14 May
2007
14 May
'07
7:50 p.m.
Hi,
I have a Transam Tuscan system and recently received a set of boot disks for
it. I am trying to back these up to a PC, but when reading the disks with
Anadisk or Imagedisk I get 'no data' errors on tracks 3 and higher. The
errors get worse from tracks 3 to 15 then nothing can be read. I have taken
the 5.25" drive out of the Tuscan and put it in the PC, and the contoller
passes the TESTFDC tests for single density and double density.
The odd thing about these floppies is that Anadisk reports them as 512 byte
/ sector and 10 sectors per track. The norm for '360K' floppies is 9 sectors
/ track. The outer sector (track 0) can be read fine with Anadisk, and I can
read the CP/M welcome message.
I was hoping someone could advise on what tricks or techniques I could use?
Would it help for example to slow the drive down below 300rpm?
I assume that with 9 sectors / track there is more time for the FDC to
recgnise the start of each track. I think the Tuscan FDC uses a WD1771
controller, is that better than the NEC765 type-controller and able to pack
more data in? I read on http://www.s100-manuals.com/Disk-drives.htm that
Kaypro also uses 10 sectors of 512 bytes.
Another thought is to try different PCs, perhaps formating, writing and
reading 10 x 512 byte sectors is another test that could be added to Dave's
TESTFDC program to help find a suitable machine :-)
Lastly (before anyone asks) yes the floppies are soft sectored.
Regards,
John
_________________________________________________________________
The next generation of Hotmail is here! http://www.newhotmail.co.uk/
15 May
15 May
10:09 a.m.
From: "John S" <john_a_s2004 at
hotmail.com>
Hi,
I have a Transam Tuscan system and recently received a set of boot disks
for it. I am trying to back these up to a PC, but when reading the disks
with Anadisk or Imagedisk I get 'no data' errors on tracks 3 and higher.
The errors get worse from tracks 3 to 15 then nothing can be read. I have
taken the 5.25" drive out of the Tuscan and put it in the PC, and the
contoller passes the TESTFDC tests for single density and double density.
The odd thing about these floppies is that Anadisk reports them as 512 byte
/ sector and 10 sectors per track. The norm for '360K' floppies is 9
sectors / track. The outer sector (track 0) can be read fine with Anadisk,
and I can read the CP/M welcome message.
I was hoping someone could advise on what tricks or techniques I could use?
Would it help for example to slow the drive down below 300rpm?
I assume that with 9 sectors / track there is more time for the FDC to
recgnise the start of each track. I think the Tuscan FDC uses a WD1771
controller, is that better than the NEC765 type-controller and able to pack
more data in? I read on http://www.s100-manuals.com/Disk-drives.htm that
Kaypro also uses 10 sectors of 512 bytes.
Another thought is to try different PCs, perhaps formating, writing and
reading 10 x 512 byte sectors is another test that could be added to
Dave's TESTFDC program to help find a suitable machine :-)
Lastly (before anyone asks) yes the floppies are soft sectored.
Hi John
It sounds like the disk has tracking errors. Either the drive
you are using is not stepping correctly or the drive it was
written on was not stepping correctly.
Things to check in general:
Make sure there is no old grease on the guide rails for the
head mech.
If belt driven, make sure the surfaces of the pullies are smooth
and clean.
Make sure the head(s) are clean. If a single head, make sure
the pad is clean and flat.
Make sure that the disk turns smoothly when in the drive
and that if belt driven, the belt isn't slipping.
Dwight
_________________________________________________________________
Like the way Microsoft Office Outlook works? You?ll love Windows Live
Hotmail.
http://imagine-windowslive.com/hotmail/?locale=en-us&ocid=TXT_TAGHM_mig…
12:45 p.m.
On 14 May 2007 at 23:50, John S wrote:
The odd thing about these floppies is that Anadisk
reports them as 512
byte / sector and 10 sectors per track. The norm for '360K' floppies is 9
sectors / track. The outer sector (track 0) can be read fine with Anadisk,
and I can read the CP/M welcome message.
