status register on the TMS9918A and the V9938/58

Pagina 1/6
| 2 | 3 | 4 | 5 | 6

Door ARTRAG

Enlighted (6234)

afbeelding van ARTRAG

20-11-2014, 21:51

As you may know the lower 5 bits of the status register, when the 5s flag is reset, assume undocumented values.

This rom
https://sites.google.com/site/devmsx/home/split-screen-on-ms...
shows as background color the value of the 5 lower bits on the status register starting from the collision among sprites on planes 26 and 27. On emulators I see a large white band (corresponding to 1Fh in the status registers).
Could anyone test this rom on a real msx1 or on a real msx2 and say what color he gets and the machine used in the test? The result could be used to improve the emulators and virtually track the raster almost line by line on the screen.

Aangemeld of registreer om reacties te plaatsen

Van hit9918

Prophet (2866)

afbeelding van hit9918

21-11-2014, 01:52

I need casette, the tools need 16k ROM.

Van ARTRAG

Enlighted (6234)

afbeelding van ARTRAG

21-11-2014, 08:30

Here it is:

https://sites.google.com/site/devmsx/home/split-screen-on-ms...

run it from basic with bload"splitdem.bin",r

Does anyone see a yellow background instead of the white one?

Van hit9918

Prophet (2866)

afbeelding van hit9918

21-11-2014, 13:28

The answer is yes and no Smile
The zone is flashing in multiple colors.
I think I see black white yellow green.

But why is the split below the colliding smileys? Is there a 5th sprite involved?

Van ARTRAG

Enlighted (6234)

afbeelding van ARTRAG

21-11-2014, 16:16

No 5s involved or you would fall into the documented behavior
Any difference with msx2?

Van Manuel

Ascended (15680)

afbeelding van Manuel

21-11-2014, 21:45

Running it on a turboR: it almost looks the same as when you run it in openMSX (R800 mode): there is at least one extra line of sprite visible and below that there is a blinking line.

Running on real MSX2 with V9958: exactly the same as on openMSX, except that one more line is shown. Only the bottom line of the smiley is missing, whilst on openMSX the bottom 2 lines are missing.
There is no blinking.

Van ARTRAG

Enlighted (6234)

afbeelding van ARTRAG

21-11-2014, 23:03

Sorry, what is relevant is only the color of the area (in the border) that on openmsx and on bluemsx appears as white and on real HW I expect yellow. The number of lines missing to the set of sprites on top is not relevant.
@hit9918
Can you confirm that the flashing area is the one you see in white on emulators?
Which machine are you using in your tests?
@manuel
Can you confirm that you see a white area in borders starting from the screen split (the same you see on emulators)?
Do you see it only on V9938 and V9958 ? What about msx1 machines?
Openmsx shows the same 1Fh value on TMS9918A and V99X8, but according to hit9918's evidences, this should be wrong.

Van Manuel

Ascended (15680)

afbeelding van Manuel

21-11-2014, 23:16

It *is* relevant, because it's different! It shouldn't be! Tongue

Anyway, on my turboR and MSX2 it was white, like on openMSX. I didn't try an MSX1 yet.

Van ARTRAG

Enlighted (6234)

afbeelding van ARTRAG

21-11-2014, 23:28

The difference you spotted has to do with the time when the register sets the collision flag.

I was focusing on another thing: the color of the border should show the differences in the content of the last 5 bits of the status register.

Van ARTRAG

Enlighted (6234)

afbeelding van ARTRAG

22-11-2014, 09:09

Could it be a way to find a difference with Toshiba

Van ARTRAG

Enlighted (6234)

afbeelding van ARTRAG

22-11-2014, 10:23

Actually it could be a good way to spot toshiba chips against tms9918a and co...
Anyone willing to try?

Pagina 1/6
| 2 | 3 | 4 | 5 | 6