Panasonic FS-A1GT 1024Kb RAM upgrade (or 1MB)

Página 3/4
1 | 2 | | 4

Por OeiOeiVogeltje

Paragon (1426)

imagem de OeiOeiVogeltje

16-02-2010, 15:52

the GT Ray2day bought at the MVM auction has 1 MB (Hapzee's GT) and i know it has a Pbc with a sim in it too

maybe i can convince ray2day to open his GT up one time

Por Repair-Bas

Paragon (1165)

imagem de Repair-Bas

16-02-2010, 15:57

Please oeioei, do not open that Turbo

I have schematics how to built in that kind of extension.
Not online, only on paper.
If you want I can make a copy for you, but is not possible anymore to built in that kind of memory on that way.

Por ray2day

Paladin (727)

imagem de ray2day

16-02-2010, 16:26

What's wrong with my new GT??

Por Gradius2

Hero (652)

imagem de Gradius2

16-02-2010, 17:24

So its really worth to upgrade or not (GT to 1024Kb) ?

Because the only useful things I can see is for MSXDOS2 and SymbOS.

Por RetroTechie

Paragon (1563)

imagem de RetroTechie

16-02-2010, 17:42

S1990      A15   A14   A12
            L     L     L     64Kbytes
            L     L     H    128Kbytes
            L     H     H    256Kbytes (ST)
            L     H     L    512Kbytes (GT)
            H     L     -    1Mbyte (see note 1)

R800        A15   A14
             L     L         64Kb  * 4-bit
             L     H         256Kb * 4-bit (ST/GT)
             H     L         1M    * 4-bit (see note 1)
             H     H         4M    * 4-bit

Source:
http://www.mail-archive.com/msx@stack.nl/msg07024.html

Really the only significant (and interesting, I admit) info in this thread sofar. For the R800 settings, it makes sense: I suppose it simply sets the size (no. of bits) of the refresh address counter. Normally this equals number of multiplexed DRAM address lines (there are exceptions). And possibly it also informs the R800 what size a DRAM page is. So that it knows when it crosses into a new DRAM page (=new RAS cycle), or stays within the page (=no new RAS cycle needed).

But what do those S-1990 settings do? Set how many mapper address lines are controlled? Determine how many bits of mapper registers can be read back (through read from I/O ports FC~FFh) ? And where does this info come from ultimately? I can't imagine that engineers that put together TurboR design, or manufacturers producing custum IC's for it, didn't have datasheets of some kind. How come these still haven't surfaced? Or: if people (in the MSX community) actually have copies of such documents, why are they sitting on them? What company has anything to lose from us knowing detailed function (not internal design!) of R800 / S-1990 chips? :-? For me this is 1 reason I never bought a TurboR - I just don't like owning black box hardware.

So only 512Kb can be used and not 1024Kb! and when I use ALLOCATE USER SEGMENT function of the MSXDOS2, it seems to work!
So that's a non-functional 1M expansion - in other words: useless :hannibal:

Here is the diagram:
http://i48.tinypic.com/ws0ld.png

Anyone else notice that schematic has an error? On the LS32, #IORQ0 and #RD0 should be swapped. Also most connections between LS367 and LS670 make no sense: if they're not coming from, or going to outside signals, it's better+easier to connect such data inputs to GND or +5. Routing them back & forth gets you nothing.

I wouldn't put much confidence in a source that can't even draw its own schematics correctly... :murdock:

Por Gradius2

Hero (652)

imagem de Gradius2

16-02-2010, 18:03

Anyone else notice that schematic has an error? On the LS32, #IORQ0 and #RD0 should be swapped. Also most connections between LS367 and LS670 make no sense: if they're not coming from, or going to outside signals, it's better+easier to connect such data inputs to GND or +5. Routing them back & forth gets you nothing.

I wouldn't put much confidence in a source that can't even draw its own schematics correctly... Murdoch

That diagram have been passed by Jipe, I just put them into understandable format (into serevals PCTs to a single png).

So far, it seems the pdf is the right one.

Por Jipe

Paragon (1594)

imagem de Jipe

16-02-2010, 18:48

the last schematic (make with Dynamic Publisher on MSX ) is the copy of intern schematic GT buy in Deutchland with just a develloper board

and yes /iorq0 and /rd0 are inverted - is correct on my paper version with annotation

see on ST PDF is the same schematic

on GT schematic are optimized by Slayerman use internal /E8 signal for decode adress but is same function

Por OeiOeiVogeltje

Paragon (1426)

imagem de OeiOeiVogeltje

16-02-2010, 22:40

well

thats nice

is there some way to test if all programs actually work in this thing?

Por RetroTechie

Paragon (1563)

imagem de RetroTechie

16-02-2010, 23:30

AFAIK that mapper test from Digital KC is pretty good - if it shows "no problems", then other software should work fine (or cause is some other issue). And if you have a problem with a program (and it's because mapper doesn't work 100%), then mapper test will usually tell you what's wrong. There's at least 1 other mapper test program that I know off. And IIRC, stable DOS2 operation is also a good indicator for how well the mapper works.

Página 3/4
1 | 2 | | 4