I hadn't realized that there was a norm for CP/M floppies. You know,
the "recommended" 512-byte sector "packing" in the uPD765 literature
is 8 sectors--and indeed, early PC-DOS used just that.
It's entirely possible to fit 10x512 sectors per track using a 765-
type controller, though it's much easier if the IAM is left off and
the space gained used to expand the inter-sector gaps (you can do
this with a uPD 2765-type controller or any of the WD 17xx/27xx
controllers.
I think the Tuscan FDC uses a WD1771 controller...
Nope, the 1771 is restricted to FM only. Maybe a 179x or
1770/1772/1773 though.
It's entirely possible that the drive used to create the diskettes
was somewhat out of alignment. You may have to "unalign" a drive to
successfully read these. This is more common than you might think.
You didn't mention if you were using a 48tpi ("360K") drive to read
these or trying to get by on a 96tpi ("720K" or "1.2MB) drive. The
former will usually produce better results.
If the first sector on the track is consistently missing (not always
the lowest-numbered; some systems skew the tracks for better
performance), you might be able to recover it by slowing the drive a
bit to allow the first sector IDAM to move outside of the 765 "blind
spot" at the beginning of a track (although the 765 formats a track
with an IAM at the start, it never reads it).
Clean heads, as Dwight's mentioned, also count for a lot.
Cheers,
Chuck
2:13 p.m.
Chuck Guzis wrote:
You didn't mention if you were using a 48tpi
("360K") drive to read
these or trying to get by on a 96tpi ("720K" or "1.2MB) drive. The
former will usually produce better results.
Is that true? I almost had this discussion with Dave a couple of weeks ago (I
was thinking it might be nice in the Imagedisk archive format recorded the
double step flag setting in the file to provide a hint as to the drive type
[1] used)
My theory (which I don't think I ever got around to explaining to Dave) was
that the head in a 40-cylinder drive would be better equipped to read a
*marginal* signal from a 48TPI disk than an 80-cylinder double-stepped drive,
purely due to the head width.
Then I got to thinking that maybe that was utter crap :-)
[1] I'm with Dave in that I think it would be wrong to restrict the recreation
of an archive based upon the drive type that created it. However, I do tend to
favour as much (automatic) recording of the parameters that were in use when
data is created as possible.
4:09 p.m.
On 15 May 2007 at 13:13, Jules Richardson wrote:
Is that true? I almost had this discussion with Dave a
couple of weeks ago (I
was thinking it might be nice in the Imagedisk archive format recorded the
double step flag setting in the file to provide a hint as to the drive type
[1] used)
In my experience, yes, it's definitely true. I suspect it's more a
matter of tolerating small misalignments better than a matter of
better S/N ratio.
I also reserve a couple of drives for "tough cases". Not all 360K
drives perform the same when recovering marginal media and I reserve
the very good drives for special occasions.
One thing that few people realize is that not all drives are equally
competent when it comes to clamping media "dead center". Almost all
(I have some very old drives that don't) spin the motor as the disk
is being clamped by the hub mechanism, but some do the clamping job
better.
Cheers,
Chuck
2:26 p.m.
It's entirely possible that the drive used to
create the diskettes
was somewhat out of alignment. You may have to "unalign" a drive to
successfully read these. This is more common than you might think.
Btw the Align/Test function of ImageDisk is very useful for this - it
you can seek to a cylinder and get an audible beep which increases
with the number of sectors IDs found match the cylinder - very handy
when you need to align a drive to a disk (which I've had to do a number
of times).
If the first sector on the track is consistently
missing (not always
the lowest-numbered; some systems skew the tracks for better
performance), you might be able to recover it by slowing the drive a
bit to allow the first sector IDAM to move outside of the 765 "blind
spot" at the beginning of a track (although the 765 formats a track
with an IAM at the start, it never reads it).
I've had to do this for a number of systems, Cromemco being the most
notable ... I find slowing the drive by about 10rpm makes a huge
difference on some disks (One of the reasons I like TEAC drives -
very easy to set the speed).
And yes - 10x512 is quite common, and can be accomodated with a
765.
