Date   
Re: MS2000 MicroDos Development System

David Schultz
 

On 12/10/19 1:45 PM, ajparent1/kb1gmx wrote:
David,

IS the copy of MPM241 yu used for source the same as the copy at
BITSAVERS.org?
After scanning I provided a copy to bitsavers.


--
https://web.archive.org/web/20190214181851/http://home.earthlink.net/~david.schultz/
(Web pages available only at the Wayback Machine because Earthlink
terminated that service.)

Re: MS2000 MicroDos Development System

ajparent1/kb1gmx
 

David,

IS the copy of MPM241 yu used for source the same as the copy at BITSAVERS.org?

As in here:
http://bitsavers.trailing-edge.com/components/rca/cosmac/ms2000/MPM-241_RCA_MicroDisk_Development_System_MS2000_User_Manual_May84.pdf

Allison

Re: MS2000 MicroDos Development System

David Schultz
 

On 12/10/19 1:25 PM, ajparent1/kb1gmx wrote:
We have the clue that SYSGEN creates a disk and system (bootable) disks
have less user space.  Totally guessing, the track 0, sector 1 disk ID
has a lot
of free space and that has the system pointers to where the system is on
the
disk outside of the directory.
See TKTABL in UT71.

TKTABL DB 1,2,28 ;TABLE CONTAINS TRACK #. STARTING
DB 2,0,36 ;SECTOR # (-1). AND BYTE COUNT (X128)
DB 3,0,32 ;FOR ALL MICRODOS LOAD
DB 0

That is where the regular file OP.SYS should be. UT71 does a sanity
check (MICTST) after loading.

--
https://web.archive.org/web/20190214181851/http://home.earthlink.net/~david.schultz/
(Web pages available only at the Wayback Machine because Earthlink
terminated that service.)

Re: MS2000 MicroDos Development System

ajparent1/kb1gmx
 

after finding MPM241 in the complete form...

The ODS is a variant of the disk inode format from the days of V6 or 7ish
and ealry linux.

There is a:

Directory on track 00, disk sector 1 disk ID, sector 2-9 file directory

Track 1 sector 1, CAT which maps the used clusters (cluster=sector)
  1 bit per used sector 1=used.   allocation is LBA (not track sector).

Then the file start pointer sector is the SSN (start sector number) locating
the RIB.

The RIB is the mapping of used sectors (called SDW or segment descriptor)

That's a cut to the bone and minimal description.
Also applies to data only disks...

70 Tracks??
No clue why they only used 70 tracks, I suspect it was because the drives
looked like the oddball 100tpI (not 46, 96, or 135!) ones that appeared and
disappeared.  Or a religious thing back then being the inner tracks were
not generally good to use, likely because someone had issues with theirs.
The only mention of 70 tracks was infoworld in 1984 and the article got a
lot of critical attention being mostly wrong.

Gleaned from MPM241 page 89-92:
A small clue is that the OS uses about 4% of the space and apparently there
is an OS disk that has roughly 25Kbytes reserved (about 50 sectors or 5-6 tracks)
if we add that to the directory and CAT that 7-10 sectors used for "housekeeping".
We have the clue that SYSGEN creates a disk and system (bootable) disks
have less user space.  Totally guessing, the track 0, sector 1 disk ID has a lot
of free space and that has the system pointers to where the system is on the
disk outside of the directory.

So unless I unearth something its a mystery.

Allison

Re: MS2000 MicroDos Development System

cellarcat
 

This is from Appendix A in the MS2000 Manual:

Each diskette has 70 tracks with 9 sectors on each track (630 sectors per diskette). However, from MicroDOS's point ofview, the diskette is divided up into clusters with 1 sector in each cluster. The system diskette has two basic configurations, one with a directory and operating system and one with a directory only. These configurations are generated with the SYSGEN command. Because the operating system requires about 4 per cent of the diskette, diskettes with directory only have more disk area for storage of the user's work files. MicroDOS assumes that a me is a string of bytes. When a me is created, a certain number of clusters is allocated to it. If more space is needed for the data than initially allocated, MicroDOS automatically allocates more space. Once a me has been created by the user, the operating system returns to the system any unused disk cluster so that the next file to be created can use this freed-up space. No cluster can be allocated to two different files. 



