Re: HP 9816 CP/M-68K

[Eric Smith via cctalk]

On Mon, Feb 12, 2018 at 3:56 PM, js--- via cctalk 
wrote:

> That's really "slick," Glen.If it's not too burdensome to give a brief
> answer, how would you keep track of the time, or know how long feeding a
> byte at a time took?
>

On an original PC or XT (without special turbo modes of clones), you could
probably measure time well enough by counting CPU cycles. Otherwise you'd
need a timer. However, if you make the decision to reset the FDC based on
when it asks for data, you just reset it when it asks for the sector
information for the next sector past the last one that you want to format.
(I haven't actually tried that. I used a timer.)

Re: HP 9816 CP/M-68K

[Guy Sotomayor Jr via cctalk]

> On Feb 12, 2018, at 3:00 PM, Chuck Guzis via cctalk  
> wrote:
> 
> 
> Most don't believe me what I say that the floppy disc dates from 1946.

And the fax was invented in 1842 (yes *before* the telephone)!

TTFN - Guy

Re: HP 9816 CP/M-68K

[Chuck Guzis via cctalk]

On 02/12/2018 01:55 PM, Fred Cisin via cctalk wrote:

> Thank you!
> Although these days, it's just curiousity and nostalgia.

Yup.  The shame is that some of the more advanced 765-based FDCs came in
just as the floppy was sunsetting.  For example, the Intel 82078 boasts,
among other things, a 2Mbps data rate option, a 16 byte FIFO and a
"Format and Write" command for single-pass copying.

Some young'uns can still identify a floppy disk, but few have seen the
8" variety.

Few realize that tape and floppy are about the same age.  The US
military discovered the German Magnetophon tape recorder, manufactured
during WW II from about 1939.  In 1946, Brush Corporation made US tape
recorders (the Sound Mirror) and what is best termed a "floppy disk
recorder", the Mail-A-Voice.

Most don't believe me what I say that the floppy disc dates from 1946.

--Chuck

Re: HP 9816 CP/M-68K

[js--- via cctalk]

On 2/12/2018 3:52 PM, Glen Slick via cctalk wrote:

On Mon, Feb 12, 2018 at 12:34 PM, Eric Smith via cctalk
  wrote:

Format a track with the sector size that occurs later on the track, with
dummy sectors ahead of them and gap sizes selected to position them
properly.
Start formatting with the sector size for the earlier sectors. Abort the
format at the time when the desired number of sectors have been written.

I'm not sure whether there's any way to abort a track format on a PC. I did
it on a machine that had control over the μPD765 reset pin.


Bit 2 of the Digital-Output Register at 0x3F2 resets the FDC when the
bit is clear.

I used that on my 5150 back in the day when I bought it new to
reproduce a copy protected disk that used sectors with bad CRCs as one
of its protection mechanisms. Feed the sector data a byte at a time to
the FDC in PIO mode so that the timing is controlled, then toggle the
FDC reset bit in the DOR just at the time when the CRC is being
written.


That's really "slick," Glen.If it's not too burdensome to give a 
brief answer, how would you keep track of the time, or know how long 
feeding a byte at a time took?


- John

Re: HP 9816 CP/M-68K

[Fred Cisin via cctalk]

But, I don't know how to FORMAT a track with multiple sector sizes with
NEC 765 type controller.  Not as hard with WD style controllers.


On Mon, 12 Feb 2018, Chuck Guzis via cctalk wrote:

It's actually not too difficult.  There were DOS utilities that could,
for example, copy IBM XDF format.
The trick is that on the 765 and its ilk, the FORMAT command writes
sector headers according to the length parameter in the *command*,
whereas the WRITE SECTOR command writes things out according to the
length in the sector ID header.


Thank you!
Although these days, it's just curiousity and nostalgia.

I learn great new [currently useless to me] fun stuff on this list every 
day.


20+ years ago, when I needed to, after a number of failed attempts,
I used double sided (and 96tpi) drives in a TRS80 model 3,
and/or "COPY"ed a disk using a flux transition board.

