30 Jun
2005
30 Jun
'05
12:50 p.m.
Since the 380Z seems to be popular at the moment, does anyone have info
on 2 bits that are not properly described in the Information File?
The first is the dual cassette controller. The pinout of the 7 pin
cassette socket is given, as of course is the schematic of the VDU board
that contains the cassette circuitry. The controller seems to contain a
couple of relays to control the recorders (were these powered from the
system 5V line or a separate power supply?) and switching to allow either
one recorder to be used, or two with one used for recording the other for
playback. It's possible to make some guesses as to how it was wired, but
some real details would be nice
The second is the SIO-2 board. This was a bit-banged serial port using,
IIRC, 3 lines of the user port. 2 ouput lines for TxD and thr reader
control relay, one input for RxD. At least knowing what port bits were
used would be a help...
-tony
1 Jul
1 Jul
8:11 a.m.
On Thu, 2005-06-30 at 20:50 +0100, Tony Duell wrote:
Since the 380Z seems to be popular at the moment, does
anyone have info
on 2 bits that are not properly described in the Information File?
The first is the dual cassette controller. The pinout of the 7 pin
cassette socket is given, as of course is the schematic of the VDU board
that contains the cassette circuitry. The controller seems to contain a
couple of relays to control the recorders (were these powered from the
system 5V line or a separate power supply?) and switching to allow either
one recorder to be used, or two with one used for recording the other for
playback. It's possible to make some guesses as to how it was wired, but
some real details would be nice
Hmm... the couple of references I've seen to it in RML documentation
suggest that it was a 3rd party thing, rather than an RML part. They
talk as though it was some standardised thing built into dual cassette
drives, or possibly that there was some magic box that sat between the
380Z and a pair of standard tape drives.
Either way, the impression is that it was nothing to do with RML...
RML documentation is all over the darn place though - they're not
particularly consistent in a lot of cases, so snippets of info tend to
be splattered right across their documentation.
The second is the SIO-2 board. This was a bit-banged
serial port using,
IIRC, 3 lines of the user port. 2 ouput lines for TxD and thr reader
control relay, one input for RxD. At least knowing what port bits were
used would be a help...
According to the COS 3.4 Reference Manual: "Example programs for driving
a 20mA or RS232 serial device via the RML SIO-2 and SIO-2B interfaces
connected to the user I/O port can be found in the Option ROM listing"
So that might help - if you can find it! I've just checked my pile of
RML docs and I haven't got a copy of the option ROM listing :(
Mind you, according to Paul Williams' site I have a few docs that aren't
in this pile in front of me - so I'll go hunting in the loft in a sec as
I must have some more somewhere! ;)
cheers
Jules
3:32 p.m.
> The first is the dual cassette controller. The
pinout of the 7 pin
[...]
Hmm... the couple of references I've seen to it in
RML documentation
suggest that it was a 3rd party thing, rather than an RML part. They
I didn;'t get that impression at all....
talk as though it was some standardised thing built
into dual cassette
drives, or possibly that there was some magic box that sat between the
380Z and a pair of standard tape drives.
It's certainly a box that goes between the 380Z's cassette socket and a
pair of normal recorders. It also would appear (from re-reading the
minimal instructions I have) that it took power from the 5V output on the
cassette socket.
It clearly contains a couple of relays and drivers for them (the 'motor'
lines on the 7 pin DIN socket are outputs from a TTL latch chip). I
assume it also cnntained switching for the audio signals depending on
whether one or two recorders were used. But I don't know the details.
Even a list of the front panel controls and connectors would be a start...
Either way, the impression is that it was nothing to do with RML...
I've never seen any other machine with this sort of cassette connector
(in particular audio I/O and TTL level motor control lines).
[Bit-banged SIO-2]
According to the COS 3.4 Reference Manual:
"Example programs for driving
a 20mA or RS232 serial device via the RML SIO-2 and SIO-2B interfaces
connected to the user I/O port can be found in the Option ROM listing"
So that might help - if you can find it! I've just checked my pile of
RML docs and I haven't got a copy of the option ROM listing :(
Incidentally, do you have the COS (monitor ROM) source listings? They
were available from RML, but I didn't get them with my machine, alas.
There is an example program for reading a tape from an ASR33 in the
Information File. It would appear that the reader relay is controlled by
output bit 1, and that RxD comes in on input bit 0. It would be logical
if output bit 0 was TzD, therefore, but I wouldn't put money on it :-)
-tony
2 Jul
2 Jul
4:56 a.m.
Incidentally, do you have the COS (monitor ROM) source
listings? They
were available from RML, but I didn't get them with my machine, alas.
I'm sure I have some listings, and my vague recollection is that they
were for COS. I scanned everything I have and sent it to Manx
so you should be able to find it there. If none of the titles
make it obvious what you want, let me know and I'll find
the paper version and confirm which document you need.
Antonio
2:47 p.m.
Incidentally, do you have the COS (monitor ROM)
source listings? They
were available from RML, but I didn't get them with my machine, alas.
I'm sure I have some listings, and my vague recollection is that they
were for COS. I scanned everything I have and sent it to Manx
1:17 p.m.
On Fri, 2005-07-01 at 23:32 +0100, Tony Duell wrote:
I don't have any kind of docs for a tape-only system though; only ones
for disk systems - which happen to describe tape procedures too. Maybe
you have some different ones. It was more that I'm surprised they didn't
say in the docs that it was an RML part, or avaialble from RML or
whatever.
