19 May
2014
19 May
'14
4:26 p.m.
On 05/19/2014 01:34 PM, Enrico Lazzerini wrote:
Hi, I have an old IBM 5170 with motherboard Type 1
without any old hard
drive.
I could can also reprogram the EPROM system to enter different geometries of
disks, but was wondering, how could I possibly replace the disc with a
cf-IDE40 adapter like this:
http://www.idotpc.com/TheStore/pc/catalog//products/accessories/cfide401big.
jpg
<http://www.idotpc.com/TheStore/pc/catalog/products/accessories/cfide401big.
jpg> Thanks for any suggestions.
You're going to need an ISA IDE adapter with its own BIOS to use with an
IDE drive or an IDE-to-CF card adapter (Promise made some of those).
You might also be able to use a BIOS-less IDE adapter if you use a DDO
(dynamic driver overlay).
Here's another option:
http://www.lo-tech.co.uk/wiki/XT-CF-lite
Note that you'll still need a floppy controller.
Hope this helps,
Chuck
20 May
20 May
2:31 p.m.
You're going to need an ISA IDE adapter with its
own BIOS to use with an
IDE drive or an IDE-to-CF card adapter (Promise made some of those).
Strictly you don't ened a BIOS extension. IDE drives were designed ot be
compatible wit hthe Western Digital controller in the original PC/AT. If
you can modify the dirve geometry tabel [1] then it will work with the
drivers in the BIOS ROMs.
[1] It is possible to ad a singel TTL IC to disable the normal ROMs for
the last 8K words and map in 8K EPROMs (2764 type) inserted in the spare
sockets on the motherboard. IIRC, the modification doens't even involve
any track cuts, there is a helpful shunt block you cna use to intercept
the ROM select signals.
FWIW, the machine I am typing thsi on has an IDE drive, a bus interdace
[2] with no BIOS extensions and the mod made to change the disk parameter
tables. So it certainly works.
-tony
3:45 p.m.
On 05/20/2014 02:31 PM, Tony Duell wrote:
Up to a point perhaps. But don't expect to get every byte of a 330GB
IDE drive. It might work with small IDEs (bad idea; they seem to be
falling like flies) or a small CF card.
In any case, you're going to need a PATA controller. Best to get one
with support built-in.
You might also get away with just a DDO, if your OS will support that
(i.e. I'm not aware of any DDOs that work with the 286 version of Xenix).
--Chuck
You're
going to need an ISA IDE adapter with its own BIOS to use with an
IDE drive or an IDE-to-CF card adapter (Promise made some of those).
Strictly you don't ened a BIOS extension. IDE drives were designed ot be
compatible wit hthe Western Digital controller in the original PC/AT. If
you can modify the dirve geometry tabel [1] then it will work with the
drivers in the BIOS ROMs. 
21 May
21 May
1:51 a.m.
On Tue, 20 May 2014, Chuck Guzis wrote:
In any case, you're going to need a PATA
controller. Best to get one with
support built-in.
No, you don't. You only need an interface. Most IDE interfaces were mainly
an address decoder and the like connected to the ISA bus. The controller
is on the drive, hence "IDE".
Christian
9 a.m.
On 05/21/2014 01:51 AM, Christian Corti wrote:
On Tue, 20 May 2014, Chuck Guzis wrote:
You're picking nits; back in the day, the word for the set of bus
transceivers and other logic to interface a PATA device to the ISA bus,
however trivial, was known as a "controller". Some were actually quite
sophisticated, offering local caching and bus-mastering.
Don't get cute. People know what I was referring to.
--Chuck
In any case, you're going to need a PATA
controller. Best to get one
with support built-in.
No, you don't. You only need an interface. Most IDE interfaces were
mainly an address decoder and the like connected to the ISA bus. The
controller is on the drive, hence "IDE".
12:57 p.m.
You're picking nits; back in the day, the word for
the set of bus
transceivers and other logic to interface a PATA device to the ISA bus,
however trivial, was known as a "controller". Some were actually quite
The PC C-crowd are not exactly known for using techncial terms correctly.
The obvious exampel is 'DB6'. But they also seem to use the terms 'VIOS'
and 'CMOS' interchangeably to mean the configuration parameter table.
These days 'A USB' seems ot be a flash memery device wit ha USB
itnerface. And so on...
