CBMSTUFF FORUM
Hard sectored media - Printable Version

+- CBMSTUFF FORUM (https://www.cbmstuff.com/forum)
+-- Forum: CBMSTUFF PRODUCTS (https://www.cbmstuff.com/forum/forumdisplay.php?fid=1)
+--- Forum: SuperCard Pro (https://www.cbmstuff.com/forum/forumdisplay.php?fid=3)
+---- Forum: Hardware (https://www.cbmstuff.com/forum/forumdisplay.php?fid=42)
+---- Thread: Hard sectored media (/showthread.php?tid=298)



Hard sectored media - Edstrom - 02-11-2016

I have a hard sectored 5 1/4" floppy that I want to preserve. I have dumped it using the no-index option and is currently deschiffering the results. Here I got a question: Will the flux start to be sampled relative the first index hole found or just at random place on the track? I understand the index hole is not checked again but at the starting point it would be good to know it is at the start of a sector.

When turning on the index check the floppy drive starts to stutter and I guess the SCP software is stepping the head 16 times per track instead of 1, which got me thinking.

1) Best thing would of course be to be able to specify the number of index holes per revolution. I think I would be capable to implement that but the firmware source code is not public, right? Any plans to provide a library for custom firmware development?

2) Next best thing would be to be able to define 16 revolutions per track and then just dump the flux to a SCP file. This would of course surprise the current samples/revolution logic so that would need to get a free running mode. This need to be described in the SCP file track header, maybe by setting the RPM to something insane like 360 x 16.

EDIT: 3) A possibly simple solution would be to set the number of ignored index pulses between head steps?!

Since I am doing this for preservation I need to restore the filesystem from the floppy on a PC from the flux data. I don't have the original hardware functioning. Is this even possible without features as proposed? How would I find the start of the track?

BR / Joakim


RE: Hard sectored media - admin - 02-11-2016

Turning off the index does not use the index at all when doing a read, so the start and end are random (and either 166.666ms or 200ms is captured, depending on the drive RPMs). You are probably better off using SPLICE mode and turning the index off when imaging disks.

It's not possible really to use a hard sectored setup with SuperCard Pro's firmware AND have the index on. With the index on, the index pulse signifies signifies the start and then end of the track. So, with a hard sectored disk only one sector per track (not the entire track) is being captured with the index on.

I will have to think about how to accomplish this.


RE: Hard sectored media - Edstrom - 02-12-2016

How exact is the rotation speed of a floppy drive?

Will the start point drift when I turn the index off and the rpm is inexact? If the drift is 1 mS per turn the result will have an overlap of 1mS for each track or worse, lack 1 mS each turn unless triggered by an index. This will also make it difficult to stitch the flux data into real sectors of course.

I can see that reading a DOS floppy with different options, eg ibm360, ibm720 and ibm1440, causes different track time generated which is strange I think.


RE: Hard sectored media - admin - 02-12-2016

If you are capturing 2 revolutions using SPLICE mode with the index off, you are guaranteed to have the entire track, so there is nothing to stitch together.

The SCP hardware uses the index pulse to trigger the hardware timers and data collection.  It's all done asynchronous to the CPU itself, which is why it works so well.  I think I would have to re-write the capture code for a hard sector disk to count the number of index pulses and stop on the correct number (8, 12, or 16... depending on type of disk).  That is going to take some thought because it's not a simple matter.


RE: Hard sectored media - Edstrom - 02-12-2016

Ahh, got it, I'll try that and see if I can extract full tracks from the flux and convert it into a mountable image, or just put a fuse file system on top. Thanks for the clarifications, I post back the results when I have some.