The Disk Information File does mention a "Firmware reference manual" as
having more info on the COS tape routines and a description of the
interface - maybe that one has more details in it about the dual tape
controller. It's not a document I've seen.
> The first
is the dual cassette controller. The pinout of the 7 pin
[...]
Hmm... the couple of references I've seen to
it in RML documentation
suggest that it was a 3rd party thing, rather than an RML part. They
I didn;'t get that impression at all.... Incidentally, do you have the COS (monitor ROM) source
listings? They
were available from RML, but I didn't get them with my machine, alas.
If you can stand a PDF I think this is the one :-)
http://vt100.net/rm/docs/RML_380Z_cassette_operating_system.pdf
489KB, so not too huge either :)
I've brought a bunch of RML floppies home for the week from the museum,
so I'll see what's on those and at least make a few backup copies. Might
well be something interesting (I don't believe we've got any RML stuff
on tape though)
cheers
Jules
3:10 p.m.
I don't have any kind of docs for a tape-only
system though; only ones
Nor do I. I've never actually seen a system without disk drives either,
although I am pretty sure they existed.
for disk systems - which happen to describe tape
procedures too. Maybe
you have some different ones. It was more that I'm surprised they didn't
say in the docs that it was an RML part, or avaialble from RML or
whatever.
The Disk Information File does mention a "Firmware reference manual" as
having more info on the COS tape routines and a description of the
interface - maybe that one has more details in it about the dual tape
controller. It's not a document I've seen.
I think I do have a COS firmware manual somewhere, but it will take some
finding...
Incidentally, do you have the COS (monitor ROM)
source listings? They
were available from RML, but I didn't get them with my machine, alas.
If you can stand a PDF I think this is the one :-)
http://vt100.net/rm/docs/RML_380Z_cassette_operating_system.pdf
489KB, so not too huge either :)
I've brought a bunch of RML floppies home for the week from the museum,
so I'll see what's on those and at least make a few backup copies. Might
well be something interesting (I don't believe we've got any RML stuff
on tape though)
Pity. I'd hoped at least one original RML cassette esisted somewhere.
From reading the manual, it appers that the phase of
the replay signal is
critical (it was with some other tape interfaces too) and I
was hoping I
could get an original tape to set up against. Oh well....
-tony
3 Jul
3 Jul
6:23 a.m.
On Sat, 2005-07-02 at 23:10 +0100, Tony Duell wrote:
We've got several 380Z's (well, two or three) that have no drives in
them - I expect those are tape-only systems. Unfortunately I've never
had time to look inside them as there's always been something more
important to do! I know we've got one of the 8" drive setups too which
are rather rare, plus we've got one 380Z-FD machine which should have
the "full disk" controller in it (on-board RAM, ROM, its own Z80 cpu
etc.).
(Said disk controller also has the SIO-4 on board, and the on-board
firmware should allow communication direct between the disks and the
serial port - I believe it's the same board as used for the 480Z shared
disk system)
What we don't have is keyboards. In some cases that's not such a big
deal as with the later serial boards they could run from a terminal
rather than system keyboard & display.
Good - RML documentation seems to be *very* thin on the ground! (Not
like the machines are common either!)
Presumably saving some data on one (or more) tape-based system(s) would
be enough to you to read it one yours and setup against that. Not as
ideal as going from an original manufacturer tape maybe, but probably
the best that can be hoped for.
Aside - ever seen a 480Z ROM cartridge? I only found out that they
existed (according to the documentation) the other day. Never heard of
anyone owning or seeing one before though...
cheers
Jules
I don't
have any kind of docs for a tape-only system though; only ones
Nor do I. I've never actually seen a system without disk drives either,
although I am pretty sure they existed. The Disk
Information File does mention a "Firmware reference manual" as
having more info on the COS tape routines and a description of the
interface - maybe that one has more details in it about the dual tape
controller. It's not a document I've seen.
I think I do have a COS firmware manual somewhere, but it will take some
finding... I've
brought a bunch of RML floppies home for the week from the museum,
so I'll see what's on those and at least make a few backup copies. Might
well be something interesting (I don't believe we've got any RML stuff
on tape though)
Pity. I'd hoped at least one original RML cassette esisted somewhere.
From reading the manual, it appers that the phase
of the replay signal is
critical (it was with some other tape interfaces too) and
I was hoping I
could get an original tape to set up against. Oh well....
11:17 a.m.
On Sat, 2005-07-02 at 23:10 +0100, Tony Duell wrote:
We've got several 380Z's (well, two or three) that have no drives in
them - I expect those are tape-only systems. Unfortunately I've never
Thay may have been used with external floppy units. The only way to be
sure is to see what PCBs are inside.
I
don't have any kind of docs for a tape-only system though; only ones
Nor do I. I've never actually seen a system without disk drives either,
although I am pretty sure they existed. had time to look inside them as there's always
been something more
important to do! I know we've got one of the 8" drive setups too which
are rather rare, plus we've got one 380Z-FD machine which should have
the "full disk" controller in it (on-board RAM, ROM, its own Z80 cpu
etc.).
I was readign the 380Z service manual on vt100.net this morning. It was
the first time I'd come across this board, it looks rather fun.
