KMK/JZ IDEa Technical Help

flashjazzcat · April 12, 2012

That makes me want to get mine going even more!

Developmental FAT32 code is just about done. Here's a not very exciting video of the directory of a FAT partition (on a CF card plugged into a SIDE cart) being listed under SpartaDOS X:

http://youtu.be/i7fOUYSFT2Q

Since we now have access to the FAT partition, I'll probably write programs to copy data from Atari partitions and back again: nothing too elaborate. A FAT32 driver for - for example - SDX seems an unlikely prospect (however tantalising) given we already designed a partition table system which is totally incompatible with the PC. But some limited cross-talk between FAT and Atari partitions is doable, and ATR mounting will also be useful.

Edited April 12, 2012 by flashjazzcat

flashjazzcat · March 22, 2013

Just finished heavily revising the APT IDEa BIOS so it's in line with the current APT spec (suddenly I have a need for a hard disk interface with no extraneous features), and it worked first time in Altirra:

I'll put this ROM up on the website in the next few days when I've made a couple of tweaks, although I'm not sure how many interested parties are still present. While the PBI ROM obviously lacks the advanced features of the Incognito/Ultimate PBI (dynamic ATR/partition mounting, etc) owing to the lack of space and the slightly inflexible RAM banking, it still sports slave drive support and pretty much everything else in the APT spec - inside under 2KB of code. Drive mounting's probably quite doable as well. Anyway - it's a nice little diversion from weightier projects.

Edited March 22, 2013 by flashjazzcat

+Larry · March 22, 2013

Hi Jon-

Can't remember -- is the IDEa rom different than the regular original KMK-JZ IDE rom?

-Larry

flashjazzcat · March 22, 2013

Can't remember -- is the IDEa rom different than the regular original KMK-JZ IDE rom?

Yeah - this one's banked but I'm only using a couple of hundred bytes of the second ROM bank (and none of the RAM), so I might be able to squeeze out a KMK-JZ version too.

Edited March 22, 2013 by flashjazzcat

flashjazzcat · March 23, 2013

Master / Slave operation works:

http://www.youtube.com/watch?v=JrFsDN1XthE

Might fit this inside the 1200XL after all... works great alongside the U1MB / SIDE PBI.

Driver here, for anyone brave and in possession of an EPROM/EEPROM flasher:

http://www.atari8.co.uk/apt/idea/

Edited March 23, 2013 by flashjazzcat

+Larry · March 25, 2013

Hi Jon-

Not trying to twist your arm on the original IDE, since I suspect that not too many folks use the IDE/IDEa since the IDE+2 came out. But wasn't Drac030's bios Version 2.x also banked? I can't remember the exact differences, but some of the later interfaces have more ram than others? And I think that the IDEa has the larger ram like the later IDE?

-Larry

flashjazzcat · March 25, 2013

But wasn't Drac030's bios Version 2.x also banked?

Yeah - he brought out a banked version of the IDEa ROM which is freely available (although he complained it wasn't too stable for some reason: it worked well for me). What's your point?

I can't remember the exact differences, but some of the later interfaces have more ram than others? And I think that the IDEa has the larger ram like the later IDE?

The interface I have here has 2 x 1.5KB ROM banks, and 2 x 0.5KB RAM banks. The RAM (unfortunately) banks in unison with the ROM, so code in bank 2 can't see variables in RAM bank 1 and vice versa. I don't use the second RAM bank at all, although I use the second bank of ROM.

So what are you asking for? An un-banked version which works with the original interface?

+Larry · March 25, 2013

Hi Jon-

Nope, not asking for anything. I interpreted your previous response to mean that the banked rom would be an issue in trying to adapt your IDEa rom to IDE-original. Since I remembered that Drac030's 2-series bios was banked that didn't seem to make sense. Just trying to understand the differences, really. No big deal at all.

-Larry

flashjazzcat · March 25, 2013

No problem. Probably had crossed wires. The point with this APT BIOS is that it will indeed be tricky to get it down to an unbanked 1.5KB to fit the original KMK/JZ. That's just because APT implementations necessarily have a somewhat richer API than their predecessors, even at a basic level.

flashjazzcat · March 30, 2013

Fixed a bug in the physical disk status command (which failed to return an error if a master/slave IDE device wasn't present), and this led me to implement something I've been thinking about doing with FDISK for a while:

This shows IDEa connected on PBI #1 with master and slave disks, and U1MB/SIDE also present on PBI #3. Soft-drivers (which work via the SDX SIO vector) are also detected. Trouble is, scanning all the PBI handlers takes about 15 seconds while the SIO times out for absent devices, so I need to try and minimise this so it's not an inconvenience when several PBI devices aren't all connected at once.

EDIT: D'Oh... the obvious thing to do was simply step through the device selection bits first and look for PBI signature bytes at $D803 and $D80B, thus creating a bitmask of available PBI handlers before issuing SIO calls. Scanning now only takes 1-2 seconds.

http://youtu.be/ADP2Wx3cM-k

Edited March 30, 2013 by flashjazzcat