Page 5 of 8

Re: Building the KLyball D-13

Posted: Thu Jun 25, 2015 12:05 am
by Jeff
I figured it out.
Connect 2 Drives.jpg
Connect 2 Drives.jpg (115.54 KiB) Viewed 48750 times
I also made a schematic to help figure out the signals when hooking up other floppy drives. See attached.
Klyball D13 Schematic.pdf.zip
(53.25 KiB) Downloaded 2091 times
/Jeff

Re: Building the KLyball D-13

Posted: Thu Jun 25, 2015 3:33 am
by Jeff
Well I am having problems getting my Superboard to boot. I have made a boot disk from OS65 v3.1 for C1P but it gets 7 tracks in and hangs. I have dumped the disk and compared it to the split files and it is perfect. I have tried 2 different drives making a boot disk with each drive. It worked a month ago and once today so I thought it was a memory problem. I wrote a ram tester and it all checks out fine. I even swapped the ram out for different ram. Still won't boot.

I would like to try a different boot disk but I do not know of any.
Does anyone have or know of other boot disk images for the C1P non rev D?
Does anyone know of a more robust Ram tester?
Does anyone have any suggestions on what else I might try?

Frustrated
/Jeff

Re: Building the KLyball D-13

Posted: Thu Jun 25, 2015 3:52 am
by Klyball
I would recheck your timings on the 610 and separator, also are these 5.25 disks on old drives or did you switches drives from the one you made the disk with
do a memory dump after it hangs

Re: Building the KLyball D-13

Posted: Thu Jun 25, 2015 4:59 am
by Jeff
I will re-check the timings, but I don't think they are out because when I do a disk dump, every track and sector reads back EXACTLY matching the data that I sent and SAved.

I am using, firstly the Epson dual 3.5 and 5.25 drive with disks I created about 3 weeks ago with that drive. I suspected the drive, so I switched it out with one of the Fujitsu 360K drives, of course, it wouldn't read the disks made with the Epson, so I made an entirely new startup disk from the split files. I disk-dumped it and compared each track and sector and it matched exactly also.

Both disks boot to track 7 and hang.

I figured that during the disk creation process I wasn't using much high memory and neither with the disk dump program, so perhaps, when booting it loads enough to hit some bad ram higher up in memory. But my BASIC memory test passed fine. The other reason I suspect memory is that last month when I made the Epson disk, I was using borrowed ram from my C4P, and since then, my new ram arrived, so I fully populated the 610.

Im not sure how to do a memory dump after it hangs.... or what to look for when I do.

Re: Building the KLyball D-13

Posted: Thu Jun 25, 2015 5:17 am
by Klyball
Watch those rams with the flimsy pins I had a batch that checked out then started dying they seemed to work but when I ran them in my ram tester they failed

Re: Building the KLyball D-13

Posted: Sun Jun 28, 2015 1:28 am
by Jeff
I have narrowed things down a bit. When the machine is cold, the disk boots fine for the first few minutes. After that, it fails around track 7. So its definitely hardware, and its likely heat related.

I ran an exhaustive memory test overnight and all the ram checked out fine. I have replaced the ACIA and the PIA and the problem persists. Im not sure what next to try.
Is there anything on the separator that changes with heat?

/Jeff

Re: Building the KLyball D-13

Posted: Sun Jun 28, 2015 4:15 am
by MK14HAK
Jeff there is a memory test on osiweb/software /utilities that Phil uploaded.

Re: Building the KLyball D-13

Posted: Tue Jun 30, 2015 3:43 am
by Jeff
I tried it out... just hangs... so Im looking at the code now.

/Jeff

Re: Building the KLyball D-13

Posted: Wed Jul 01, 2015 7:35 pm
by BillO
Hi Folks,

Trying to figure out why I can only write $24 to a disk.

In this scope image, the yellow trace is the inverted data coming from the floppy ACIA as I write $AA to it. The blue trace is pin-9 going to the D-13. Does this look correct?
write-data.jpg
write-data.jpg (61.24 KiB) Viewed 48692 times

Re: Building the KLyball D-13

Posted: Thu Jul 02, 2015 3:52 am
by dave
At a glance, it looks correct. 1 start bit, 10101010, 1 stop bit.

The blue trace looks like clock + data pulse for 1, clock only for 0, providing the frequency modulation. Each pulse in the blue trace should be close to 400 ns.