31 Aug
2001
31 Aug
'01
2:41 p.m.
Richard Erlacher wrote:
Well, if the written spec's can be believed, you
can't SEE the difference
between DD and SD media. That doesn't change the fact that they have different
spec's. Further, DS media are certified on BOTH sides, while the emulsion on
one side may be defective.
Jerome Fine replies:
Then some more information may be helpful. Note that since this is only my
own experience, it might have been different for others if they had installed
the same switch. However, since I believe that 8" floppies are no longer very
popular, it is unlikely that this information will actually be put to the test.
Over about 5 year from 1988 to 1993, I formatted about 900 SSDD diskettes
from about a dozen manufactures as DSDD. They were my
PRIMARY
backup and I still have a number which I need to transfer to a hard disk
drive.
I seem to remember that ALL SSDD floppies that I attempted to format
as DSDD floppies in the DSD 880/30 worked except for some from one
particular manufacturer that were done in 1993. At the very end, I acquired
(rescued from pile on the way to the dumptser - i.e. still inside the building
in a normal environment, so storage conditions would NOT have caused a
problem) a couple of dozen of DEC diskettes with software for RSX-11.
(Sorry, back 8 years ago, I did not realize that they might have been worth
something.) About half of the DEC diskettes would not format correctly.
After the format operation was finished, I always did a READ check of
every block. What was interesting was that when there was a problem,
it was on the original side of the floppy and the so-called second side was OK.
The other bit of information is that when these SSDD floppies were sometimes
used again in a DEC RX02 drive, there was NEVER any problem with the
original side - except of course with some of the DEC floppies.
It's true that DS media have their index hole
located differently than SS media.
The index timing is different, according to some drive makers, albeit not VERY
different. At least one mfg spec's the DS drives with 160 microsecond +/- 50
from the start of /Index to the data burst, while SS media are at 200. I've
never given much thought to this and how it affects things, but I am not
comfortable with the notion of switching, mechanically, between the two sensors.
Dick
Since I had no specs, I just tried it. It worked. Not one floppy that was formatted
as DSDD was unable to be used in a DEC RX02 drive. In addition, one some of
the SSDD floppies (before I had installed the switch) I have punched a second
set of holes in the DSDD index position. I initially covered up the SSDD index
holes, but found that these same floppies worked without a problem in both the
DSD 8" "dual" drive and a DEC RX02 drive. I never worked out the timing
between the two index holes, but since they are about 1/2" apart, it should be
trivial based on the number of revolutions per second. The distance from the
centre of the floppy at that location is about 1 1/2" (from memory - it is probably
between 1" and 2"). At 300 rpm or PIE * 3" * 6 per second = about 50"
per second,
I get a timing difference of 10 milliseconds. Since 1/10" is still about 2
milliseconds,
I suspect that any timing requirements less than 1/2 milliseconds would be too
difficult to meet. Of course, my calculations might be totally wrong.
As far as the comment from: Don Maslin
Probably of little consequence, Dick, since his DSDD disks will not be
readable by anyone else unless the have performed the same modification
to a single-sided drive.
Oops! Obviously, I meant and should have said "double-sided drive".
As I mentioned above, the so-called first side could always be accessed
via a single-sided DEC RX02 or any other single-sided drive. As for an
unmodified double-sided drive, I have one on an SMS-1000 box that is
very interesting. It uses 8" floppies and is also both single-sided and
double-sided in operation in that it senses both index hole locations.
However, it is VERY different in operation. FIRST, although it obviously
uses the same interleave scheme as DEC, the emulation is MSCP, i.e.
in RT-11 it uses the DEC DUX.SYS device driver. The first time I
tried to boot RT-11, I was stumped. Finally it dawned on me. On
the DSD 880/30, I did the command:
COPY/BOOT:DU DY0:RT11XM.SYS DY0:
with a SSDD floppy that had RT11XM.SYS and both the DYX.SYS
and DUX.SYS device drivers. Worked just fine. SECOND, although
the number of blocks is 988 in single sided-mode and 1976 in double-sided
mode, in double-sided mode the second side is accessed right after the
same track for the first side. Which means that since the DYX.SYS
device driver on the DSD 880/30 accesses first all 988 blocks on one
side and then all of the 988 blocks on the other side, the same DSDD
floppy can't be used on both systems. If the 8" floppy is used to transfer
any files in RT-11, I can use ONLY SSDD media after which I can
write the contents of two SSDD floppies to one DSDD floppy from
a temporary file since the SMS-1000 has only one 8" drive.
All of the above is very much in the past over 10 years ago and I have
probably not touched an 8" floppy in many years except to send some
with TCP/IP to the UK.
So, yes Tony, I do play with hardware to some extent, but only in direct
relationship to a software need.
Sincerely yours,
Jerome Fine
----- Original Message -----
From: "Jerome Fine" <jhfine(a)idirect.com>
To: <classiccmp(a)classiccmp.org>
Sent: Friday, August 31, 2001 7:42 AM
Subject: Re: 8" Floppies
> >Absurdly Obtuse wrote:
>
> > For those who asked, I need DSDD 8" Soft-Sectored Disks.
>
> Jerome Fine replies:
>
> I don't know if this will help, but here goes.
>
> I was using the DSD 880/30 - an RL02/RX02 compatible. It
> holds a single 8" floppy drive which can use both SSDD and DSDD
> diskettes (soft-sectored). I found that since I had a small number of
> DSDD diskettes and the media seemed to be identical, the ONLY
> difference that I could observe was that the index hole was in a
> different location.
>
> I wired a DPDT switch into the index hole detection circuit. In the left
> hand position, it was wired to duplicate the existing circuit. When I
> wanted to REVERSE the operation, I used the right hand position of
> the switch. In the second position, the index hole for a SSDD diskette
> caused the drive to route the signal from the sensor to the portion of
> the circuit which normally detected the signal from the DSDD index
> hole. Thus, the drive and the software "THOUGHT" that a DSDD
> media was present and I was able to use a SSDD diskette as a DSDD
> diskette without having to punch a second set of index holes. The
> use of a DSDD diskette as a SSDD diskette is also possible if that
> is useful - I never found it to be needed.
>
> Naturally, this required that I do a LLF (Low Level Format) on both
> sides of the diskette. This the DSD 888/30 is able to do all by itself
> and that solved the problem. As for using the DSDD diskette with
> DEC software, V4.00 of RT-11 had some code which showed how
> to access the "second" side and use the floppy drive as an RX03.
> I used that code as a starting point since I also wanted to allow the
> DYX.SYS device driver in a 22-bit environment with V5.03 of
> RT-11. Note that the DSD 880/30 and the DEC RX02 with an
> M8029 can use only the first 1/4 MByte (18 bit addresses) when
> a transfer is made to/from the DMA silo from/into a buffer in the
> memory of a PDP-11. Software in the DYX.SYS device driver
> is able to use all 4 MBytes of the PDP-11 physical address space
> in a manner totally transparent to the user of the device driver.
>
> I realize that Sridhar is not using RT-11, but I just thought I might
> mention the above for those of you that do.
>
> Sincerely yours,
>
> Jerome Fine
>
>