Re: HP 9816 CP/M-68K

[Chuck Guzis via cctalk]

On 02/12/2018 12:23 PM, Fred Cisin via cctalk wrote:
>> Imagedisk can't handle mixed sector sizes on a track.
> 
> Reading or writing multiple sized sectors can be done with multiple passes.
> 
> But, I don't know how to FORMAT a track with multiple sector sizes with
> NEC 765 type controller.  Not as hard with WD style controllers.

It's actually not too difficult.  There were DOS utilities that could,
for example, copy IBM XDF format.

The trick is that on the 765 and its ilk, the FORMAT command writes
sector headers according to the length parameter in the *command*,
whereas the WRITE SECTOR command writes things out according to the
length in the sector ID header.

So, for example, you can specify that, say, 18 256 byte sectors are to
be formatted, but each ID header would have a code of 3 or 1, indicating
that the data part of the sector is 1024 or 256 bytes.  Pick the
appropriate sectors to write and you'll clobber any of the 256 byte ones
that fall under the 1024 byte shadow of the larger ones.

It does take a bit of planning, but it does work.

In fact, this is a handy way to "wipe" a track free of any ID
information at all.  Just format up a couple of sectors with a length
code of 6 or 7 and then write them.  They'll eat their own heads,
ignoring the index.

FWIW,
Chuck

Re: HP 9816 CP/M-68K

[Fred Cisin via cctalk]

But, I don't know how to FORMAT a track with multiple sector sizes with
NEC 765 type controller.  Not as hard with WD style controllers.


On Mon, 12 Feb 2018, Eric Smith wrote:

Format a track with the sector size that occurs later on the track, with
dummy sectors ahead of them and gap sizes selected to position them
properly.
Start formatting with the sector size for the earlier sectors. Abort the
format at the time when the desired number of sectors have been written.
I'm not sure whether there's any way to abort a track format on a PC. I did
it on a machine that had control over the ??PD765 reset pin.


Thank you!


Definitely much easier with the WD style controllers.
It is possible to format disks on an NEC style controller with a format
that cannot be created on a WD style controller, for instance using
particular track or sector numbers above 0xf0, which are specially
interpreted by the WD during a track write.


I still feel that "multiple sector read/write" on NEC765 is less useful 
than "track read/write" of WD 179x.

In spite of WD's handling of what it interprets as address marks, etc.
'Course OUR needs are not always the same as the general public's, . . .


Is there any truth to the rumors that the choices of Data Address Marks 
(DAM-it!) on TRS80 was due to a misprinted or misread data sheet?
And that Tandy wasn't a big enough player for WD to consider a variant of 
the 179X that could support those DAMs?
We were not amused that the Model 3 could not write a true model 1 disk, 
and that Doubler, etc. had to contain BOTH 177x and 179x chips!

Re: HP 9816 CP/M-68K

[Glen Slick via cctalk]

On Mon, Feb 12, 2018 at 12:34 PM, Eric Smith via cctalk
 wrote:
>
> Format a track with the sector size that occurs later on the track, with
> dummy sectors ahead of them and gap sizes selected to position them
> properly.
> Start formatting with the sector size for the earlier sectors. Abort the
> format at the time when the desired number of sectors have been written.
>
> I'm not sure whether there's any way to abort a track format on a PC. I did
> it on a machine that had control over the μPD765 reset pin.
>

Bit 2 of the Digital-Output Register at 0x3F2 resets the FDC when the
bit is clear.

I used that on my 5150 back in the day when I bought it new to
reproduce a copy protected disk that used sectors with bad CRCs as one
of its protection mechanisms. Feed the sector data a byte at a time to
the FDC in PIO mode so that the timing is controlled, then toggle the
FDC reset bit in the DOR just at the time when the CRC is being
written.

Re: HP 9816 CP/M-68K

[Al Kossow via cctalk]

On 2/12/18 12:23 PM, Fred Cisin via cctalk wrote:
>> Imagedisk can't handle mixed sector sizes on a track.
> 
> Reading or writing multiple sized sectors can be done with multiple passes.