I thought the 'full disk' controller was the 8" one (which was
essentially the same as the normal 380Z controller), and that the one
with the on-board Z80 was the Intellegent Disk Controller. But I could
well be wrong.
(Said disk controller also has the SIO-4 on board, and the on-board
FWIW, the normal, non-intellegent disk controller has a SIO-4 on the same
PCB, but it's nothing to do with the disk controller function. Obviously
on that board you can go from serial port to disk without using the rest
of the machine.
I believe the SIO-4 board (also SIO-5, SIO-6) was just a disk controller
board with the serial chips fitted and the disk controller chips not
fitted. Much as the RAM expansion board is the same PCB as the CPU board,
but you leave off the CPU and buffers, etc.
firmware should allow communication direct between the
disks and the
serial port - I believe it's the same board as used for the 480Z shared
disk system)
That's what the service manual implied. I have a 480Z (one of the older
metal-cased ones), but no disk unit for it.
What we don't have is keyboards. In some cases that's not such a big
deal as with the later serial boards they could run from a terminal
rather than system keyboard & display.
The keyboard interface spec is in the Information File and service
manual. The latter also contains a scheamtic of _a_ keyboard, but from
what I rmemeber the one I have is very different. Not that said schematic
is a lot of use for making one, since the contents of the microcontroller
firmware ROM are not know. The keyboard I have (and from what I remember
we had at school) has a single 40 pin custom chip in it, nothing more.
It would eb a fairly simple microcontroller project to make an interface
from a PC keyboard (the old XT, AT, PS/2 ones) to the
RML keyboard connector.
Pity. I'd
hoped at least one original RML cassette esisted somewhere.
Presumably saving some data on one (or more) tape-based system(s) would
be enough to you to read it one yours and setup against that. Not as From reading the manual, it appers that the phase
of the replay signal is
critical (it was with some other tape interfaces too) and
I was hoping I
could get an original tape to set up against. Oh well....
3:02 p.m.
On Sun, 2005-07-03 at 19:17 +0100, Tony Duell wrote:
It's on my to-do list... it's just been there for about a year now :-)
Hmm... I've only seen schematics for two different disk controllers -
the simple one (still with an SIO-4 on board) with an FD1771 IC on
board, and the one that I thought was the 'full disk' controller with
the on-board CPU etc.
Far as I know, the former board is FM only and can put about 70KB on a
disk, whilst the latter one can do MFM and gets about 240KB on a disk.
I'll have to have a look inside our FD machine and see what disk board
it has...
Aha, I've got a 480Z disk unit here - never tried it as I need to sit
down and figure out the cable wiring which puts a bit of extra time on
the task.
It does seem that there were a few different keyboards, but I'm not sure
how they differed internally. Far as I know, they all talked the same
protocol to the CPU though.
I think hacking a PC keyboard to work with a 380Z would spoil the
machine though :-) I'd rather use a period terminal and drive things
that way. Hitting a key on the terminal at system start should be enough
to wake things up and tell the 380Z to use the terminal as input rather
than the usual keyboard. Can't remember if there's a way of telling it
what device to use as output, though...
We've got
several 380Z's (well, two or three) that have no drives in
them - I expect those are tape-only systems. Unfortunately I've never
Thay may have been used with external floppy units. The only way to be
sure is to see what PCBs are inside. important to
do! I know we've got one of the 8" drive setups too which
are rather rare, plus we've got one 380Z-FD machine which should have
the "full disk" controller in it (on-board RAM, ROM, its own Z80 cpu
etc.).
I was readign the 380Z service manual on vt100.net this morning. It was
the first time I'd come across this board, it looks rather fun.
I thought the 'full disk' controller was the 8" one (which was
essentially the same as the normal 380Z controller), and that the one
with the on-board Z80 was the Intellegent Disk Controller. But I could
well be wrong. FWIW, the normal, non-intellegent disk controller has
a SIO-4 on the same
PCB, but it's nothing to do with the disk controller function. Obviously
on that board you can go from serial port to disk without using the rest
of the machine.
Yep. :-)
I believe the SIO-4 board (also SIO-5, SIO-6) was just
a disk controller
board with the serial chips fitted and the disk controller chips not
fitted.
Ahhh, that's an SIO-4C according to the disk system info file - the
board for the SIO-4 always has the disk controller circuitry with it
according to that doc. RML didn't half make things confusing.
Much as the RAM expansion board is the same PCB as the
CPU board,
but you leave off the CPU and buffers, etc.
Actually, info on the various PAL chips would be handy; RML didn't half
like stuffing PAL chips with different colour codings into all their
stuff, and I'm not aware that the equations are documented anywhere.
(Your mention of the RAM board made me think of that, as the PAL is
different for a RAM board versus the equivalent on the CPU board)
firmware
should allow communication direct between the disks and the
serial port - I believe it's the same board as used for the 480Z shared
disk system)
That's what the service manual implied. I have a 480Z (one of the older
metal-cased ones), but no disk unit for it. What we
don't have is keyboards. In some cases that's not such a big
deal as with the later serial boards they could run from a terminal
rather than system keyboard & display.
The keyboard interface spec is in the Information File and service
manual. The latter also contains a scheamtic of _a_ keyboard, but from
what I rmemeber the one I have is very different. In the xase of the 380Z, what I want to do is to try
to load a real RML
tape, sort out the replay phase using that, then sort out the recording
phase so I can load my own recordings.
