MFM emulator - can't read disk

[Vladimir Romanov]

Hello!

I am using MFM emulator to read Miniscribe 3425 MFM disk. It comes from AT&T UnixPC 7300. I can boot from this MFM disk on UnixPC (it takes some time to boot from this MFM disk).

# cd mfm

# ./setup_mfm_read 

Rev B Board

# ./mfm_read --analyze --emulation_file ../miniscribe_3425.emu --extracted_data_file ../miniscribe_3425.data --transitions_file ../miniscribe_3425.transitions

Found cyl 0 to 615, head 0 to 3, sector 0 to 16

Expected 41888 sectors got 41613 good sectors, 256 bad header, 19 bad data

I cannot boot from this image on UnixPC.

But I have successfully read seagate st-251-0 disk and i can boot from this image on UnixPC. So i think the MFM emulator/reader is not a problem.

Can i attach full log and transitions file to this topic?

Thank You

Vladimir

[David Gesswein]

Upload them to
http://www.pdp8online.com/upload/abuploadab.html
Compress the transitions file.

> --
> You received this message because you are subscribed to the Google Groups "MFM Discuss" group.
> To unsubscribe from this group and stop receiving emails from it, send an email to mfm-discuss...@googlegroups.com.
> To view this discussion on the web visit https://groups.google.com/d/msgid/mfm-discuss/a509a109-00ac-4b2f-907b-ed1fb09fa898n%40googlegroups.com.

[Vladimir Romanov]

done

miniscribe_3425.tar.gz

[David Gesswein]

Issue is Bad sectors on cylinder 0 head 3: 10 12 13. The read of the disk
with my board was unable to recover those three sectors. Plotting the
transitions it looks to be a bad enough error burst that I don't see how
the 3b1 was able to read the sectors. Its only a 16 bit CRC so its
possible that with enough reads it was able to get one where the CRC matched.
I assume the boot slowly is it printed the block character on the screen
for a while before finally booting.

I looked at images I have and your and it appears those sectors have a fill
pattern but no useful data. The read error can be removed with these commands.
The data itself will still be wrong.

~/corvus/mfm/mfm_util --tran miniscribe_3425.transitions --ext miniscribe_3425.ext

~/corvus/mfm/ext2emu --emu miniscribe_3425.emu --ext miniscribe_3425.ext --heads 4 --cyl 616 --format wd_3b1

I verified miniscribe_3425.emu will boot on my 3b1.

On Fri, Dec 03, 2021 at 12:06:04AM -0800, Vladimir Romanov wrote:

[Vladimir Romanov]

3b1 indeed boots slowly with this disk.

Thank You!

[Forgotten Machines]

What about setting the retries number higher on the MFM reader command line?  I did this trick with a number of marginal drives, and especially if there's an indication that the UNIX PC can retry enough to get value from these sectors, if we set the MFM Reader retries up to 50 or 100, I'll bet it can catch those few bad sectors too....it's worked for me on the very margina. Convergent AWS drives anyway...

Just a thought...

Best always,

AJ Palmgren

515.460.6393

http://MightyFrame.com

http://Entrex480.com

To view this discussion on the web visit https://groups.google.com/d/msgid/mfm-discuss/445e3935-faaf-4f17-a13d-37663a9855e9n%40googlegroups.com.

[Vladimir Romanov]

I tried different settings (default value is 50,4)

--retries 100,8

--retries 200,16

No luck. I can't boot from this images.

Now i can't boot from physical miniscribe mfm disk on my unixpc. But i can boot from fixed emu image.