floppy data separator (re)build woes

[Dokken]

edit. this separator is the klyball v2.

I've been trying to get my klyball data separator with gotek working for a couple of weeks. the frustrating thing is I had this functioning for a time before I had some other failures. I now have my C4P working, a glitch RAM board, serial works, all seems to work fine (I actually have two working C4Ps).

I have several floppy images from Mark which are known to work. I had them working on my system. Now I cannot get It working again.

Possibly an issue setting the timing? I don't really understand what I'm doing other than looking for a negative 6ms signal on 121 pin 1. also seems the positive 6ms should be on pin 6?.

reset D will cause the gotek display to stutter and look like it is doing something. then nothing.

There has to be something stupid I'm doing. if anyone has any troubleshooting tips, additional test points, etc. I'd be grateful. thanks for any thoughts or obvious places to check.

[Dokken]

it turns out this does work. kind of. with one image that Dave made for me. A floppy disk test program image. this will load and run from the klyball gotek. I also built up another floppy interface which behaves the same.

The floppy test program will run, read/write all 40 tracks, and seems to function as it should. so I'm fairly confident my floppy adapters are ok.

however, other images do not seem to ever get past track 13. No other images I have will fully load. they all seem to try, but stop.

My systems are C4PMF (505 revB, 540 Rev 1b, and a glitch RAM board).

I was told "that Pressing ‘D’ loads the bootstrap from Trk 0 to RAM address $2200. It then passes control to the software it just loaded to load Trk 1 to RAM address $2A00. Depending on what disk you are loading ‘BASIC, Assembler and Extended Monitor’. It is normal for it to stop at TRK 13 and then transfer control to BASIC and Load “BEXEC*”

Is this a memory issue/glitch ram config? Will my board configuration only boot with certain images? I now have serial access and can load the memory test program. I do get X at some locations.

[bxdanny]

Yes, track 0 loads track 1, and together they then load BASIC from tracks 2 through 6. The disk head should then skip ahead to track 12, where it first loads an overlay to BASIC from sector 3, then reads sector 1 to search the directory for BEXEC*. Finally, BEXEC* should be loaded from the track(s) the directory indicates for it. Depending on the OS version, there may also be a read from track 6 before starting to load BASIC from track 2.

Are you sure the disk images you are using are for the C4P, and not for the C1P? What version of 65D are you trying to load? If this was a real floppy disk, I would suspect the stepping rate could be a problem. The test disk may test the tracks one at a time, while the real OS skips from track 6 to track 12. But I'm not sure that ever was much of an issue for 5.25" disks anyway. And it certainly shouldn't affect the Gotek, I wouldn't think. Does that numeric display on the unit show what track the "disk" is currently on, or simply which disk image on its memory card it is working with?

You can load just track 0 and return to the Monitor by typing .FFF3G when in the Monitor. You should see track 0 starting at $2200, the first five bytes should be A9 01 8D 5E 26 (LDA #1/STA $265E).

[Dokken]

Thanks for the reply. Can anyone point me to good Gotek images to try with the C4PMF in this scenario? Mark sent me several in another thread, but the floppy test is the only one that fully loads. this has been an on again/off again project for me.

[Mark]

I created a bootable disk image from the tiny 1K OSI memory test program.
Normally you'd load it from serial/cassette, but since you have a Gotek it'll be faster to put it on the flash drive.

The memory test program loads from $0150 to $03FF, and the boot loader needs 1 page of RAM at $2200 + zero page

Anyway it should load and dump you into the memory test program.
It's limited in that you can only Test, Fill & Compare memory but it should tell you if you have working RAM.

T:0400,8000 should test the RAM from 1K to 32K. If it finds no errors, it will just keep incrementing the pass count.
If it finds errors it will print the address, the write value and the read value. Press <Return> to go back to the menu.