I'll double check regarding tapes. I've got about half of Bletchley's
RML software here, and one of the other guys has the rest - so it may be
that he has one or more tapes...
cheers
Jules
8:13 p.m.
I thought the
'full disk' controller was the 8" one (which was
essentially the same as the normal 380Z controller), and that the one
with the on-board Z80 was the Intellegent Disk Controller. But I could
well be wrong.
Hmm... I've only seen schematics for two different disk controllers - Far as I know, the former board is FM only and can put
about 70KB on a
The 1771 is certainly FM (single density) only.
disk, whilst the latter one can do MFM and gets about
240KB on a disk.
I've not seriously looked at those schematics yet...
Much as the
RAM expansion board is the same PCB as the CPU board,
but you leave off the CPU and buffers, etc.
Actually, info on the various PAL chips would be handy; RML didn't half That's
what the service manual implied. I have a 480Z (one of the older
metal-cased ones), but no disk unit for it.
Aha, I've got a 480Z disk unit here - never tried it as I need to sit
down and figure out the cable wiring which puts a bit of extra time on
the task. It does seem that there were a few different
keyboards, but I'm not sure
how they differed internally. Far as I know, they all talked the same
protocol to the CPU though.
They did. 7 bit ASCII code and a strobe line. That much is documented in
the Infroation File.
I think hacking a PC keyboard to work with a 380Z would spoil the
machine though :-) I'd rather use a period terminal and drive things
COnsidering my Infromation File is dated November 1981, a PC keyboard
_is_ period :-)
that way. Hitting a key on the terminal at system
start should be enough
to wake things up and tell the 380Z to use the terminal as input rather
than the usual keyboard. Can't remember if there's a way of telling it
Are you sure about that? I've never seen it documented anywhere.
what device to use as output, though...
I'll double check regarding tapes. I've got
about half of Bletchley's
Thanks. But absoutely no hurry -- this is very much a 'to do sometime'
project, and it's been on the list for many years!
-tony
4 Jul
4 Jul
5:37 a.m.
On Mon, 2005-07-04 at 04:13 +0100, Tony Duell wrote:
That's what I'm thinking of :) I haven't seen their contents noted down
anywhere. My programmer here might well handle them, I just don't want
to risk toasting one out of a running machine!
Indeed. I suppose given a dump of the ROM out of my 480Z disk unit, plus
the schematics for the board, and the 480Z details it would be possible
to figure it out - but somewhat time-consuming.
I just pulled the lid on my disk unit and it definitely has all the
right bits to be the intelligent disk controller; it's just missing all
the circuitry to interface to the 380Z bus (it just has all the serial
side of things present + the Z80 part of everything)
It's in the COS 3.4 Information File - under misc functions (reset and
initialisation):
"COS checks to see if there is a VDU plugged into the SIO-4 socket. If
there is, all subsequent output via the scroller output EMTs is sent to
this VDU. Graphical output, such as the front panel, or low or high
resolution graphics, still appear on the RML screen. If a character is
received from the VDU at 9600 baud, COS also switches the keyboard EMTs
to take all subsequent input from the VDU ignoring the RML keyboard."
So it should work with an SIO-4 (or presumably 4C). Indeed, we've
actually got some Cifer terminals doing nothing which have nice black
keyboards very similar in looks to the 380Z ones...
The text implies that it's not possible to have a terminal as input-only
though; it'll always redirect text output if a terminal's present. How
much RML software actually uses the COS routines is another matter -
there must be at least some bits and pieces out there that talk to the
keyboard / display directly rather than through COS, and those then
wouldn't work...
Heh, I know what you mean! In the case of RML stuff I like to try and
keep track of who has what anyway as it's reasonably uncommon stuff.
cheers
Jules
There are only 2 boards. The 8" controller is the
same as the
non-intellegent 5.25" one. There are notes on the schematic for the link
settings for the FDS (== 8") and MDS (== 5.25") versions.
Not me yet... more interested in their SASI controller at this stage to
be honest as I really need a way of getting the 380Z fileserver backed
up!
Far as I know, the former board is FM only and
can put about 70KB on a
The 1771 is certainly FM (single density) only.
disk, whilst the latter one can do MFM and gets
about 240KB on a disk.
I've not seriously looked at those schematics yet... Actually, info
on the various PAL chips would be handy; RML didn't half
THere were no PALs in the original set of boards (CPU, VDU, hi-res, disk,
etc). WHat there were -- and were to excess -- were small bipolar PROMs.
Those were often colour-coded. Aha, I've
got a 480Z disk unit here - never tried it as I need to sit
down and figure out the cable wiring which puts a bit of extra time on
the task.
It would be useful to know the protocol that the 480Z uses to talk to the
disk unit. From what I remember, though,, it uses the serial chips in
synchronous mode, which means finding another machine that could act as a
disk server would not be totally trivial. that way.
Hitting a key on the terminal at system start should be enough
to wake things up and tell the 380Z to use the terminal as input rather
than the usual keyboard. Can't remember if there's a way of telling it
Are you sure about that? I've never seen it documented anywhere. I'll
double check regarding tapes. I've got about half of Bletchley's
Thanks. But absoutely no hurry -- this is very much a 'to do sometime'
project, and it's been on the list for many years!
5:09 p.m.
Not me yet... more interested in their SASI controller
at this stage to
be honest as I really need a way of getting the 380Z fileserver backed
up!