I perefer to use terms correctly, it save a lot of confusion. And FWWI, I
object to the RX11, RX211, etc being called controllers for much the same
reason (bus interfaec is the term I use) -- in the RX01/RX02 ssytem the
'controlelr' is on of the PCBs in the drvie chassis, cotnaining a
microprogremd systerm
-tony
3:22 p.m.
On 05/21/2014 12:57 PM, Tony Duell wrote:
The PC C-crowd are not exactly known for using
techncial terms correctly.
The obvious exampel is 'DB6'. But they also seem to use the terms 'VIOS'
and 'CMOS' interchangeably to mean the configuration parameter table.
These days 'A USB' seems ot be a flash memery device wit ha USB
itnerface. And so on...
Such as a z9600 baud modem? Or DB-9 connector? Or my favorite: a CGA
controller. But then IBM uses "adapter" to mean just about anything--it
can be a simple mini-DIN to DIN bit of cable for attaching "PS/2" style
keyboards to older systems with DIN keyboard receptacles. Or it can be
a very complex graphics controller (e.g. PGA).
You use the lingua franca of the great unwashed to be understood. It's
just easier.
--Chuck
22 May
22 May
1:37 p.m.
Such as a z9600 baud modem? Or DB-9 connector? Or my favorite: a CGA
Yes, I meant to say 'DB9',but my fingers won't type that string of
characters easily...
controller. But then IBM uses "adapter" to
mean just about anything--it
can be a simple mini-DIN to DIN bit of cable for attaching "PS/2" style
keyboards to older systems with DIN keyboard receptacles. Or it can be
a very complex graphics controller (e.g. PGA).
I thhouhg the Professional Graphics board set was called a PGC
officially. Maybe to reduce confusion with the alternative expansion of
the acronym (Pin Grid Array).
FWIW, HP used the term 'controller' for just about every custom IC in the
HP85 series of machines.
You use the lingua franca of the great unwashed to be
understood. It's
just easier.
I thinki it depends who you are talking to. In a PC shop, even I might
ask for a 56Kbaud modem, knowing full well that is technicaly incorrect.
A a meeting of telecoms engineers, I would be a lot more precise.
The obvios esample is the 3.5" HD floppy disk. We all know that '1.44M
Bytes' implies a megabyte is 1000*1024 bytes. But equally, it's what we
all used to ask for in the lcoal shops. But if I was discussing a floppy
drive controller with anohter engineer, I would not use that name.
-tony
5:05 p.m.
On 05/22/2014 01:37 PM, Tony Duell wrote:
I thhouhg the Professional Graphics board set was
called a PGC
officially. Maybe to reduce confusion with the alternative expansion of
the acronym (Pin Grid Array).
Could be--I'm too lazy to check. But the A in CGA was definitely an
"adapter". So I was sloppy. Brevis esse laboro, obscurus fio.
Mea maxima culpa.
As an aside, would a PATA "adapter" with a BIOS ROM be an "interface",
"adapter" or "controller"? After all, isn't the software that
does the
controlling?
In disk formats, I'll sometimes say MFM 80x2x18x512, for example. But
only when I'm trying to be precise.
--Chuck
5:22 p.m.
On May 22, 2014, at 5:05 PM, Chuck Guzis <cclist at sydex.com> wrote:
On 05/22/2014 01:37 PM, Tony Duell wrote:
Internally (at IBM) it was always referred to as the PGA as I recall. In addition there
were CGA, EGA, and VGA (of course).
TTFN - Guy
I thhouhg the Professional Graphics board set was
called a PGC
officially. Maybe to reduce confusion with the alternative expansion of
the acronym (Pin Grid Array).
Could be--I'm too lazy to check. But the A in CGA was definitely an
"adapter". So I was sloppy. Brevis esse laboro, obscurus fio.
24 May
24 May
12:08 p.m.
On Fri, 23 May 2014, Tony Duell wrote:
My Optionas and Adapters Techref has the section called 'Monochrome
Display and Printer Adapter', the pages labelled 'Monochrome adapter' and
the schematics captioned 'Monochrome Display Adapter' (and FWIW the
schematics do contain the printer port too). You choose :-)
Waht wa all call the CGA is called the 'Color/Graphics Monitor Adapter'.
(Yes, Monitor, not Display)
-tony
What's happened to MDA ? :-)
It was replaced by "Hercules".