I don't believe the .imd image format supports it either, so I just found 
another
tool that could handle the job.

I really need to look at getting the Supercard Pro I bought set up so I can do 
lower-level imaging

Re: HP 9816 CP/M-68K

[Eric Smith via cctalk]

On Mon, Feb 12, 2018 at 1:23 PM, Fred Cisin via cctalk <
cctalk@classiccmp.org> wrote:

> Reading or writing multiple sized sectors can be done with multiple passes.
> But, I don't know how to FORMAT a track with multiple sector sizes with
> NEC 765 type controller.  Not as hard with WD style controllers.
> With multiple sector sizes, can squeeze 440K on a "360K" disk.
>

Format a track with the sector size that occurs later on the track, with
dummy sectors ahead of them and gap sizes selected to position them
properly.
Start formatting with the sector size for the earlier sectors. Abort the
format at the time when the desired number of sectors have been written.

I'm not sure whether there's any way to abort a track format on a PC. I did
it on a machine that had control over the μPD765 reset pin.

Definitely much easier with the WD style controllers.

It is possible to format disks on an NEC style controller with a format
that cannot be created on a WD style controller, for instance using
particular track or sector numbers above 0xf0, which are specially
interpreted by the WD during a track write.

Re: HP 9816 CP/M-68K

[Fred Cisin via cctalk]

Imagedisk can't handle mixed sector sizes on a track.


Reading or writing multiple sized sectors can be done with multiple 
passes.


But, I don't know how to FORMAT a track with multiple sector sizes with 
NEC 765 type controller.  Not as hard with WD style controllers.



With multiple sector sizes, can squeeze 440K on a "360K" disk.

Re: HP 9816 CP/M-68K

[Al Kossow via cctalk]

On 2/11/18 7:35 PM, Glen Slick via cctalk wrote:

> Any technical reason you used Teledisk instead of ImageDisk? Can
> Teledisk do something with these disks that ImageDisk cannot?
> 

Imagedisk can't handle mixed sector sizes on a track.

Re: HP 9816 CP/M-68K

[Glen Slick via cctalk]

On Sun, Feb 11, 2018 at 7:12 PM, Al Kossow via cctalk
 wrote:
>
> also, I may have discovered a bug in the HcX disk analysis tool
> the track map really confused by the .td0 images I created, though they 
> appear to work
> when written to a new disk

Any technical reason you used Teledisk instead of ImageDisk? Can
Teledisk do something with these disks that ImageDisk cannot?

Re: HP 9816 CP/M-68K

[Al Kossow via cctalk]

got it running this afternoon
it won't run on a 9000/216 with 256k, adding a 1meg card made it happy (ie. 
programs load and run)
still need to figure out why B: isn't working
the memory test with 1.25mb is slw


also, I may have discovered a bug in the HcX disk analysis tool
the track map really confused by the .td0 images I created, though they appear 
to work
when written to a new disk


On 2/9/18 12:05 PM, Al Kossow via cctalk wrote:
> floppies recovered and uploaded to 
> http://bitsavers.org/bits/HP/HP_9000/cpm-68k
> 
> I'm pretty sure this will only work in a 9121 single-sided drive but I'll be 
> trying
> to boot it soon
> 
> 

Re: HP 9816 CP/M-68K

[Paul Berger via cctalk]

On 2018-02-09 4:05 PM, Al Kossow via cctalk wrote:

floppies recovered and uploaded to http://bitsavers.org/bits/HP/HP_9000/cpm-68k

I'm pretty sure this will only work in a 9121 single-sided drive but I'll be 
trying
to boot it soon


I would suspect the issue is that it only support Amigo devices so it 
may also work with 9895 or 82901/2  like other early 68K 9800 software 
like HPL, but yeah the boot disk is would seem to be for a 9121 or 
possibly one of the 5.25 drive they quite conveniently have the same 
number of tracks per disk.


Paul.