is the MegiChip (2000 ver.) outfitted with a 8375, or is it
just a Ram extension (+1MB Chipmem)?

Best Regards,

Daniel Mandic


Clockmeister

unread,
May 4, 2007, 4:35:48 PM5/4/07
to

"Daniel Mandic" <daniel...@aon.at> wrote in message
news:463b669a$0$25625$91ce...@newsreader02.highway.telekom.at...

> Clockmeister wrote:
>
>>
>> "Daniel Mandic" <daniel...@aon.at> wrote in message
>> news:463a13e4$1$2298$91ce...@newsreader01.highway.telekom.at...
>> > Clockmeister wrote:
>> >
>> > > The 8372A already is a 1MB Agnus.
>> >
>> >
>> > Where?
>> >
>>
>> Everywhere, why?
>>
>> http://www.amiga-hardware.com/showhardware.cgi?HARDID=1448
>
> Not in a std. AMIGA, AFAIK.

There is no standard Amiga, just revisions.
Later revisions only need a jumper or two changed to obtain 1MB chipRAM
instead of 512K chip/512 fast.

That doesn't change the fact that the 8372A is a 1MB Agnus in either case.


You have to patch the hardware to real 1MB
> Chipmem, like on an A500 with the later 8372, and even A2000 (rev.4.3).

Later A500's and A2000's only need a jumper changed to select 1MB chipRAM
and earlier ones are not pin compatible with the 8372 because the 8372 is
not pin compatible with 8370.

> Just the Rev.6 A2000 is with 1MB Chipmem, as far as I can remember. My
> Rev4.3 A2000 (8372) has been patched...
>

Rev4.3 is an early A2000 that doesn't support the 8372 on the board without
some changes because it was never designed to accommodate it.

> By the way, is the MegiChip (2000 ver.) outfitted with a 8375, or is it
> just a Ram extension (+1MB Chipmem)?
>

It has to be a 2MB part or it won't be 2MB chipRAM, which the 8375 is. Looks
like the 8372B and 8372AB may also be 2MB.


Michael Bourne

unread,
May 4, 2007, 8:20:36 PM5/4/07
to

I have the Megachip kit built and installed in my rev 6.2 A2000. It consists
of a daughter board with PLCC plug plugging into the empty FAT Agnus socket and
a flyer wire to pick up another address wire on Gary. The original Agnus was
an MOS 8372A (1MB) part. On the daughter board (PCB) are 1MB RAM and a 2 MB
Agnus - in this case the 8375B part. There were other candidates, read the
hard/hack docs for this project. These are 84 pin PLCC IC packages.

This results in 2MB chipram:- the original motherboard 1MB and the added 1MB
on the daughterboard plugged into the motherboard PLCC Agnus socket.

: A2000 B6.2 ECS 2MB, GVP-M68060/50, 64MB, SCSI-2 w/ scsi-tower :
: PicassoIV, BuddahIDE, MFC-III, AriadneII, AOS3.9, Picasso96 :
: Dynalink ADSL router, LiNuX RH9 Athlon1.6GHz, Planet Hub, etc :

--
Michael Bourne mfbourne #AT# brisnet |period| org |period| au
Edit the address above; Ignore From: header

Clockmeister

unread,
May 4, 2007, 9:11:15 PM5/4/07
to

"Michael Bourne" <mic...@gallium.mantle.net> wrote in message
news:463bcdd3$1...@news.comindico.com.au...

Yep. Early revision ie 4.x A2000's would end up with 1.5MB chipRAM and 512k
fast without some additional changes to the motherboard, assuming the
Megachip was compatible with those revisions at all.

I have the installation manual here somewhere in my webpage archives.


zipper

unread,
May 5, 2007, 6:48:07 AM5/5/07
to
On May 5, 4:11 am, "Clockmeister" <whowh...@andwhy.com> wrote:
> Yep. Early revision ie 4.x A2000's would end up with 1.5MB chipRAM and 512k
> fast without some additional changes to the motherboard, assuming the
> Megachip was compatible with those revisions at all.
>
> I have the installation manual here somewhere in my webpage archives

Should be (re manual); just alter J101 and cut J500 - but you need
that extra 0.5 MB Chip
to be installed somehow.

Daniel Mandic

unread,
May 5, 2007, 9:47:35 AM5/5/07
to
Clockmeister wrote:

> It has to be a 2MB part or it won't be 2MB chipRAM, which the 8375
> is. Looks like the 8372B and 8372AB may also be 2MB.


You mean the A3000 8372.

That what I meant was a Rev.6 A2000 with continous 1MB ChipMem (I
have). So it must have been a 8375 and the Rev.4.3 holds a 8372A, which
is a 1MB Agnus Type with the typical 512K Chip and 512K Fast (Slow RAM
in fact), by default.

