Some 8" drives, notably those from Shugart, had optional circuitry which you
could install yourself, BTW, to do things like recover clock from the data
stream, hence the SEPDAT and SEPCLK signals, and mask the sector pulses if
you wanted to use hard sectored media in a soft-sectored mode, or whatever.
The signal you are probably going to want is the raw data, which is common
to all of the various versions. Oddly enough, the half dozen or so SIEMENS
drives I have all have this logic as standard.
the use of the same signal as both clock and data stems from the fact that
FM data has the clock embedded in it. Hence, if you don't separate them on
the drive, the controller does that by taking the signals whether they're
separated or not, and feeding them to the respective circuits whose job it
is to process them. They then use the signals either combined or separated,
and process them as they need.
These circuits are the things which got many people into the habit of
referring to a drive as being either single or double density capable.
Without them, they're all capable of either, since the rate at which flux
reversals are written on the medium is still the same. MFM simply
integrates the clock and data in a way which doesn't waste half the
bandwidth on clock pulses.
Dick
-----Original Message-----
From: Bill Sudbrink <bill(a)chipware.com>
To: Discussion re-collecting of classic computers
<classiccmp(a)u.washington.edu>
Date: Tuesday, July 27, 1999 4:53 PM
Subject: RE: Cromemco 4FDC, How do you format a disk?
<snip>
>
>According to the docs, you have to tell the controller
>which drive type it is talking to (bit 4 at I/O address
>34 hex). Looking at the schematics, it sure looks like
>it blindly shoots signals to both connectors. FDAT is
>connected to pin 30 on the 34 pin header and pin 48 on
>the 50 pin header. Hmmm... what's going on here? Pin 48
>is SEP DATA, used with hard sectored drives isn't it?
>Pin 50 is connected to FCLK as well.
>
>What gives?
It used to be so nice.
But the dehumidifier hose got clogged while I was gone.
The room filled with mildew.
I came home and rushed to save my stuff, mostly successfully. The TU10
was affected, though. I don't know for sure if the dampness was the
cause, but I'm guessing it was. Here are the symptoms:
under rt11 5.00, it won't init a tape. used to do this w/ no problem. rt
returns "output error on mt0:" The tape spins and tries to init, but
after going back and forth a number of times, it gives up. So the
mechanism and control logic seems fine. I've tried different tapes and
cleaned and degaussed the head. No change. Jiggled all the cards in the
controller and device section to 'assure' a good connection. Nothing. I
even tried using an old copy of ROLLIN, just to see if it'd do a backup,
but it moved the reels and instantly gave a FATAL DEVICE ERROR on MT0.
Unfortunately, my xxdp pack is thrashed (moldy), so until I can
replace it, I'm stuck with trial and error here... help if you can,
please!
The subject pretty much says it all...
I can't seem to figure out how to initialize
a disk with this controller. I'm sure you do
it via the WRITE TRACK command, but I don't
follow what it is that you are supposed to
OUT to the DATA REGISTER. Also, just for
laughs and because I have hundreds of disks
(thanks AOL!), I want to hook up a 3 1/2"
floppy to it. This should be OK, shouldn't
it? How many tracks and sectors-per-track?
What interleave? Finally, I gathered up all
of the CP/M 2.2 stuff I could find on the web...
I've been looking through all of the docs...
They all seem to make the assumption that you
already have a CP/M system disk, on the correct
media, properly formatted and all you need to do
is get your system specific BIOS onto it.
Looking at the BIOS level implementation, it
seems that CP/M doesn't have a native format
command?
Thanks,
Bill Sudbrink
>Anyone know the procedure? Can it be done? Do I need some special
>diagnostic program (do vaxes run XXDP?)
I forget if they can be formatted, but I do know that they can
be erased and scanned... but you need to have something which
speaks DU Protocol (DUP)... if you have a uVaxII or uVaxIII,
you have what you need...
You have to use one of the maintenance commands from the uVax
console...