What about hacking around with the Acorn SASI controller, e.g. the one in
your ACW? I assume you have schematics, etc for that.
Actually,
info on the various PAL chips would be handy; RML didn't half
THere were no PALs in the original set of boards (CPU, VDU, hi-res, disk,
etc). WHat there were -- and were to excess -- were small bipolar PROMs.
Those were often colour-coded. "COS checks to see if there is a VDU plugged into
the SIO-4 socket. If
What does it look for? One of the handshake lines being asseted?
there is, all subsequent output via the scroller
output EMTs is sent to
this VDU. Graphical output, such as the front panel, or low or high
resolution graphics, still appear on the RML screen. If a character is
Argh!. While twin monitor setups are useful in some cases, this sounds
like something of a kludge.
received from the VDU at 9600 baud, COS also switches
the keyboard EMTs
to take all subsequent input from the VDU ignoring the RML keyboard."
So it should work with an SIO-4 (or presumably 4C). Indeed, we've
actually got some Cifer terminals doing nothing which have nice black
keyboards very similar in looks to the 380Z ones...
The text implies that it's not possible to have a terminal as input-only
though; it'll always redirect text output if a terminal's present. How
much RML software actually uses the COS routines is another matter -
The EMTs (one of the Z80 RSTs followed by a function code byte) are
certainly used in the CP/M CBIOS. I suspect a lot of software uses them.
there must be at least some bits and pieces out there
that talk to the
keyboard / display directly rather than through COS, and those then
wouldn't work...
Anything doing graphics, for a start...
Personally, I'd try to use the RML's video card for everything, which
means kludging up a keyboard. It can't be that hard, the interface is as
simple as it gets.
Heh, I know what you mean! In the case of RML stuff I
like to try and
keep track of who has what anyway as it's reasonably uncommon stuff.
By my standards they're fairly common....
-tony
6:19 p.m.
On Tue, 2005-07-05 at 01:09 +0100, Tony Duell wrote:
Yep, could do - plus I've got the schematics for the Torch one
somewhere. Just wanted to check that there wasn't anything funny about
the RML one (there shouldn't be) - SASI should be SASI no matter what.
Doesn't say, unfortunately. I actually think that RML started out with
the best intentions when producing their documents and wanted to
document everything - with the result being that time constraints meant
things were somewhat patchy! Related stuff seems to be spread over lots
of docs seemingly randomly in places, plus there are some big holes...
Yep I agree - still, it's nice that the functionality's sort of there.
I'll have to try it sometime - just got my own 380Z up and running
again.
Oh sure, not super-rare - but it's much easier to pick up an S100 system
(say) than a 380Z or other systems of similar age / performance /
flexibility. I seem to have been pretty lucky with RML finds over the
years.
seeya
Jules
Not me yet...
more interested in their SASI controller at this stage to
be honest as I really need a way of getting the 380Z fileserver backed
up!
What about hacking around with the Acorn SASI controller, e.g. the one in
your ACW? I assume you have schematics, etc for that. [PROMS]
I'll send you a uuencoded tar.gz file of the ones in my machine (CPU, RAM
expansion, 40 column video, hi-res). May be a start.
Brilliant - ta muchly for that. I'll see if I can work my programmer out
and read the others that I have...
"COS
checks to see if there is a VDU plugged into the SIO-4 socket. If
What does it look for? One of the handshake lines being asseted? there is, all
subsequent output via the scroller output EMTs is sent to
this VDU. Graphical output, such as the front panel, or low or high
resolution graphics, still appear on the RML screen. If a character is
Argh!. While twin monitor setups are useful in some cases, this sounds
like something of a kludge. The EMTs (one of the Z80 RSTs followed by a function
code byte) are
certainly used in the CP/M CBIOS. I suspect a lot of software uses them.
Well that's a good thing in this case :)
Personally, I'd try to use the RML's video
card for everything, which
means kludging up a keyboard. It can't be that hard, the interface is as
simple as it gets.
Yep, true. I'll have to have a look inside one of the Cifer keyboards
sometime as they look the part - maybe one of those can be hooked up via
an interface...
Heh, I know
what you mean! In the case of RML stuff I like to try and
keep track of who has what anyway as it's reasonably uncommon stuff.
By my standards they're fairly common....
5 Jul
5 Jul
12:13 p.m.
On Tue, 2005-07-05 at 01:09 +0100, Tony Duell wrote:
Yep, could do - plus I've got the schematics for the Torch one
somewhere. Just wanted to check that there wasn't anything funny about
IIRC Torch made a SCSI controller for the Beeb, it was used with a ROM
called SCISFS. If you've found one of those you're very lucky, I am told
that fewer than 100 were made.
Not me
yet... more interested in their SASI controller at this stage to
be honest as I really need a way of getting the 380Z fileserver backed
up!
What about hacking around with the Acorn SASI controller, e.g. the one in
your ACW? I assume you have schematics, etc for that. the RML one (there shouldn't be) - SASI should be
SASI no matter what.
IIRC the RML hard disk unit used a standard ST506-interfaced drive and a
standard bridgeboard to SASI (Xebec?) In which case the host adapter is
going to be pretty normal too.
[PROMS]
I'll send you a uuencoded tar.gz file of the ones in my machine (CPU, RAM
expansion, 40 column video, hi-res). May be a start.