On Dec 10, 2019, at 11:28 AM, ajparent1/kb1gmx wrote:

cellarcat,

That is one of the stranger drives.  All of the 3.5" drives I ahve (maybe 2 dozen or more) 
are either the early 40 track or later 80.  For that size i'd never seen 70 tracks.

For the 5.25size there was an oddball that tied to be a 77track (like 8") but tht was 
a rare beast.  Again there was a 5.25 that was two sided 35 cylinders (70 tracks).
However that was short lived and went to 40 cylinders to stay in line with
Sa400b/TM100 and others and the two sided versions SA450 were 40 cylinders
and two sided.  Same for the half height 5.25" drives like TEAC FD55 series
(both 40 or 80 track versions were both available in single sided and two sided).

Micropolus, Inovex has a few oddbals and there were the 3.0inch drives.
Most oall of those were extinct before they got traction.

What maybe the case is RCA used 10 tracks for system tracks or other hidden uses.
We can speculate....
An example of that is 8" SSSD (Single sided single density) was 77 tracks but the
first two (0 and 1) were reserved for the system and the directory and data were
stored on track 2 through 76 netting 75 usable tracks.  That practice was retained
even on CP/m on 1.44mb 3.5" floppies where two tracks would easily store the
system twice over.  Why?  Tradition!  No other good reason.  FYI the only reasonable
one I've heard is that keeping the directory off the outer track (000) was a bit more
robust but over time that proved unlikely.

FYI: I have not seen any information on the ODS (on disk structure) or organization
of how the tracks and sectors are used.   I could offer that MAYBE a far stretch the
first 10 tracks were boot tracks for the system (maybe 2-3tracks worth) and directory
for up to 80 files where each sector has a file name, details (meta), and allocated
spaces listed in that 512byte block. That's a wacky scheme but I've seen
many others.  Without the ODS information it is speculation.

So ODS?  Where did you get he Microdos 70track info?

Allison

Re: MS2000 MicroDos Development System

ajparent1/kb1gmx
 

cellarcat,

That is one of the stranger drives.  All of the 3.5" drives I ahve (maybe 2 dozen or more) 
are either the early 40 track or later 80.  For that size i'd never seen 70 tracks.

For the 5.25size there was an oddball that tied to be a 77track (like 8") but tht was 
a rare beast.  Again there was a 5.25 that was two sided 35 cylinders (70 tracks).
However that was short lived and went to 40 cylinders to stay in line with
Sa400b/TM100 and others and the two sided versions SA450 were 40 cylinders
and two sided.  Same for the half height 5.25" drives like TEAC FD55 series
(both 40 or 80 track versions were both available in single sided and two sided).

Micropolus, Inovex has a few oddbals and there were the 3.0inch drives.
Most oall of those were extinct before they got traction.

What maybe the case is RCA used 10 tracks for system tracks or other hidden uses.
We can speculate....
An example of that is 8" SSSD (Single sided single density) was 77 tracks but the
first two (0 and 1) were reserved for the system and the directory and data were
stored on track 2 through 76 netting 75 usable tracks.  That practice was retained
even on CP/m on 1.44mb 3.5" floppies where two tracks would easily store the
system twice over.  Why?  Tradition!  No other good reason.  FYI the only reasonable
one I've heard is that keeping the directory off the outer track (000) was a bit more
robust but over time that proved unlikely.

FYI: I have not seen any information on the ODS (on disk structure) or organization
of how the tracks and sectors are used.   I could offer that MAYBE a far stretch the
first 10 tracks were boot tracks for the system (maybe 2-3tracks worth) and directory
for up to 80 files where each sector has a file name, details (meta), and allocated
spaces listed in that 512byte block. That's a wacky scheme but I've seen
many others.  Without the ODS information it is speculation.