One of the commands of the form
SET HOST/MAINT/UQSSP/SERVER n
where n is the number of the UQSSP port.
Once the DUP starts up, it will prompt for a command or
filename... type DIRECT and it will give you a listing
of what is available. Some drives have different numbers
of programs available, so try what looks obvious.
BTW - the DUP protocol is essentially a protocol which
causes the VAX and the KFQSA to exchange packets of text.
You type something, it is packetized and sent to the KFQSA
where it is passed to code running in the disk. The output
of that code is passed back to the VAX in a packet.
I don't know of anything else which speaks it (unless there
is something on the user test TK50 under MDM).
And it is a proprietary protocol, so I can't explain the
specifics.
Megan Gentry
Former RT-11 Developer
+--------------------------------+-------------------------------------+
| Megan Gentry, EMT/B, PP-ASEL | Internet (work): gentry!zk3.dec.com |
| Unix Support Engineering Group | (home): mbg!world.std.com |
| Compaq Computer Corporation | addresses need '@' in place of '!' |
| 110 Spitbrook Rd. ZK03-2/T43 | URL: http://world.std.com/~mbg/ |
| Nashua, NH 03062 | "pdp-11 programmer - some assembler |
| (603) 884 1055 | required." - mbg |
+--------------------------------+-------------------------------------+
<> The 1771 is a single density only controller and not suitable for use wit
<> 3.5" disks. You want a board with a 765 (compupro) or 1793 CCS and a raf
<
<Are you claiming that you can't format/use a 3.5" disk at single density?
<Now, I can't think of a machine that did, but equally I can't think of a
<reason why you can't. The floppy drive really doesn't care how you
One, at the lower data rates the 1771 is running below the 3.5" floppies
720khz rates and the 1771 does not do the faster one! So to be compatable
with a 3.5" drive you need to run the 1771 at the 8" single density rate to
get a something approximating 720k disk. At that rate your running at
250Kbits/Sec wich is the bottom data date for 3.5" floppies.
Personally that is is a lot of effort as you will have to rewrite the bios
for the different tracks/sectors and also allow for sides. OH, forgot that
most SD controllers did only one side. If your going to the effort why not
use a good DD controller and run the dsik up to the 1.44-1.6MB range? At
that amount of storage CP/M feels like your using a slow hard disk.
I know this from putting 3.5" floppies on my Vt180s, Ampro, Kaypro, SB180,
CCS and compupro systems. All use 1793 or 765 DD controllers.
Allison
<According to the docs, you have to tell the controller
<which drive type it is talking to (bit 4 at I/O address
<34 hex). Looking at the schematics, it sure looks like
<it blindly shoots signals to both connectors. FDAT is
<connected to pin 30 on the 34 pin header and pin 48 on
<the 50 pin header. Hmmm... what's going on here? Pin 48
<is SEP DATA, used with hard sectored drives isn't it?
<Pin 50 is connected to FCLK as well.
<
<What gives?
Some 8" drives have data seperation on the disk drive and some boards used
that. NOTE: there will also be raw data from the disk and that may actually
be used. Of the 50 pins there are often many unused.
Allison
Ok, finally got a moment to put together some bits in my uVAX 3900 in the
BA213 cabinet. I've got it built as follows (from right to left)
KA655
32MB Memory
32MB Memory
DECSQA (ethernet)
KFQSA (DSSI)
TQK70 (TK70 contoller)
Then wide open space. :-) [meta question, presumably DEC had fill plates
for the empty slots?]
In the top part I've got a TK70, and two RF71's. The RF71's are connected
to the KFQSA via an internal cable that daisy chains first the rightmost
drive then the left drive and then ends at a bulkhead that has a mini
connector (looks like SCSI2). On that connector I've got a DSSI terminator
plugged in.
When I boot the cpu does all of its diagnostics but it doesn't "see" the
KFQSA. One of the drives blinks fault the other rattles around a bit and
then blinks ready.