Brilliant - ta muchly for that. I'll see if I can work my programmer out
and read the others that I have... From what I remember there are no PROMs on the standard
(non-intellegent)
disk controller or on the PIO card that I have somewhere. The
latter is a
board with an address decoder, 3 Z80-PIOs and a Z80-CTC on it, along with
a wire-wrap area. I don't have schematics, but it would be trivial to
work out.
"COS
checks to see if there is a VDU plugged into the SIO-4 socket. If
What does it look for? One of the handshake lines being asseted? Personally,
I'd try to use the RML's video card for everything, which
means kludging up a keyboard. It can't be that hard, the interface is as
simple as it gets.
Yep, true. I'll have to have a look inside one of the Cifer keyboards
sometime as they look the part - maybe one of those can be hooked up via
an interface...
1:50 p.m.
It should be piossible to work out what COS looks for
from the source
listings. I am not sure, though, if the version on vt100.net is recent
enough to support this. It's a pity nobody has the COS 3.4 or COS 4.0
listings.
I cannot find the COS listings in there at all. I must have missed
a manual somewhere. I'll traipse up into the attic in the next day
or two and see if I can find the right box.
Antonio
4:23 p.m.
It should be piossible to work out what COS looks
for from the source
listings. I am not sure, though, if the version on vt100.net is recent
enough to support this. It's a pity nobody has the COS 3.4 or COS 4.0
listings.
I cannot find the COS listings in there at all. I must have missed
a manual somewhere. I'll traipse up into the attic in the next day
or two and see if I can find the right box.
6 Jul
6 Jul
11:46 a.m.
IIRC you go to http://www,vt100.net/rm/ and follow the
obvious link for
documentation. Near (at?) the top of the list is a
source listing for
an
early version of the COS monitor. It's about half a
megabyte, pdf.
My mistake is assuming I was the source for the manuals there -
obviously some others snuck in that's one of them!
But I'm sure I've seen a listing in paper form, so I still need
to go hunting in the attic.
Antonio
1:40 p.m.
IIRC you go to http://www,vt100.net/rm/ and follow the
obvious link for
documentation. Near (at?) the top of the list is a
source listing for
an
early version of the COS monitor. It's about half a
megabyte, pdf.
A quick trip to the attic restored my faith in my sanity (not a good
measure of anything, but it's all I have to go on :-)).
I found "Resaerch Machines COS 4.0 Assembly Listing" for Release 4.0B
from October 1981.
The reason I've not scanned it is that it is a perfect bound book (I
think that's the term, pages glued into the spine). So I can and will
scan it, but it will take a little time since I cannot spend hours
at the office scanner :-)
There are other similarly unscanned RML manuals up there, so I'll
make a proper list at the weekend.
Antonio
5 Jul
5 Jul
1:56 p.m.
On Tue, 2005-07-05 at 20:13 +0100, Tony Duell wrote:
IIRC Torch made a SCSI controller for the Beeb, it was used with a ROM
called SCISFS. If you've found one of those you're very lucky, I am told
that fewer than 100 were made.
Hmm... my Torch 725 has a Torch SCSI board in it - although I recall
that it's not a complex board at all, just a handful of TTL chips.
Whether that's the same board as available from them for the BBC I don't
know, but as the 725 uses a BBC as the I/O processor it's reasonable to
suggest that it is. I'll have to check what ROMs it has... (it's got
one of the Torch 68000 coprocessors in it and should run Unix - *but* I
have a feeling that the hard drive's toast and I don't think it's
something I have media for)
The big brown-cased C-series machines that had hard disks in used a
Torch SASI board; I expect those are somewhat more common...
Can't remember what make the bridge board is now, but yep, it's one of
the common ones. The controller shouldn't be doing anything funny with
signal lines or the data, but I thought I'd check!
Nice - I've not seen one of the PIO boards before. And agreed, I don't
think there's a PROM on the standard FDC either.
What about hacking around with the Acorn SASI
controller, e.g. the one in
your ACW? I assume you have schematics, etc for that.
Yep, could do - plus I've got the schematics for the Torch one
somewhere. Just wanted to check that there wasn't anything funny about the RML one
(there shouldn't be) - SASI should be SASI no matter what.
IIRC the RML hard disk unit used a standard ST506-interfaced drive and a
standard bridgeboard to SASI (Xebec?) In which case the host adapter is
going to be pretty normal too. From what I
remember there are no PROMs on the standard (non-intellegent)
disk controller or
on the PIO card that I have somewhere. The latter is a
board with an address decoder, 3 Z80-PIOs and a Z80-CTC on it, along with
a wire-wrap area. I don't have schematics, but it would be trivial to
work out. It should be piossible to work out what COS looks for
from the source
listings. I am not sure, though, if the version on vt100.net is recent
enough to support this. It's a pity nobody has the COS 3.4 or COS 4.0
listings.
Hmm yep, that is a shame. I think a lot changed between COS 3.0 and
3.4...
I still wonder if the PC keyboard interface would be a
good idea. It
would have an adantage to you that the public at Bletchley could bash
away on a PC keyboard without risking anything too rare.
Yep, that is a good point. Mind you, we've got a couple of VT220s with
dead LOPTs - they'd make nice input-only devices!
cheers
Jules
4:31 p.m.
