Dumping special format rom

Pagina 3/5
1 | 2 | | 4 | 5

Van Louthrax

Prophet (2280)

afbeelding van Louthrax

05-07-2020, 01:02

Ok, so now we need someone to buy the knitting machine and knit some MSX related stuff at the next MSX (Nijmegen?) meeting Running Naked in a Field of Flowers

Van Takamichi

Champion (361)

afbeelding van Takamichi

05-07-2020, 07:46

As you've seen, both the knitting machine and U-BOX were and are fairly common. There are classes that teach how to hand draw and scan the knitting patterns. A lot of you might know textile machine programming preceded computers ;)
Here's KD2 manual. Page 5 illustrates the MSX2+ based system that has been sold as a set and must have costed 100K+ yen.

One more about KD2 I/O behavior; it ALWAYS asks to insert the memory card when accessing, regardless it's already inserted or not. The card capacity is 32KB as written on its label.
Edit: Added the PCB back photo for clarity. The memory card has 38 pins (photo).

Van gdx

Prophet (3673)

afbeelding van gdx

05-07-2020, 10:02

I did a wiki. You can complete it if you want.
https://www.msx.org/wiki/Silver_Reed_KD2

Van PAC

Guardian (5810)

afbeelding van PAC

05-07-2020, 21:35

Louthrax wrote:

Ok, so now we need someone to buy the knitting machine and knit some MSX related stuff at the next MSX (Nijmegen?) meeting Running Naked in a Field of Flowers

Hahaha... and organize a contest for the most beautiful dress... LOL!

Van NYYRIKKI

Enlighted (5588)

afbeelding van NYYRIKKI

06-07-2020, 02:13

Takamichi wrote:

After the titles screen, does the designing screen with mouse-controlled cursor show up? If so it's fine.

No, I was hopeful, but we are not there yet... It seems there is still lot to work out...
In the HW picture I can see the device has 128kB of ROM and 128kB of RAM (+memory card) After investigating the data, it seems that so far we have managed to dump 96kB of ROM only as these two bits that we have manipulated seems to only select the bank to #8000-#BFFF. It has no effect to memory area #0000-#7FFF. We also have not seen anything that would look like RAM memory.

Before crash the last value written to port 4 seems to be #10... All writes before that use #C as lower nibble and the computer boots from page 3 (on #8000), so the logic could be something like:
bits 0-1: Page select for #0000-#3FFF
bits 2-3: Page select for #4000-#7FFF
bits 4-5: Page select for #8000-#BFFF
bits 6-7: Page select for #C000-#FFFF

... but this is just wild guess right now...

Could you run another set of dumps like this:

BLOAD"SUPER-X.LDR",R
_@
PO7,81H
PO4,00H
S#PAGE0.ROM,0#1,FFFF
PO4,55H
S#PAGE1.ROM,0#1,FFFF
PO4,AAH
S#PAGE2.ROM,0#1,FFFF
PO4,FFH
S#PAGE3.ROM,0#1,FFFF

... let's see what it spits out then...

If you are willing to do some testing on your own, it would not hurt, but I don't blame you if you don't feel like it... Since this hardware is still quite a mystery, it is hard to tell exactly what/how to test... Here are anyway some tips:

If you ie. write "D4000#1" Super-X will give you HEX-dump of the memory area starting from #4000 slot 1... try if you can write it... If you ie. type "12" and cursor moves, but the value does not change then it is ROM, not RAM. ESC-key will get you back to command mode. See how these PO (I/O port out) commands to port 4 affect the data that you see on this address... see also if something changes in address 0 or #C000

You can also use binary numbers to test what the bits do, if it feels more comfortable... so ie. "PO4,10101010B" is equal to "PO4,AAH" and equal to "PO4,170" ... Small help of commands can be seen by pressing GRPH+SPACE, but manual for this software is in the .zip for a reason... With "QT" you can get back to BASIC. I hope these tips help rather than just messes up your head.

The port 7 is also still 100% mystery to us... The value that we use now is written there once early in the boot and then I can't see any routine that would touch it... I think we have not dumped that part of the code yet that uses it.

Van NYYRIKKI

Enlighted (5588)

afbeelding van NYYRIKKI

06-07-2020, 14:06

Ha,

As the HW picture was so blurry, it took me this long to recognize the chip on the left... but hey, I just realized it seems to be our belowed 8255 PPI! This also explains the mystery of #81 written to port 7. That practically defines that I/O ports 4 and 5 are output. Port 6 bits 0-3 are input and bits 4-7 are output... Now we only need to figure out what these ports do.

Van Takamichi

Champion (361)

afbeelding van Takamichi

06-07-2020, 17:07

PO7,81H
PO4,00H
S#PAGE0.ROM,0#1,FFFF
PO4,55H
S#PAGE1.ROM,0#1,FFFF

These worked but both PO4,AAH and PO4,FFH returned Error. So I attach only PAGE0.ROM and PAGE1.ROM. PAGE1.ROM is the "fullest" dump I've seen so far, 0000 to C000H pretty occupied.
After "D4000#1" I tried to type everywhere between 0000 and FFFF but none of them changed, so there probably is no ram. I haven't tested other subslots 1-*. How to select subslot? D0000#1-1/2/3 simply returned Error.

I took precise pictures of the PCB divided to three. 1, 2 and 3 I wanted to upload them to the wiki but had no idea how. Also they are shrunk so here's original sizes. The model name of IC2 (one below IC6) is worn out but it's probably same with IC1; TOSHIBA JAPAN TC54512AP-15 VPP 12.5V 8905EAI

Van Takamichi

Champion (361)

afbeelding van Takamichi

06-07-2020, 16:58

Buying Silver Reed knitting machine in Europe might not be difficult since they were successful in UK and around in ebay UK.
Edit: Another thing, the OCRed KD2 manual can be Google translated. It's crappy but much better than nothing.

Van NYYRIKKI

Enlighted (5588)

afbeelding van NYYRIKKI

06-07-2020, 17:41

Takamichi wrote:

These worked but both PO4,AAH and PO4,FFH returned Error.

Ah, my bad this time... This "old intel standard" requires leading zero, so the correct format is PO4,0AAH and PO4,0FFH

Quote:

How to select subslot? D0000#1-1/2/3 simply returned Error.

Yes, you have correct format, but because this cartridge has no subslots, you can't select them. You can use the "M-S" format only when the slot is really expanded.

Quote:

I took precise pictures of the PCB divided to three.

Great, thanks! I'll look in to them right away!

Van NYYRIKKI

Enlighted (5588)

afbeelding van NYYRIKKI

06-07-2020, 18:09

I tried to look at the pics, but maybe it is more easy if you can just list IC7 pins that don't seem to connect anywhere... This way we can rule out most of the things that we don't need to test further. This chip has 40 pins and the pin numbering makes a little circle so that on bottom right next to little hole mark is pin number 1 then the numbering goes up to pin 20. Pin 21 is on left side top and pin 40 is bottom left next to hole mark again. My problem is that from the bottom side pic I unfortunately could not see any tracks and on top side pic the chip it self hides the connections on left side.

To me it seems that after doing PO7,081H and PO4,0 the memory area between #0000-#7FFF looks like RAM, but naturally I can't say for sure... Can you please test this scenario once again by writing something in there after these commands?

Pagina 3/5
1 | 2 | | 4 | 5