DPF-550 Floppy drive | MSX Resource Center (Página 4/5)

Por Manuel

Ascended (19463)

12-07-2012, 10:35

msxholder: thanks! Then something is wrong with the emulation still.

Por zeilemaker54

Champion (347)

12-07-2012, 20:00

@Manuel, I will take a look at it....

Por zeilemaker54

Champion (347)

12-07-2012, 21:04

@Manuel, about the AVT-DPF550:
I/O 0D0H-0D3H r/w WD1793 or WD2793
I/O 0D4H r b7 IRQ (1 = IRQ)
b6 DRQ (0 = DRQ)
I/O 0D5H w b0 drive 0 select (1 = select drive 0)
b1 drive 1 select (1 = select drive 1)
b2 side select (1 = side 1)

Is this what you like to know ?

Por Manuel

Ascended (19463)

12-07-2012, 22:21

I was hoping you had an idea why I get "File not found" where I would have expected "DIsk offline"...

Por Manuel

Ascended (19463)

14-07-2012, 11:30

msxholder wrote:

The drive is always A: because I/O configuration
THERE can't be 2 interfaces of avt or daewoo used in one msx.

I remembered this remark when I saw MSX Technical Databook page 42: "The FDC may be placed in the I/O area; however, it must have a mechanism to disable it, and it must be enabled only if the system does access to the FDC. This makes it possible for the system to have more than one FDC interface for handling different media types."

Anyway, an extra dump of the AVT DPF-550 would surely be helpful, msxholder. And also if you have the Daewoo, because that one is also different, if I'm not mistaken...

Por Manuel

Ascended (19463)

14-07-2012, 22:48

Update on AVT DPF-550 emulation: we found the Disk offline missing issue. To quote Wouter:

aha the WD1770 doesn't have a READY input pin (instead that pin is a motor-control output pin)
the more important thing is that we need to emulate the timing of the read/write sector command in more detail
the thing is that when a sector is read that won't be found, we currently emulate that by immediately generating an error, while in reality the disk has to rotate and the it takes up to 5 rotations before the FDC decides the sector really is missing
so on successful reads we currently already emulate the correct timing, but not yet for errors
and the disk-offline detection in this AVT diskROM relies on this extra time (on failure)

So, emulating this should fix it... we'll see in the coming days