The following is an exerpt from a text file written/complied by one
A.D. Longton of Rockville, MD. I have omitted the discussion of how
to make a 1.44M disk from a 720K disk with a soldering iron since I
don't want to be a party to furthering that practice. I did, however,
find the remainder of the information quite interesting and have
included it here. As You can see, it comes directly from the
'brain trust' at Big Blue and may tend to lend some measure of credibility
to what I've been saying all along. The original file was dated 5/10/89,
I'm not sure when the information spewed forth from Boca Raton.
3.5" DISKETTE FORMATS
Boca Raton, Florida
"Reprinted by permission of the
IBM Personal Systems Technical Journal."
Page 42, issue 2, 1989
"The original recommendations about the proper formatting and use
of PS/2 diskettes have undergone revision. This article explains
why the recommendations have changed.
THE ORIGINAL CAUTION
Personal System/2 shipping cartons include a sheet of paper that
cautions users not to format a 2.0 MB diskette to 720 KB, because
the diskette becomes unusable and should be discarded.
This caution was issued because of the physical properties of 720
KB diskettes versus 1.44 MB diskettes. The 720 KB format uses a
higher write current, and the 1.44 MB format uses a lower write
current. To accommodate the higher write current, the oxide
coating on a 1.0 MB (720 KB formatted) diskette is denser than the
oxide coating on a 2.0 MB (1.44 MB formatted) diskette.
When you format a 2.0 MB diskette to 720 KB, you apply the higher
write current to the less dense oxide coating. The hardware
developers originally felt that this meant the 720 KB formatting
pattern is written too deeply into the 2.0 MB oxide coating,
causing intermittent data errors and unreliable use. Furthermore,
the developers felt that if you attempted to reformat the diskette
to 1.44 MB, which uses the lower write current, the 1.44 MB format
would not completely write over the "deeper" 720 KB format.
Therefore the developers' recommendation was to discard a 2.0 MB
diskette that was formatted to 720 KB.
THE SUBSEQUENT FINDINGS
[aka a RETRACTION]
Since the time that this caution was issued, the developers have
performed additional testing, and have concluded that there is no
need to discard a 2.0 MB diskette that was formatted to 720 KB.
It is still true that a 2.0 MB diskette formated to 720 KB will
cause intermittent data errors. However, the latest assessment is
that you will be able to reformat the diskette to 1.44 MB and use
it reliably after that.
The same logic applies to a 1.0 MB diskette formatted to 1.44 MB.
You cannot use it with the 1.44 MB format, but you can reformat it
to 720 KB and use it reliably after that.
Therefore, the current recommendation is: If you format a
diskettte to the wrong capacity, do not discard it; instead,
reformat it correctly and use it."
With all those feelings and recomendations on those feelings it
makes me wonder how much experimentation was actually being done
on a strictly scientific level. Note that the one mention of
formatting 1.0mb disks to 1.44 MB does not say that you will get
errors if you use them. What it does say is that if you reformat
that wrongly formatted disk, you can reliably use it at 720 KB.
The implication is that since there were errors with 2.0mb disks
formatted to 720 KB "logic applies" that there will be errors if
the reverse is done. This is not necessarly the case, and we are
not told why, we are just told.
FYI, here are the specifications for the 720 KB, 1.44 MB, and 360k
5.25" disk drives as listed in the same issue on pages 43-44.
Note the large similarity between 360k and 720 KB disks and 720 KB
disks and 1.44 MB disks.
720 KB and 1.44 MB Diskette Drives
720 KB 1.44 MB 360 KB (5.25")
Track-to-track 6 ms 6 ms 6 ms
Head settle time 15 ms 15 ms 15 ms
Motor start time 500 ms 500 ms 750 ms^
Disk rotational speed: 300 rpm 300 rpm 300 rpm
Maximum Latency 200 ms 200 ms 200 ms
Formatted Characteristics: 720 KB 1.44 MB^ 360 KB^
Tracks (actual) 80 80 40 ^
Tracks per inch 135 tpi 135 tpi 48 ^
Sectors per track 9 18 ^ 9
Bytes per sector 512 512 512
Bytes per track 4608 9216 ^ 4608
Data heads 2 2 2
Sector interleave factor 1:1 1:1 1:1
Sector skew factor 0 0 0
Sectors per cluster 2 1 ^ 2
Transfer rate 250,000 500,000^ 250,000
(bits per second)
(All ^'ed numbers are numbers that are different from the 720 KB
"...if they think you're technical, go crude. ....
These days, though, you have to be pretty technical before
you can even aspire to crudeness."
--From William Gibson's short story