Copying DEC VAX set up disks rx50, followup

Dave Dunfield dave06a at
Mon Mar 5 05:59:34 CST 2007

> All is well,  the disks booted a VAX 8250 with out  any problems
> as I said in a earlier post it was the controller that caused the problem.
> I guess that why I have 6 image disk machines. 

Funny - I have one main IMD machine - that one supports single-density
and two drives and has taken everything I have thrown at it. Find a
machine with a good FDC and stick with it - you should not need 6 separate

>     1) question.  I was reading some Cromemco Cromix  disks
>     which are 360k  DD with sd boot tracks on a  1.2  meg dirve 
>     and a SD controller. Disks where all original Cromemco
>     or Dynatech software.

I have found that with Cromemco disks, I almost always have to slow
the drive by about 10rpm - otherwise, depending on the machine it
cannot reliably read either the first or last sector on a track.

[This is the main reason IMDU can merge images - until I sorted this
 out, I read Cromemco disks on two machines, one that missed the first
 sector, and one that missed the last - I could recreate the complete
 image by merge them...]

>    I would  get 
>    0/0: unable to determine interleave
>         :300k SD - 18 sectors of 128 bytes  G1:7 G2:9
>    0/1 300k DD -10 sectors of 512 bytes   G1:13 G2:20
>    and on....................  (xxx data, xx compresses. 0 unavil)
>    This did not happen on all of the disks (50%) but the first one
>    which is bootable, did have problems. 
>    It seemed to read them OK,  but I switch to a 360 drive 
>    to be safe. I did  not try to write any of these but would 
>    they still work with the interleave error.  The 360k drive setup
>    had  no problems  at all.

I would recommend doing Cromemco disks on a DD drive. - In general,
I recommend using the drive type matching the target system, but this
is especially true with Cromemco disks.

When ImageDisk analyzes a track, it waits for the index pulse, then
scoops up all the sectors until they repeat, recording the sector
numbers which it also uses to calculate interleave. It rechecks this
by reading a few more revolutions, and "unable to determine interleave"
means that additional sectors turned up during the check passes - It
tries this a few times and could not get a consistant read.

As long as all the sectors are there, it will probably work without
the same interleave as the original - although it might take a little
longer to read contiguous blocks of sectors.

If you know what the sector interleave should be, you can use IMDU to
re-interleave the image to that value, or you can tell ImageDisk to
write the image with that interleave when you recreate the disk.

Another trick you can use is to copy the disk on the target system
after you make it with ImageDisk - this will normally restore the
interleave to the native formatting for the system.


dave06a (at)    Dave Dunfield
dunfield (dot)  Firmware development services & tools:
com             Collector of vintage computing equipment:

More information about the cctalk mailing list