You can restart the program (assuming it hasn't been overwritten) by entering the monitor and typing .0236G

So to build this I used "A65" to create a binary object file from MemTest.asm
Then I used "BootThis" to create a bootable disk image from the binary, specifying the load address and start address
Then I used "OSIHFE" to convert the ".65d" image to .HFE for the Gotek. You can do that for any .65d 5.25" image.

These disk images have been added to the TestMem archive on my OSI Tools page.
It looks like the program on C1/C4 might be too high on the display depending on overscan....

By the way, what firmware are you using on your Gotek drive?

[Dokken]

thank you, Mark. it boots fine to this image. I started it running about 3+ hours ago. Is it still running? it is still counting up. how/where do I view the addresses associated with the Xs?

I'm using the Glitch RAM board, so I would doubt there would be an issue there. the video board does have 5 or 6 RAM chips. should be no other ram chips in my system?

[Dokken]

RTFM

looks like the repeating Xs mean all is good?

The dynamic portion of the Memory Test Program utilizes a romping bit test successively through memory and on video systems, it prints out the pass that it is on. When the dynamic memory test makes 256 passes through memory, it wiil place an X on the video screen or on the serial system. If a block of memory makes it through this test twice--that is, places two X's--there is a 99.5% probability that it has no dynamic problems.

So I need to use C4P-DOS3-2.65D for my system? Can you outline the specific steps, command, syntax to get to an HFE file from the 65D file?

you said...

So to build this I used "A65" to create a binary object file from MemTest.asm
Then I used "BootThis" to create a bootable disk image from the binary, specifying the load address and start address
Then I used "OSIHFE" to convert the ".65d" image to .HFE for the Gotek. You can do that for any .65d 5.25" image.

step 1:I'm unclear if I need to do step 1 with this file. Do I? syntax?
step 2: I looked at the boot this readme, but I don't follow the command line syntax, memory load, etc.
step 3 I can probably do..

sorry for the questions, but I feel like I'm close. once I get the syntax specifics, I can repeat it myself.

thank you for your continued patience and help!

[Mark]

It looks like the memory test passes successfully.

If you already have a 90K ".65D" disk image (aka .dsk, .img etc.), download OSIHFE.zip and extract it in a folder with the .65D file.
You can open a command prompt and use the command-line to convert the file "OSIHFE filename.65d", OR you can use Windows Explorer to drag the .65D file onto OSIHFE.EXE, which will generate a .HFE file automatically. Likewise if you have an HFE image you can convert to 65D with the same procedure. "OSIHFE infile.hfe outfile.65d"

Some disk images may be unprocessed raw dumps that contain extra noise characters in the disk dump that do not need to be there.
You can use OSIHFE to process the files & generate a cleaned image.
OSIHFE -c infile.65d cleanfile.65d

You can identify disk type/contents using:
OSIHFE -i infile.65d

You can see directory contents of a disk image with:
OSIHFE -d infile.65d

You can test a disk image (HFE and 65D) with
OSIHFE -t infile.65d
(add the -v flag with commands for verbose decoding information when processing files, especially HFE files.)

[Run OSIHFE with no arguments to see a list of all commands.]

To create the bootable memtest disks I used the following.

Code: Select all

#REM create .bin file
A65 -b MemTest.asm

#REM create disk images (see asm file for starting address & load address)
bootthis -8 -l0150 -e0236 -o -n" OSI   MEM   TEST G0236" MemTest.bin MemTest8.65d
bootthis -5 -l0150 -e0236 -o -n" OSI   MEM   TEST G0236" MemTest.bin MemTest5.65d

#REM create HFE files
osihfe MemTest5.65d
osihfe MemTest8.65d

Hope this helps!

[Dokken]

Well I'll be dipped!

I dragged D65 files from your site onto osihfe, it created the hfe, dropped them on the Gotek and they all booted right up!! Fantastic!

thank you so much for your help!!! Off to play with this thing some more.

[Mark]

Great news!

And yes, it looks like the Memory Test screen output writes too high on the display... need to fix that.

Have fun with your C4PMF!