So ODS?  Where did you get he Microdos 70track info?

Allison

Re: MS2000 MicroDos Development System

cellarcat
 

Thanks, Dave! I think I will just fire up Linux and follow your instructions.

On Dec 9, 2019, at 6:26 PM, David Schultz <david.schultz@...> wrote:

On 12/8/19 11:51 PM, cellarcat wrote:
I should have looked more closely at the origins of the .img file. Dave
Schultz produced it and Dave, I assume, from reading your web page that
you used dd under Linux. I always thought dd and rawrite were similar
and could read each other's .img files but maybe not!
I have no idea what rawrite is looking for but that disk image is just
that, an image. It contains zero information on the disk geometry and is
just a linear collection of the disk sectors and tracks.


My notes say that this is how I read the data:

sudo ./setfdprm -p /dev/fd0 SS DD sect=9 cyl=80 ssize=512
sudo dd if=/dev/fd0 of=image bs=512 count=630




--
https://web.archive.org/web/20190214181851/http://home.earthlink.net/~david.schultz/
(Web pages available only at the Wayback Machine because Earthlink
terminated that service.)


Re: MS2000 MicroDos Development System

cellarcat
 

Thanks, Allison! I think the 630 may be because the Microdos disks are actually 70 tracks and not 80. Why on earth RCA would do that is beyond me unless the Sony drives in the MSIM50 unit are limited to 70 tracks. I would really like to know what drives they used. 

On Dec 9, 2019, at 8:01 PM, ajparent1/kb1gmx <kb1gmx@...> wrote:

On Mon, Dec 9, 2019 at 04:26 PM, David Schultz wrote:
sudo ./setfdprm -p /dev/fd0 SS DD sect=9 cyl=80 ssize=512
sudo dd if=/dev/fd0 of=image bs=512 count=630
OK so you set the FD parameters to:
9 sectors of 512 bytes (double density clock rate) and
80 cylinders single sided.

That is the geometry.
IF we do the math  we get 720 blocks of 512 bytes.  DD will read them
sequentially from start of media to end of media.

I am mystified that the count is 630 not 720 (720=9x80)?

But if we do the reverse of that to a formatted but bank disk with:

sudo ./setfdprm -p /dev/fd0 SS DD sect=9 cyl=80 ssize=512
Establish parameters again...

sudo dd if=image o
f=/dev/fd0 bs=512 count=630
"if" is the input file "image" and "of" is the output or /dev/fd0.

That should create a copy of the originating disk in full fidelity.
There was no interpreting of the sector content or order other than
sequential read and write.

DD and rawrite are similar but that's about it.  How to talk to them
differs.  However rawwrite should have the parameters from teh above somehow.

One last thing the media (floppy) must be prefromatted with 9 sectors,
512 bytes sector, for 80 tracks starting at 000.

Allison

Re: MS2000 MicroDos Development System

David Schultz
 

On 12/9/19 8:01 PM, ajparent1/kb1gmx wrote:
I am mystified that the count is 630 not 720 (720=9x80)?
Read errors on the last few tracks.


--
https://web.archive.org/web/20190214181851/http://home.earthlink.net/~david.schultz/
(Web pages available only at the Wayback Machine because Earthlink
terminated that service.)

Re: MS2000 MicroDos Development System

ajparent1/kb1gmx
 

On Mon, Dec 9, 2019 at 04:26 PM, David Schultz wrote:
sudo ./setfdprm -p /dev/fd0 SS DD sect=9 cyl=80 ssize=512
sudo dd if=/dev/fd0 of=image bs=512 count=630
OK so you set the FD parameters to:
9 sectors of 512 bytes (double density clock rate) and
80 cylinders single sided.

That is the geometry.
IF we do the math  we get 720 blocks of 512 bytes.  DD will read them
sequentially from start of media to end of media.

I am mystified that the count is 630 not 720 (720=9x80)?