The IBM product was called "MDP". In 1981, that printer port was a
selling point.
22 May
22 May
1:45 a.m.
On Wed, 21 May 2014, Chuck Guzis wrote:
You're picking nits; back in the day, the word for
the set of bus
transceivers and other logic to interface a PATA device to the ISA bus,
however trivial, was known as a "controller". Some were actually quite
sophisticated, offering local caching and bus-mastering.
Sure I'm picking nits, sometime I just have to ;-))
But over here, we mostly said "interface", sometimes "adapter", only
seldom "controller". Think of "SCSI" where you have a host bus adapter
and
not a controller, even though most adapaters (e.g. with integrated SCSI
IC) were controllers in the sense of controlling themselves the protocol
between the host system and the drive. Any pre-DMA IDE interface doesn't
control anything in this regard.
Christian
21 May
21 May
12:41 p.m.
On 05/20/2014 02:31 PM, Tony Duell wrote:
Up to a point perhaps. But don't expect to get every byte of a 330GB
IDE drive. It might work with small IDEs (bad idea; they seem to be
falling like flies) or a small CF card.
My experience is that fr larger disks, the BIOS routines will correctly
access the first <n> MBytes (the vlaue of n was famous at one point, I've
long forgotten it). OSes that don't sue the BIOS vaiurly obviosuly don't
use BIOS extensions either, so they work fine with a plain bus adapter.
One Gotcha is that often the BIOS is used to boot jsut abotu any OS. So
what I did on this machine as to have a few small MSDOS partitions, then
the rest of the disk up to that boundary as linux partition contianing
tyhe root file system/kernel (which can be read by the BIOS with no
problems, so it can be booted from) and the test of the disk as a larger
linus partition mounted as /usr. The fact that hte latter can't be read
by the BIOS doesn't matter.
You're going to need an ISA IDE adapter with its own BIOS to use with an
IDE drive or an IDE-to-CF card adapter (Promise made some of those).
Strictly you don't ened a BIOS extension. IDE drives were designed ot be
compatible wit hthe Western Digital controller in the original PC/AT. If
you can modify the dirve geometry tabel [1] then it will work with the
drivers in the BIOS ROMs.
In any case, you're going to need a PATA controller. Best to get one
s/controller/bus interface/
with support built-in.
That depends on what you can find, and whether you like undocuemtned bits
of firmware in your system.
-tony
20 May
20 May
6:34 p.m.
On Tue, 20 May 2014, Tony Duell wrote:
[1] It is possible to ad a singel TTL IC to disable
the normal ROMs for
the last 8K words and map in 8K EPROMs (2764 type) inserted in the spare
sockets on the motherboard.
My 5150 had one spare ROM socket. I put in an optional enhancement ROM
from MBI.
My 5170s did NOT have any spare ROM
sockets. Or did you mean the sockets
used by the BASIC ROMs?
But, YES, many of my aftermarket XTs and ATs had extra sockets.
Or were there more variants of the 5170 motherboard than I was aware of?
Hard drive spinning up does not check both 5V and 12V, and power LED
doesn't mean much.
21 May
21 May
12:45 p.m.
My 5150 had one spare ROM socket. I put in an
optional enhancement ROM
from MBI.
My 5170s did NOT have any spare ROM sockets. Or did you mean the sockets
used by the BASIC ROMs?
But, YES, many of my aftermarket XTs and ATs had extra sockets.
Or were there more variants of the 5170 motherboard than I was aware of?
There were at least 2 versions of the 5170 motherboard. The obbious
difference is the RAM. The older one has 4 rows (36 packages) of those
strange IBM piggybacked chips, a pair of 64K*1 DRAMs with slgihtly
different pinouts. The later version ahs 2 rows (18 chips) of 41256s
The later board i have has 4 EPROM sockets. There was also a DIP shunt
block associated with them at the very back edge of the board. I think
you could eitehr fit 16K byte of 32K byte EPROMS (4 of the former, 2 of
the latter) with the correct jumpering of that DIP shunt.
-tony
3870
days inactive
3875
days old
17 comments
6 participants
participants (6)
-
ard＠p850ug1.demon.co.uk -
cc＠informatik.uni-stuttgart.de -
cclist＠sydex.com -
cisin＠xenosoft.com -
ggs＠shiresoft.com -
ploopster＠gmail.com