Page 3 of 3

Re: Texas DS990/1 Dump

Posted: Thu Jul 16, 2020 10:24 pm
by pieroandreini
Thanks David, you are right. I discovered that DS990/1 use signals DIR and SIDE mixed together. As you said and described on Qume DT/8 manual, the signal SIDE may be use on different ways. I'm gonna try to change drive jumpers to factory setup

Re: Texas DS990/1 Dump

Posted: Fri Jul 17, 2020 8:33 am
by pieroandreini

Re: Texas DS990/1 Dump

Posted: Fri Jul 17, 2020 7:33 pm
by ZrX
Ok this time the header shows the head/side flag changing between sides on every track. But the bit order of that value is weird, as in the byte value is 0 for first head and 8 for the second head. All the other header values are in a typical order.

Also, in all the previous dumps the head value was 0 for both sides of the first track and showed another value for all the other tracks.

Re: Texas DS990/1 Dump

Posted: Sat Jul 18, 2020 12:28 am
by pieroandreini
@ZrX What could this strange bit order for the heads depend on?

Re: Texas DS990/1 Dump

Posted: Sat Jul 18, 2020 10:49 am
by brightcaster
Nice to see this progress! Which tool did you use to convert from RAW to DSK?

Shouldn't the DSK file be much bigger if there are 2 sides, 77 tracks and 26 sectors with 288 bytes each? According to my calculation it should be about 1.2 MByte (including the overhead)!

David

Re: Texas DS990/1 Dump

Posted: Sun Jul 19, 2020 11:55 am
by pieroandreini
A really nice person made a tool for this, I ask him the permit to share it. I guess it's ok but for respect I will ask before to share his software
The firs version of the tool had some issue, that's why dimension is just Half.

Here the last dump, still some bad sector

https://www.dropbox.com/s/eq8n2r8pdgcl8 ... d.img?dl=0
https://www.dropbox.com/s/y0x1vi6zyuimi ... d.log?dl=0
https://www.dropbox.com/s/7695818a0x8j1 ... d.zip?dl=0

May be really useful to add a profile with this sector encoding on the kryoflux software to dump Texas Instruments floppy disk

Re: Texas DS990/1 Dump

Posted: Sun Jul 19, 2020 7:54 pm
by ZrX
The tool is my experimental tool I use to manage and study odd disk formats.

The FD1000 MFM manual states "The head address is in bits 1 through 4 and the cylinder address is in bits 9 through 15. Bits 5 through 8 are always zero." Yet the graph shows HEAD to use bits 0 through 4. But it doesn't specify what order the bits on HEAD value should be.

So it could be they just used bit 3 for the upper head, or it could be any of the bits could be used for the upper head on the floppies.

Re: Texas DS990/1 Dump

Posted: Sun Jul 19, 2020 8:27 pm
by pieroandreini
FD1000 don't have any cpu or rom, even the controller on DS990 is made just with ttl logic, no custom or whatever, so should be the operating system that manage this, right?
If this is correct this information should be found on operating systems technical manual I guess

Re: Texas DS990/1 Dump

Posted: Mon Jul 20, 2020 8:48 am
by brightcaster
@pieroandreini:

Only a few floppy disk drives were equipped with their own CPU (mainly CBM). So, yes, it should be a question of the operating system, the kernel, or whatever it is called on that specific system ...

@ ZrX:

Ahh, I see it's your tool again. I wondered if anyone else had gone that far! I also haven't seen anything to read this strange sector size ...

David

Re: Texas DS990/1 Dump

Posted: Mon Jul 27, 2020 8:16 pm
by pieroandreini
Thanks to everybody for the big support