Regards,
Dave
--
dave06a (at) Dave Dunfield
dunfield (dot) Firmware development services & tools: www.dunfield.com
com Collector of vintage computing equipment:
http://www.classiccmp.org/dunfield/index.html
3:52 p.m.
Dave Dunfield wrote:
I've had to do this for a number of systems,
Cromemco being the most
notable ... I find slowing the drive by about 10rpm makes a huge
difference on some disks (One of the reasons I like TEAC drives -
very easy to set the speed).
That's nice to know - FWICT my modified YE Data drive is running at about
290RPM instead of 300 (based on the index pulse timing). Now I don't have to
try and figure out how to speed it up by another 10RPM :)
That said, the drive motor appears to be crystal-locked, so increasing the RPM
might be a little difficult anyway.
BTW Dave, thanks for the page on modifying 5.25" drives for 300RPM - the
method you mentioned worked like a charm. I'd post the wire colours for the
high/low speed switch, but the drive's downstairs. Basically, there's one line
that's always at +5.0V, one that's always +12.0V, one that's on/off depending
on the motor's state (on or off) and one that sits at around 4V. Ground the 4V
line at the motor controller board (and disconnect said wire from the drive
controller/head amp board first!) to get 300RPM.
--
Phil. | (\_/) This is Bunny. Copy and paste Bunny
classiccmp at philpem.me.uk | (='.'=) into your signature to help him gain
http://www.philpem.me.uk/ | (")_(") world domination.
10:10 p.m.
That's nice to know - FWICT my modified YE Data
drive is running at about
290RPM instead of 300 (based on the index pulse timing). Now I don't have to
try and figure out how to speed it up by another 10RPM :)
That said, the drive motor appears to be crystal-locked, so increasing the RPM
might be a little difficult anyway.
Thats odd - I have a number of drives that are crystal controlled and non-
adjustable, and they are all "bang on" 300 or 360 rpm ... Thats a pretty
big error.
BTW Dave, thanks for the page on modifying 5.25"
drives for 300RPM - the
method you mentioned worked like a charm. I'd post the wire colours for the
high/low speed switch, but the drive's downstairs. Basically, there's one line
that's always at +5.0V, one that's always +12.0V, one that's on/off depending
on the motor's state (on or off) and one that sits at around 4V. Ground the 4V
line at the motor controller board (and disconnect said wire from the drive
controller/head amp board first!) to get 300RPM.
The exact wire colors very from drive to drive which is why I describe the
method (pretty much what you describe above) to find the right one...
Dave
--
dave06a (at) Dave Dunfield
dunfield (dot) Firmware development services & tools: www.dunfield.com
com Collector of vintage computing equipment:
http://www.classiccmp.org/dunfield/index.html
2:39 p.m.
On Tue, 15 May 2007, Chuck Guzis wrote:
I hadn't realized that there was a norm for CP/M
floppies. You know,
the "recommended" 512-byte sector "packing" in the uPD765 literature
is 8 sectors--and indeed, early PC-DOS used just that.
I once asked Gary Kildall, "What is the standard format for 5.25" disks
with double density?"
He replied, " 8" single density."
If the first sector on the track is consistently
missing (not always
the lowest-numbered; some systems skew the tracks for better
performance), you might be able to recover it by slowing the drive a
bit to allow the first sector IDAM to move outside of the 765 "blind
spot" at the beginning of a track (although the 765 formats a track
with an IAM at the start, it never reads it).
For that same problem, I prefer "masking off" the index pulse. As was
discussed earlier, that can be done in almost any way that permits
interrupting that signal in the cable, or, with MOST drives (but NOT
Teac!), by covering the index hole of the disk jacket.
--
Grumpy Ol' Fred cisin at xenosoft.com
6272
days inactive
6274
days old
8 comments
7 participants
participants (7)
-
cclist＠sydex.com -
cisin＠xenosoft.com -
classiccmp＠philpem.me.uk -
dave06a＠dunfield.com -
dkelvey＠hotmail.com -
john_a_s2004＠hotmail.com -
julesrichardsonuk＠yahoo.co.uk