We have still not found an AMIGA with real 1MB Chip, beside the A3000
8372AB (2MB ECS) and the 8375 (A500 plus and later A2000, both with
8375, of course. Also ECS), which is also to be found as 512K Versions
in the A500 plus. An extension would give 1.5MB continous Chip MEM, the
old A501-RAM (512K Slow RAM) would give 1MB continous Chipmem w/o the
slow ram effect (due to 8375, and other Mainboard design with no
slowRAM at all. Kick 2.0 Machines....). Soldering another 512K inside
the A500plus (Rev. 6A) would give the Max. 2MB Chip, what the 8375 can
do with the MegiChip onto the Agnus-Extension Port (+1MB Chip). Rev. 6B
obviusly is the 1MB A500plus. Whereas the A2000 were always with 1MB
Chip (newer Rev's.)

I know you mean the 8372 can 1MB cont. ChipMem, but only on patched
machines...
My A2000 6.0 (codename: A2000D. C was Rev. 4.x, AFAIK) was outfitted
with an ECS Chipset by default, also before the Megichip upgrade!
So, 1MB continous real ChipMEM, due to the 8375 and the new Board
design. And Kick 2.0 by default....

http://www.fh-luh.de/~spider/Amiga/Hacks/Agnus-Typen.txt

Best Regards,

Daniel Mandic

Daniel Mandic

unread,
May 5, 2007, 9:58:09 AM5/5/07
to
Daniel Mandic wrote:

> I know you mean the 8372 can 1MB cont. ChipMem, but only on patched
> machines...
> My A2000 6.0 (codename: A2000D. C was Rev. 4.x, AFAIK) was outfitted
> with an ECS Chipset by default, also before the Megichip upgrade!
> So, 1MB continous real ChipMEM, due to the 8375 and the new Board
> design. And Kick 2.0 by default....
>
>
>
> http://www.fh-luh.de/~spider/Amiga/Hacks/Agnus-Typen.txt

hmmm, it seems the 8372B can 2MB as well, but it's not listed....


http://www.fh-luh.de/~spider/Amiga/Hacks/2MB-Chip1.txt

a1_noc...@hotmail.com

unread,
May 5, 2007, 4:22:45 PM5/5/07
to
> >http://www.fh-luh.de/~spider/Amiga/Hacks/Agnus-Typen.txt
>
> hmmm, it seems the 8372B can 2MB as well, but it's not listed....

That list is old and has errors. Everyone out there seems to use it as
a reference, duplicating the errors across the net.

I'm going to get the Big Book of Amiga Hardware Agnus page updated and
upload lots of examples.

Shame that no Commodore Semiconductor Group engineers read usenet
anymore.

I think the secrets of the different flavours of Amiga chips may have
"died" with them.

Alex


Clockmeister

unread,
May 5, 2007, 10:37:26 PM5/5/07
to

<a1_noc...@hotmail.com> wrote in message
news:1178396565.8...@l77g2000hsb.googlegroups.com...

>> >
>>
>> hmmm, it seems the 8372B can 2MB as well, but it's not listed....
>
> That list is old and has errors. Everyone out there seems to use it as
> a reference, duplicating the errors across the net.
>
> I'm going to get the Big Book of Amiga Hardware Agnus page updated and
> upload lots of examples.
>
> Shame that no Commodore Semiconductor Group engineers read usenet
> anymore.
>
> I think the secrets of the different flavours of Amiga chips may have
> "died" with them.
>

National Amiga used to have a list that seemed fairly complete at the time
but... well, there is interesting stuff there nonetheless.

http://www.titan.co.nz/amigaak/AA060100.htm

or more specifically

http://www.titan.co.nz/amigaak/AA060193.htm

but not much help there.

Dave Haynie is still around and as one of the old hardware engineers he
might be able to help. He's also still fairly active in the C= scene and a
nice guy so try sending him an email.

http://www.frogpondmedia.com/index.html

Just out of interest, why do you need such specific info?

a1_noc...@hotmail.com

unread,
May 8, 2007, 7:34:26 AM5/8/07
to
On May 6, 3:37 am, "Clockmeister" <whowh...@andwhy.com> wrote:
> Just out of interest, why do you need such specific info?

I wanted to put a MiniMega chip in my CDTV. I bought one and it came
with an NTSC 8375 Agnus on it. I wanted to replace it with a PAL 8372B
one but there was no information about which chips were pin compatible
and a lot of conflicting information.

I'd like to dispel the Myths surrounding this chip to help other
people thinking of upgrading.

Plus it is always good to have accurate information about these
things ;-)

Alex


Sam

unread,
Aug 9, 2007, 11:17:42 AM8/9/07
to
Can't say much for 318069-19 but I've tried an 8375B 318069-03 Agnus in my
A3000/ Rev 9/03 board.

I have also read on a German website that this variant was produced to
replace 8372AB and seems to be compatible.

Although it's not proof, I would definately put my money on it based on my
results :)

http://www.jochi.de/speichertabelle.html

Cheers.


<a1_noc...@hotmail.com> wrote in message
news:1178209159.0...@y80g2000hsf.googlegroups.com...

0 new messages