[I am using the popular names (such as "360K", rather than more
technically correct designations.]
The most important parameter is the magnetic coercivity.
"360K" is 300 Oerstedt.
"1.2M" is 600 Oerstedt. Therefore, the two will not interchange
Sure. Some of the IBM Techrefs get this wrong, BTW. At least one edition
gives the same coercivity for the 2 types of disk.
satisfactorily. Unlabeled disks can sometimes
be differentiated by
color. If a disk has a reinforcer around the center hub, then it is
probably a 360K, although it COULD be an after-market reinforcer (jigs
used to be available.) If there is NOT a reinforcer, then it is either a
1.2M, or a very early "360K"
Don't forget that '720K' 5.25" disks also exist, although they're
not
common on IBM PCs. By that I mean 80 track, double density, double sided,
300 oersted coercivity. Most of those do not have the reinforcing ring IIRC.
<RANT> There ain't no sech thing as a
1.44M disk. The IBM style of HD 3.5
Agreed. But I've got fed up with people correcting me when I call them
1.4Mbyte disks and have adopted popular usage :-(
has 2 sides, 80 tracks per side, 18 sectors per
track, and 512 bytes per
sector. If you multiply that out, you get 1.406 HONEST Megabytes
(1048576). The only way to get 1.44 out of that is to creatively redefine
a Megabyte to be 1024000 bytes. That leaves IBM in the position of
1440Kbytes is OK, though :-)
1 Megabyte is often defined as 10^6 bytes by hard disk manufacturers.
Alas that doesnt' work for floppy disks either.
DS v SS (5.25"): The disks are manufactured
the same, but the difference
is whether BOTH sides are tested and/or "certified". Using DS for SS is
acceptable. Using SS for DS is taking a chance on using untested media.
Sure. But I have _never_ had a problem with using an RX50 as double
sided, unlike attempting to use '360K' disks as 80 track ones, or even
trying to format some lesser brands of disks as even 360K.
I would not suggest using an RX50 like this for data that really matters
(in fact I'd not keep that sort of data on any floppy disk apart from
perhaps an 8" one). But if you need a few disks for a TRS-80 or BBC micro
with 80 track drives then this is a possible solution. It works a lot
better than using '360K' disks, even good name-brand ones.
"Flippy": The second side can often be
used in a single sided drive by
flipping the disk over. In the case of Apple ][ and Commodore, it
requires punching a write enable notch. (Which does NOT need to be square.)
On TRS-80, IBM, etc, it is necessary to also punch an additional
(symmetric) access hole for the index hole. (jigs for marking and
punching used to be available.)
THe Siemens drives in my Zenith Z90 have mounting holes in the chassis
and pins on the PCB to connect a second index sensor and WP sensor. So
you can use any disk as a flippy in those. My drives don't have the extra
sensors fitted, though.
DS V SS (8"): The index hole opening is in a different location.
Punching a new hole through the jacket normally works OK.
Be careful if doing this to turn an SS disk into a DS one. Double sided 8"
drives have 2 index sensors so they can tell what sort of disk is
inserted (there is a 2-side line on the SA850 interface). Some drives get
_very_ confused if both sensors are active.
720K 5.25" v 360K 5.25": again, an issue of testing/certification,
similar to SS v DS. At least for a while, they were manufactured the same,
but were tested/certified for 48tpi or 96 tpi.
I have had very little success in formatting 48tpi disks in 96tpi drives.
As I mentioned earlier, this includes name-brands like 3M..
Interesting! I have never had a problem using even generic 48tpi disks
at 96tpi.
- don
Hard-sectored
v soft-sectored: In a system that does not use the index
hole (Commodore, Apple ][), it doesn't matter. In all others, the only
way to use the wrong diskette would be to modify the drive to index off
of the spindle instead of using the sensor.
Using hard sectored disks as soft-sectored should be possible. Just add a
monostable and a couple of gates to only output a pulse when 2 holes are
close together.
In other words trigger a non-retriggerable monostable (with a time period
of 3/4 of a sector time) off the trailing edge of the index pulse. AND
the index pulse from the drive with the Q output of the monostable and
feed the result to the controller.
On the sector holes the monostable has already timed out, so the AND gate
blocks the pulse. On the index hole, the monostable hasn't timed out from
the previous sector pulse, so that one gets through.
To do the reverse is a lot more difficult - you need to have the sector
pulses accurately positioned wrt the data on the disk.
I have thought about taking an old 5.25" drive (something that is not
hard to get) and stripping it to get the chassis + spindle + disk clamp.
Then mount a punch/die in place of the index sensor and a disk with
suitable notches in place of the motor pulley. Add a detent on the drive
to go into the notches. Line up the existing hole with the punch and die,
set the detent into one of the notches, and clamp the disk. Now move
round the notches and punch the extra holes. I must try this...
3", 3.25": Many newbies will get sloppy in reference to 3.5" diskettes,
without realizing that there actually were 3" and 3.25" diskettes. 3"
were used by Amstradt, some non-US Canon?, and Amdek add-on drives for Coco
The most common use in the UK was the Amstrad machines. But they were
also used on the Tatung Einstein (I am looking at one) and on the Oric
drives. And in some other machines, I'll bet (I have seen them used on a
BBC micro).
-tony