On Tue, 2005-07-05 at 20:13 +0100, Tony Duell wrote:
IIRC Torch made a SCSI controller for the Beeb, it was used with a ROM
called SCISFS. If you've found one of those you're very lucky, I am told
that fewer than 100 were made.
Hmm... my Torch 725 has a Torch SCSI board in it - although I recall
that it's not a complex board at all, just a handful of TTL chips.
That sounds like the Beeb one. Maybe it's Torch Beeb SCSI systems that
are rare -- there's a plinth to go under the monitor that contains a
floppy drive, hard drive, SCSI interface, etc.
What about hacking around with the Acorn SASI
controller, e.g. the one in
your ACW? I assume you have schematics, etc for that.
Yep, could do - plus I've got the schematics for the Torch one
somewhere. Just wanted to check that there wasn't anything funny about Nice - I've not seen one of the PIO boards before.
And agreed, I don't
think there's a PROM on the standard FDC either.
Mine is not currently in my 380Z, in fact it's never been in my 380Z. It
turned up in a box of Apple ][ cards I was given (and FWIW, some of those
Apple boards are not common -- things like speech input boards, etc). I
must find time to trace out the schematic, wire-wrap the PIOs to
connectors in the prottyping area, and stick it in the 380Z.
The PIOs and CTC are in wire-wrap DIL sockets. The bus side is wired up
by traces on the PCB, the user is intended to wire-wrap to the
appropriate pins on the sockets. There's space on the PCB for a few more
ICs if you need to fiddle around with signals, and space for at least one
IDC connector to bring your I/O signals off-board to connectors on the
back of the 380Z.
It should be piossible to work out what COS looks
for from the source
listings. I am not sure, though, if the version on vt100.net is recent
enough to support this. It's a pity nobody has the COS 3.4 or COS 4.0
listings.
Hmm yep, that is a shame. I think a lot changed between COS 3.0 and
3.4... I still wonder if the PC keyboard interface would
be a good idea. It
would have an adantage to you that the public at Bletchley could bash
away on a PC keyboard without risking anything too rare.
Yep, that is a good point. Mind you, we've got a couple of VT220s with
dead LOPTs - they'd make nice input-only devices!
7 Jul
7 Jul
1:42 p.m.
I've just found that there are at least two case variants for the 380Z
too; they have exactly the same connectors and panel blanking plates /
cut-outs, but in a totally different arrangement, which seems a strange
thing to do. (One styles does have the stencils for RGBS video signals
on the back, but both types have the blanking plate / hole for the BNC
video signals, so it doesn't seem to be the case that one type was never
intended to have a colour board)
I'm still hoping to find out one day why at least one of RML's adverts
shows a white-cased 380Z; I've never heard of anyone who owns such a
thing.
I think at least one of my manuals mentions Tony's prototyping board in
passing, but I've not seen any details on it though.
cheers
Jules
2:46 p.m.
On 7/7/05, Jules Richardson <julesrichardsonuk at yahoo.co.uk> wrote:
I've just found that there are at least two case variants for the 380Z
too; they have exactly the same connectors and panel blanking plates /
cut-outs, but in a totally different arrangement, which seems a strange
thing to do. (One styles does have the stencils for RGBS video signals
on the back, but both types have the blanking plate / hole for the BNC
video signals, so it doesn't seem to be the case that one type was never
intended to have a colour board)
I'm still hoping to find out one day why at least one of RML's adverts
shows a white-cased 380Z; I've never heard of anyone who owns such a
thing.
I think at least one of my manuals mentions Tony's prototyping board in
passing, but I've not seen any details on it though.
cheers
Jules
We had some external harddrives for the 380Z server we had at the
school I went to. They where in the same case as a 380Z but they where
white (same cream colour as the newer 480Z) and the reset button was
just a power light. Inside was 2 ST225N's with 2 xebec boards. I'm not
sure if this was an original spec though. The drives could of been
upgraded, they where the same as they used in the Nimbus.
Dan
3:15 p.m.
On Thu, 2005-07-07 at 22:46 +0100, Dan Williams wrote:
We had some external harddrives for the 380Z server we
had at the
school I went to. They where in the same case as a 380Z but they where
white (same cream colour as the newer 480Z) and the reset button was
just a power light. Inside was 2 ST225N's with 2 xebec boards. I'm not
sure if this was an original spec though. The drives could of been
upgraded, they where the same as they used in the Nimbus.
Hmm, now that's interesting. My 380Z fileserver has the same black case
to it for the hard disk as the normal 380Z. As with the ones you
remember, there's no reset button, just a power light, and there's a
Xebec board inside.
Can't recall what drive mine has inside, but it's a full-height unit so
definitely not a '225.
I suppose it makes more sense for them to be white/cream rather than
black, given that they were supposed to be used with the cream 480Z
systems - in which case maybe my black one's something of the odd one
out!
One day I'll bump into an ex-RML employee who has all the definite
answers to everything (I hope :)
I don't *think* the fileservers were meant to be used with a network of
380Z machines - but that's only going by the fact that I've never seen a
380Z with a network interface in it, except for my fileserver
controller. Maybe they did exist as a standard network shipped by RML
and that's why my fileserver unit is black...
(mine was certainly used with 480Z's though as I got both fileserver and
480Z machines in the same pile from a school)
At least that's confirmation that the cream/white machines did actually
exist though!
cheers
Jules
5:27 p.m.