Some questions:
The front panel has the disk lights bulkhead configured for
three disks. I connected the small connector of the rightmost
disk to disk one connector and the left disk to the second drive
connector. Is that correct?
When I type SHOW DEV to the uvax prompt it show the ethernet and
tape but NOT the KFQSA. Is that normal if the disks are not inited?
When I type Megan's command "SET HOST/MAINT/QSSP/SERVER 0"
to access the internal controller on the KFQSA I get
>>>SET HOST/MAINT/UQSSP/SERVER 0
?28 UNK SW
Any ideas anyone?
--Chuck
You'll need to spend a bit of time studying out how the controller you're
using handles the drive select or even a cable select if it has separate
selection on the 34 and 50-conductor cables, and whatever it does to
manipulate the clock extraction circuits to derive the appropriate clock for
whatever data rate you're wanting. BE CAREFUL! Some controllers switch the
clock frequency as they make whatever assumptions they may make about the
nature of what's on each cable. In order to drive 3.5" diskettes, you need
to fool the controller and software into believing it's talking to a drive
on the 50-pin cable. Though the 8" drives have only 77 tracks, they have
the 360 rpm spin rate and most other parameters in common with the little
teensy drives. If your controller switched the clock assuming its top data
rate will be 250 kbps, you may have problems.
Dick
-----Original Message-----
From: Chuck McManis <cmcmanis(a)mcmanis.com>
To: Discussion re-collecting of classic computers
<classiccmp(a)u.washington.edu>
Date: Tuesday, July 27, 1999 3:29 PM
Subject: RE: Cromemco 4FDC, How do you format a disk?
>At 04:38 PM 7/27/99 -0400, Allison wrote:
>>The 1771 is a single density only controller and not suitable for use with
>>3.5" disks. You want a board with a 765 (compupro) or 1793 CCS and a raft
>>of others.
>
>Which defines, for the most part, the difference between a 4FDC and a
>16FDC. The latter used the 1793 I believe. I've got a copy of the Western
>Digital databook that has these ICs in it if someone needs a copy.
>
>BTW, I wrote the formatter for CP/M on the 16FDC (in Turbo Pascal no less
>:-)) and it simply loaded up a buffer, then did a write track and stuffed
>the buffer with an OUTIR instruction. (the board did wait states so this
>worked.) The process involved creating a buffer and using special 8 bit
>codes for "sector gap" and "index mark" and "sync mark". I've long since
>given away that code but the person I gave it to is lurking on this list...
>
>--Chuck
>
This thing will supposedly execute 'highlighted' Forth and 6800 assembly
code with some (leap)key gyrations... Anyone know?
Thanks
- Mike: dogas(a)leading.net
Yes! I finally scored a Pro380, well, actually it's a "VAX Console", but
hey, same thing right?
Question, does it use the same monitor as the DECmate and Rainbow's? I
know there was a colour and a mono version of the monitor, is do they both
work on the same computer, or do you have to match specific computers to
specific monitors? Having asked that, I should be getting the monitor in
the next few days from the person I just got the computer from.
I know it's got a RD52 in it, and I suspect it's a working system as it is
off a system that is currently being deinstalled (which is why I don't have
the monitor yet).
Yes, it's safe to assume I'm fairly clueless about Pro's.
This is so COOL!!!!
Zane
OK, I'm calmed down now.... NOT!!! :^)
Yep, I'm bouncing :^)
You'd never guess I've been wanting one of these would you? :^)
| Zane H. Healy | UNIX Systems Adminstrator |
| healyzh(a)aracnet.com (primary) | Linux Enthusiast |
| healyzh(a)holonet.net (alternate) | Classic Computer Collector |
+----------------------------------+----------------------------+
| Empire of the Petal Throne and Traveller Role Playing, |
| and Zane's Computer Museum. |
| http://www.aracnet.com/~healyzh/ |
