Slot and Memory mapper on SVI 738 DIY to MSX2/MSX2+

Page 3/3
1 | 2 |

By NYYRIKKI

Enlighted (5396)

NYYRIKKI's picture

10-04-2019, 19:38

lkpalwa wrote:

NYYRIKKI any ideas or problems as you see on this

I believe you are now fixing the part that is not broken...

How I see what you should do would be mounting an extra LS133 on back of IC41, add A2 to pin 2... Connect pin 6 to "ADDRESSOUT" (IC52) and the result in pin 9 back as new "ADDRESSOUT"

I must warn you though that this is all just untested speculation...

Once you have the slot select signals fixed you can swap the slots around as you like.

By lkpalwa

Expert (92)

lkpalwa's picture

10-04-2019, 20:47

Okej Smile but I try to keep it simple, my idea was to keep the design for slot 3-0, 3-1 as it was and just make a work around for slot signal 3-2 and 3-3 according to the expert3 design the A14, A15 togeather with D0-D7
so when 1Y-2Y on LS154 is LL - SLTX0 is H, HL - SLTX1, LH -SLT X2 and HH - SLT X3
On the LS153 A14 and A15 will handle which gate is select Y or 2Y and the D0-D3 in combination will apply when A or B on LS139 is selected .. so that is my though ... as I can see IC41 in SVI is handle to handle DISKROM and RS232ROM
Or are I am wrong ? oO oO oO

By lkpalwa

Expert (92)

lkpalwa's picture

10-04-2019, 21:01

because the ULA ship IC52 is mysterius, can find really a datasheeat that explain to 100% how it works and the design behind - it seems like a 8255 with some extra logic for some more signal if I compare with PPI desing for expert3 schematics
but I can be 100% sure that is the case Smile

By lkpalwa

Expert (92)

lkpalwa's picture

10-04-2019, 21:10

the idea is that the extra LS139 togeather with LS153 will only handle the SLT3-2 and SLT3-3 signal so pins 12 and 11 will
not be connected at all and the SLT3 signal that comes in is from IC35 where it handle by D0-D1 and A0-A1, A14-A15.
where also the PPISEL is handle by IC34.

By NYYRIKKI

Enlighted (5396)

NYYRIKKI's picture

11-04-2019, 11:24

You must have this somehow messed up in your head as your plan makes no sense to me... If you look expert3 schematics, your problem is not on the right side of the picture, but on the left side of the picture...

You miss the point that the 1Y and 2Y signals that you get out of LS153 in expert3 picture are same as EXTSLTA and EXSLTB you get from IC52 in SVI, so adding such is useless... You also can't add that since you would be taking value from databus instead of RAM. You also miss that the signal coming to LS139 EN pin is broken... You can't connect that directly to SLT3 as that would disable output regardless of what address is read or write.

Practically to my eyes the problem on SVI is that the FFFF is not FFFF although the schematics say so. As we don't have IC52 datasheet, we can't read it from there, but we can see that it CAN'T BE complete FFFF signal as the address pins can't get properly decoded as they are NOT even CONNECTED to anywhere.

I think only thing you need to do is to make FFFF real FFFF and all your slot decoding problems disappear from all slots as the subslot select register is not anymore all over the place on page 3.

By NYYRIKKI

Enlighted (5396)

NYYRIKKI's picture

11-04-2019, 11:22

Maybe this helps you a bit to understand the big picture:
In expert3 schematics circuits U48, U49 and U50A make up the address decoding. (FFFF)
Circuits U50B, U50C, U55, U56A and U56B make up an 1 byte read write RAM memory that is attached to the decoded address.
Content of this 1 byte RAM is then fed in to the sub slot select signals (by U57) and decoded if we are not actually reading or writing that 1 byte RAM address (U58B)

Inside SVI IC52 there must be logically U50B, U50C, U55, U56A, U56B and U57

By lkpalwa

Expert (92)

lkpalwa's picture

11-04-2019, 12:27

sorry, that really make sense NYYRIKKI thx Smile will try to make a new analyse oO oO oO
as the big german guy say I will be back Tongue

By lkpalwa

Expert (92)

lkpalwa's picture

11-04-2019, 16:22

Ok I think I will try this Smile IC52 (on SVI pdf schematics) is on socket, so I will disconnet PIN 40 ADDRESSOUT, and add U48 + U48 + U50A, that way I 100% to get the correct FFFF signal and connect to IC42 pin 5 (on SVI pdf schematics)
that way IC63 should now give all SL3-X signal and SLT30, SLT31 are working as before. Thanks for the hint NYYRIKKI Smile
(I use 2 x LS30 + LS32 because I have them at home, but your idea of LS133 should also work, but I think this is more safe)
Any suggestions ?? oO oO oO oO

By NYYRIKKI

Enlighted (5396)

NYYRIKKI's picture

11-04-2019, 16:55

lkpalwa wrote:

IC63 should now give all SL3-X signal and SLT30, SLT31 are working as before.

No, these SLT30 and SLT31 will be fixed as well... The problem is that none of the SLT3-X slots work correctly above #C000

Quote:

Any suggestions ?? oO oO oO oO

Now this finally sounds like a proper plan... I would say go for it.

By LS120

Expert (98)

LS120's picture

16-06-2019, 09:10

how you going with this so far?

Page 3/3
1 | 2 |