But if we do the reverse of that to a formatted but bank disk with:

sudo ./setfdprm -p /dev/fd0 SS DD sect=9 cyl=80 ssize=512
Establish parameters again...

sudo dd if=image o
f=/dev/fd0 bs=512 count=630
"if" is the input file "image" and "of" is the output or /dev/fd0.

That should create a copy of the originating disk in full fidelity.
There was no interpreting of the sector content or order other than
sequential read and write.

DD and rawrite are similar but that's about it.  How to talk to them
differs.  However rawwrite should have the parameters from teh above somehow.

One last thing the media (floppy) must be prefromatted with 9 sectors,
512 bytes sector, for 80 tracks starting at 000.

Allison

Re: MS2000 MicroDos Development System

David Schultz
 

On 12/8/19 11:51 PM, cellarcat wrote:
I should have looked more closely at the origins of the .img file. Dave
Schultz produced it and Dave, I assume, from reading your web page that
you used dd under Linux. I always thought dd and rawrite were similar
and could read each other's .img files but maybe not!
I have no idea what rawrite is looking for but that disk image is just
that, an image. It contains zero information on the disk geometry and is
just a linear collection of the disk sectors and tracks.


My notes say that this is how I read the data:

sudo ./setfdprm -p /dev/fd0 SS DD sect=9 cyl=80 ssize=512
sudo dd if=/dev/fd0 of=image bs=512 count=630




--
https://web.archive.org/web/20190214181851/http://home.earthlink.net/~david.schultz/
(Web pages available only at the Wayback Machine because Earthlink
terminated that service.)

Re: Software serial and overclocked 1802's #microboards #system1

Lee Hart
 

Mark Moulding wrote:
Jameco is now selling some of the same fake remarked ICs that you
can get on ebay. I bought a batch of 32k RAMs from them that were
obviously re-marked, and all bad. Caveat emptor./

I hope you called them about this. Jameco is a family-owned company
(just down the street from me, in fact), and I've found that they are
quite responsive to customer feedback.
I did; and they refunded my money. But it caused extra grief, because I had shipped a couple to customers who found them bad.

I really wish that they'd re-spin their god-awful web site.
Agreed. Its search capability is pitiful, and it's hard to find things that I *know* they have because there are so many errors in their descriptions.

I too still use their paper catalog (which is great).

Lee Hart
--
ICEs have the same problem as lightbulbs. Why innovate and make
better ones when the current ones burn out often enough to keep
you in business? -- Hunter Cressall
--
Lee Hart, 814 8th Ave N, Sartell MN 56377, www.sunrise-ev.com

Re: Software serial and overclocked 1802's #microboards #system1

Mark Moulding
 

But I have to say... they are now getting into el-crappo Chinese junk parts. Jameco is now selling some of the same fake remarked ICs that you can get on ebay. I bought a batch of 32k RAMs from them that were obviously re-marked, and all bad. Caveat emptor.

I hope you called them about this.  Jameco is a family-owned company (just down the street from me, in fact), and I've found that they are quite responsive to customer feedback.  I have noticed this trend from them also; I just bought some quarter-watt resistors from them, and it's amazing just how thin the lead wires are...

I really wish that they'd re-spin their god-awful web site.  I spoke with one of their executives about it, and they're well aware that its usability is poor, especially compared to the exemplary models of DigiKey or Mouser.  Apparently they took a wrong turn many years ago, and creating a new web site would be prohibitively expensive.  Frankly, I believe that it's currently so bad that they're costing themselves a lot in lost sales, even though their prices are very competitive.  When I'm looking for some components, I actually find it easier and definitely more reliable to look at their paper catalog rather than try to find it on-line.

I believe that a simple cleansing of their data, with their current web site, would help a great deal (same searchable fields for all similar components, correct and consistent value units for reliable searching, etc.) would go a long way towards improving it at a relatively low cost.  It shouldn't take too many man hours to go manually through their 50,000 item catalog, with a data entry person and a knowledgeable engineer.
~~

Mark Moulding

Re: Software serial and overclocked 1802's #microboards #system1

Lee Hart
 

ajparent1/kb1gmx wrote:
Already been there. Jameco has been a long time supplier I've used
going back to the 70s.
Same for me. I borrowed one of their 1970's ads from an old Byte magazine to use in my Z80MC manual (http://sunrise-ev.com/z80.htm page 13). You can see they had just about all the same parts needed for vintage computing today (and their prices).

