Kaypro 10 & CoPower 88

[ktmchrisv]

Ok, I have my system somewhat back together. I'm currently on the Universal board with 81-478a ROM. I am using the MSDOS.COM and RAMDISK.COM from http://www.retroarchive.org/maslin/d...o/copwrcpm.td0
and booting MSDOS disk http://www.retroarchive.org/maslin/d...o/copwrdos.td0.
The system can also boot the MS-DOS 2.11 disk at http://www.zorba.z80.de/files/swp/dos211-1.td0 However, with this disk, I get some errors in the protocol running between MSDOS.COM and the firmware on the COPOWER 88. You can demonstrate this type of failure by running basica.com of the disk from zorba.
If I remember correctly, the DOS 3.3 disk from Zorba also works: http://www.zorba.z80.de/files/swp/dos330.td0, but, not the CPM86 disk that is there.
I am using CPM 2.2u1 when running the utilities, and not the CPM that are on the disks.
The port number used in these programs is 07Eh. This was the result of my disassembly. I'm including pictures of my board, where you can see where I cut a trace to get my board to respond. Please see the 2nd picture.
IMG_0793.jpg
IMG_0794.jpg
IMG_0795.jpg

Here is a picture of my system booting the Zorba Dos 2.11 disk. However, the Maslin disk works better.
IMG_0792.jpg

So far, the MSDOS.COM off of the K10 disk you sent only seems to hang. But, upon hex comparison, it has a lot more data than the file that I retrieved off the Maslin archive. I'm still looking into how to get this file working / running. More to come.

Unfortunately, I accidentally used a CPM 2.2F PUTSYS.COM command which trashed the hard disk under CPM 2.2u1 and I have to rebuild the hard disk again.....So,

That is what I have for now. I'm going to work on getting the K10 MSDOS.COM to work. I might switch over to the 81-302 board to see if it has something to do with the BIOS / Board.

Let me know if you have any questions.

Chris V.

[shirsch]

Interesting. I'm going to try duplicating your results step-by-step. Did you start by using Teledisk to write a floppy, or did you convert to IMD (or BIN) and extract the files? IIRC, I was unable to write a floppy from the copwrcpm.td0 image. Your success with 7E and 7F is particularly intriguing considering that I couldn't even get the RAMDISK.COM program to work at that port address.

[ktmchrisv]

Interesting. I'm going to try duplicating your results step-by-step. Did you start by using Teledisk to write a floppy, or did you convert to IMD (or BIN) and extract the files? IIRC, I was unable to write a floppy from the copwrcpm.td0 image. Your success with 7E and 7F is particularly intriguing considering that I couldn't even get the RAMDISK.COM program to work at that port address.

I typically convert to IMD with the TD02IMD command from Dunfield. Since I blew away my installation last night, I recreated a K10 set of H autoload disks, but with CPM 2.2u1 loaded on the system tracks, and with utilities in USER 1. I copied the MSDOS.COM and RAMDISK.COM into disk1, so now the whole system will set itself up with the COPOWER support.

I have found Teledisk to provide inconsistent results between the many versions that are out there. Especially with that K10COPWR disk on MASLIN. Each version of Teledisk seems to output a different disk with different bad sectors.......translate I use IMD for consistency.... The problems with Teledisk is consistent with TD V1.4 and V1.5 being the version of TD that ripped the original disk. YMMV.

Back to work. I'll try and update more tonight.

Chris V.

[ktmchrisv]

I experimented last night with my 81-302 board and have duplicated your results of having massive screen garbage on execution of both MSDOS.COM and RAMDISK.COM from the COPOWR.TD0 disk. RAMDISK.COM works with the screen garbage on initialization and when it is writing to the disk, but not when it is reading. This is with the board strapped to higher order nibble port bits of 7 and addressed as 7E,7F. MSDOS.COM from K10COPWR just hangs on my system when run at default. I decompiled MSDOS.COM to figure out that the port address for addressing the board is set at address 102,103 and is set to 1B and 1A respectively. I put a header on my board, so I could change the decode of the high order nibble. I changed the high order bits to decode on 01 and ran MSDOS.COM. Hang. I then re-strapped to 7 and loaded in DDT and changed 102 and 103 to 7F and 7E respectively. Repeatedly prints a bunch of garbage on the screen; but, tries to do something....i.e. the 81-302 board communicates with the COPOWER board.