I've just found that there are at least two case variants for the 380Z
too; they have exactly the same connectors and panel blanking plates /
cut-outs, but in a totally different arrangement, which seems a strange
thing to do. (One styles does have the stencils for RGBS video signals
on the back, but both types have the blanking plate / hole for the BNC
video signals, so it doesn't seem to be the case that one type was never
intended to have a colour board)
I don't think I've seen one with the RGB sockets labelled on the back
panel. My RGB board, and the one I fitted to a machine back at school,
had coloured dots stuck on the bracket hodling the BNCs to label them.
The intallation notes for the RGB board in my machine mention at least 3
case variants :
'For most black case 380Zs : facing the rear panel, remove the
rectangular blanking plate on the lower right of the rear panel (behind
the drives on an MDS-2). There may be one or two blanking plates, if two
use the one nearest the righthand side. On a few black case machines a
horizontally alighend rectangular hole was provided; this will need
slightly enlarging by the user before an RGB board can be fitted
For blue-case (early model) 380Zs : consult Research Machines'
There were also at least 3 different front panels : no drives, 2
horizontally-mounted full-height drives, and 2 vertically mounted 2/3rd
height drivwa.
I'm still hoping to find out one day why at least one of RML's adverts
shows a white-cased 380Z; I've never heard of anyone who owns such a
thing.
I've seen a hard disk unit in much the same case as at 380Z but white-ish
in colour.
I think at least one of my manuals mentions Tony's prototyping board in
passing, but I've not seen any details on it though.
It's mentioned in the Information File IIRC...
-tony
8 Jul
8 Jul
4:38 a.m.
On Fri, 2005-07-08 at 01:27 +0100, Tony Duell wrote:
For blue-case (early model) 380Zs : consult Research
Machines'
There were also at least 3 different front panels : no drives, 2
horizontally-mounted full-height drives, and 2 vertically mounted 2/3rd
height drivwa.
Yep - plus front panels that say 380Z-FD rather than just 380Z, and of
course one that says Winchester Disk rather than 380Z. Plus it sounds
like there are at least five different rear case panels - two variants
with the same holes, one with two RGB-type holes, one with a horizontal
RGB-type hole, and the one for the Winchester unit.
Then there are white-cased and blue-cased machines on top of that.
Actually, I'd not heard of the blue-cased machines before. Maybe they're
in standard off-the-shelf instrument cases rather than with RML logos,
panel cut-outs etc.?
I don't know of anyone with anything other than a black-cased machine
(albeit there are a few variants out there) - shows how much RML
hardware and knowledge has been lost I suppose :-(
Oh, Xebec board in my fileserver is an S1410, for which I have a manual
and a couple of spare boards, so I can do some messing around with a PC
parallel port driver. There's even a sample Z80 interface schematic in
the back of the manual that's probably adaptable...
Still concerned about timings though as some of the timing diagrams do
have upper constraints on signal changes (a lot of them only specify a
minimum limit). Time will tell if the PC's port is quick enough...
cheers
Jules
3:51 p.m.
Actually, I'd not heard of the blue-cased machines
before. Maybe they're
in standard off-the-shelf instrument cases rather than with RML logos,
panel cut-outs etc.?
The black case is a standard (Vero iIRC) case. It was designed to take a
19" rack chassis, but RML just mounted the PSU and card guides inside.
OK, the cutouts and logos are custom, but that's all.
Oh, Xebec board in my fileserver is an S1410, for
which I have a manual
and a couple of spare boards, so I can do some messing around with a PC
parallel port driver. There's even a sample Z80 interface schematic in
the back of the manual that's probably adaptable...
Still concerned about timings though as some of the timing diagrams do
have upper constraints on signal changes (a lot of them only specify a
minimum limit). Time will tell if the PC's port is quick enough...
I think (wihtout trying it), you can go as slowly as you like.
-tony
10 Jul
10 Jul
1:51 p.m.
On Fri, 2005-07-08 at 23:51 +0100, Tony Duell wrote:
Yep, that's certainly what it looks like anyway...
Having had a better read of the timing diagrams, that looks to be the
case. The critical one is the time between the target asserting REQ and
the controller asserting ACK - but in all 'simple' SASI/SCSI interfaces
that bit's done in hardware via a flipflop, so time taken for the
software-driven part doesn't actually matter.
Time to throw a few chips together and have a play around I think...
cheers
Jules
Actually,
I'd not heard of the blue-cased machines before. Maybe they're
in standard off-the-shelf instrument cases rather than with RML logos,
panel cut-outs etc.?
The black case is a standard (Vero iIRC) case. It was designed to take a
19" rack chassis, but RML just mounted the PSU and card guides inside.
OK, the cutouts and logos are custom, but that's all. Oh, Xebec
board in my fileserver is an S1410, for which I have a manual
and a couple of spare boards, so I can do some messing around with a PC
parallel port driver. There's even a sample Z80 interface schematic in
the back of the manual that's probably adaptable...
Still concerned about timings though as some of the timing diagrams do
have upper constraints on signal changes (a lot of them only specify a
minimum limit). Time will tell if the PC's port is quick enough...
I think (wihtout trying it), you can go as slowly as you like.
7110
days inactive
7120
days old
27 comments
4 participants
participants (4)
-
a.carlini＠ntlworld.com -
ard＠p850ug1.demon.co.uk -
julesrichardsonuk＠yahoo.co.uk -
williams.dan＠gmail.com