Jameco has always sold surplus; manufacturer over-runs, stuff they bought at an auction when a company went out of business, etc. But I have to say... they are now getting into el-crappo Chinese junk parts. Jameco is now selling some of the same fake remarked ICs that you can get on ebay. I bought a batch of 32k RAMs from them that were obviously re-marked, and all bad. Caveat emptor.

Looking at the various board there is a lot of repeated circuitry
(mostly for bus) so I'm
looking at if I do this how to cut the total parts
One feature of CMOS is its negligible input current. You often don't need bus drivers; that can save a lot of chips. You just get more capacitive loading, so the ultimate maximum speed isn't as high.

Lee

--
ICEs have the same problem as lightbulbs. Why innovate and make
better ones when the current ones burn out often enough to keep
you in business? -- Hunter Cressall
--
Lee Hart, 814 8th Ave N, Sartell MN 56377, www.sunrise-ev.com

Re: Software serial and overclocked 1802's #microboards #system1

ajparent1/kb1gmx
 

Already been there.  Jameco has been a long time supplier I've used going back to the 70s.

Looking at the various board there is a lot of repeated circuitry (mostly for bus) so I'm
looking at if I do this how to cut the total parts to soem much smaller amount but
full ram/eprom as needed.  The FDC board was a cluttered mess, that was a an
easy reduction with the first section to go being the PLL data separator it was
complex and tended to misbehave.

Allison

Re: MS2000 MicroDos Development System

cellarcat
 

I should have looked more closely at the origins of the .img file. Dave Schultz produced it and Dave, I assume, from reading your web page that you used dd under Linux. I always thought dd and rawrite were similar and could read each other's .img files but maybe not!

Re: MS2000 MicroDos Development System

Lee Hart
 

cellarcat wrote:
I tried using the microdos.img file from our file section with rawrite3
but it comes back with the message that it cannot figure out how many
sectors per track for the new diskette. Does anybody know what program
was used to write the original .img file?

I used Dave Dunfield's IMD to make .imd copies of my two original RCA
disks but one of them was corrupt and the other is suspect. Hardly
surprising considering that they are 35 years old.
I went through the same thing to recover Ron Cenker's source code for BASIC3. He supplied it to me on an 8" floppy from an RCA development system. But it was damaged, so I had to laboriously read all the sectors and piece them together again. It was a joint effort of many people on this list. But we succeeded in resurrecting it!

Lee

--
ICEs have the same problem as lightbulbs. Why innovate and make
better ones when the current ones burn out often enough to keep
you in business? -- Hunter Cressall
--
Lee Hart, 814 8th Ave N, Sartell MN 56377, www.sunrise-ev.com

Re: MS2000 MicroDos Development System

cellarcat
 

I tried using the microdos.img file from our file section with rawrite3 but it comes back with the message that it cannot figure out how many sectors per track for the new diskette. Does anybody know what program was used to write the original .img file?

I used Dave Dunfield's IMD to make .imd copies of my two original RCA disks but one of them was corrupt and the other is suspect. Hardly surprising considering that they are 35 years old.

Re: Software serial and overclocked 1802's #microboards #system1

cmdrcosmac
 


Allison,
See:

https://www.jameco.com/z/CDP1854AE-INTERSIL-CORPORATION-Programmable-Universal-Asynchronous-Receiver-Transmitter-UART-DIP-40_2290543.html

They may not have many left. You'll need one to run UT71. I ordered some and the worked fine. Marked RCA CDP1854AE RCA Z 745.
-Chuck

Re: Software serial and overclocked 1802's #microboards #system1

ajparent1/kb1gmx
 

That and bitbash and performance in the same sentence is only for 
contrast or a discussion of electrical simplicity.

The 1854s... gotta get a few.

Allison