When thinking about the hardware design, the issue really is the lower order nibble A and B of port 1A and 1B. The board will select when using E and F and not A and B. The experiments prove that the hardware is decoding beyond the high order nibble and the LSB. I'm going to review the schematic available to see if this observation matches up with the schematic we have. I'm worrying that there may be a specific CPU board for the 81-302 motherboard. Because, right now, my board cannot decode the standard address being used by the MSDOS.COM and RAMDISK.COM on the K10 disk....... The issue could also be a custom ROM implementation on the COPOWER 88 main board itself.....i.e. the firmware on that board needs to match the protocol used by RAMDISk.COM and MSDOS.COM.......

Still thinking about this.

Chris V.

[shirsch]

Hi, Chris. Do you have the schematics? If not, they are attached here. I was able to setup for decode of 1A and 1B by doing this:

• Pull U4 and U6, then install sockets in those locations
• Bend out pin 2 of U6 and jumper to pin 1
• Bend out pin 13 of U6 and run a jumper to pin 32 of the Z80 header that connects to motherboard
• Bend out pin 12 of U4 and jumper to pin 12 of U6 (you can pick this up at an adjacent via if you like)
• Jumper only A4 on the address selection pins (I mounted a Berg header there)

[shirsch]

I experimented last night with my 81-302 board and have duplicated your results of having massive screen garbage on execution of both MSDOS.COM and RAMDISK.COM from the COPOWR.TD0 disk. RAMDISK.COM works with the screen garbage on initialization and when it is writing to the disk, but not when it is reading. This is with the board strapped to higher order nibble port bits of 7 and addressed as 7E,7F. MSDOS.COM from K10COPWR just hangs on my system when run at default.

I'm very confused as to which exact software does work for you and what the port settings are for that successful use.

[tehotaone]

Amazing work... love the project!

[ktmchrisv]

I'm very confused as to which exact software does work for you and what the port settings are for that successful use.

What works for me:
My K10 Universal board with rom 81-478a using the MSDOS.COM and RAMDISK.COM from http://www.retroarchive.org/maslin/d...o/copwrcpm.td0
and booting MSDOS disk http://www.retroarchive.org/maslin/d...o/copwrdos.td0.
My board is strapped to 7 for the high order bits.

This combination works fairly well. It does not work on my older board with ROM 81-302.

BTW, I started a disassembly of the MSDOS.COM from the K10COPWR disk you sent me. The bad news is that while the port address is stored at offset 102 (1B and 1A), there are entire portions of the code that are hard coded to particular port addresses; so changing the ports at 102 and 103 will not work.

BTW, I am out of town until next Wednesday; so, I will be not replying for a while.

Chris V.

[shirsch]

BTW, I started a disassembly of the MSDOS.COM from the K10COPWR disk you sent me. The bad news is that while the port address is stored at offset 102 (1B and 1A), there are entire portions of the code that are hard coded to particular port addresses; so changing the ports at 102 and 103 will not work.

When you are back I'd appreciate a copy of that disassembly. I started this exercise a while back, but I really am not good at it. Always suspected there were hard-coded port addresses lurking somewhere and I have a feeling that is part of my problems. Port 7E and 7F has some sort of conflict on the 81-302 ROM (and TurboROM apparently).

[ktmchrisv]

When you are back I'd appreciate a copy of that disassembly. I started this exercise a while back, but I really am not good at it. Always suspected there were hard-coded port addresses lurking somewhere and I have a feeling that is part of my problems. Port 7E and 7F has some sort of conflict on the 81-302 ROM (and TurboROM apparently).

Hi shirsch,

Here is the disassembly. I did some more work on it today. There is a dispatch routine at 13E8 that includes hard coded references to IO Port 1C/1D, where the code earlier in the file uses indirect addressing of the port using the addresses at 102, which are 1A/1B. Unfortunately, all of these ports are not able to be decoded by the hardware; which is using an AND gate on the lower order address bits B3-B1. I can't help not believing that this code was meant to address a different board than we both have.....Let me know what you think. Also, I'm thinking that their may be some embedded 8088 code which is transferred over; so some of the disassembly may be completely garbage....

Thank you,

Chris V.