Searching for IBM S/23 Datamaster users

[RetroAND]

Thanks' for the rest of the information
..
As for a case to re enclose it, If you can get an old PC like tower case pre ATX that would be the easiest thing to modify. Yes you still might have to do some metal work. Likely make a mounting bracket for the diskette drive and make some new port holes and a drive hole, build some power adapter cables, and etc.
..
There are other cases that have 8 inch drive bays , but there availability could be an issue, Of course if you have access to a custom case/enclosure manufacture that could also be a way to go at a higher cost of course.
..
As for replacing the missing keyboard issue from a Sys/23 or having to build one your self from scrap parts.
> The simplest thing might be to make an Ps/2 keyboard to Sys23 adapter instead of making a hole new keyboard.
>There are a flew other IBM keyboard that will also work with the Sys/23 but trying to get one at a realistic price won't be very likely.
..
Your note about the monitor and the System23 video standard could very complex to the to explan in a test message. If you can live with a monitor adapter ( to SCART) and an existing external monitor that is the easiest.
>If not or you want to use an original monitor with a resolution better that 640x240 pixels I would try to find a old IBM monochrome display monitor, make a cable and manually recalibrate it.
..
Hopefully this message helps you along it this process.

Hello Nathan,

I will seek for the cases you say that can hold 8" drives. I also have two other drives that would like to encase as a separate project, so it may be a good opportunity to find cases for both of them.

The keyboard issue will be the most complicated thing to solve. And I am not that good with microcontrollers, so I will eventually ask for help.

The picture issue is almost resolved, we found the RGBtoHDMI project at github and with a combination of it and the logic from my RGB interface will solve the issue of the missing screen parts not only for this caseless 5322, but with monitorless 5324 models too.

Thank you for the information!

 

[N. Brown]

A note about using other cases for the rehousing of IBM system 23 Computer system components.
..
If you cant wait to get or find the right size of enclosure (or case) they you will likely have to modify
an existing case or build a new one your self. As trying to find a reasonable price desktop vintage
computer case with internal 8-inch disk drive bays is almost impossible today.
S-100 Bus Cromemco System 100 and 300 series would be a good example of an existing unit.
If you are looking for a small mini size system then a case like a NABO 1100 would do.
The quickest and sometime the cheapest way to go is to modify an existing full size fat PC tower
case or to build a case your self (depending on your skill set, time and resources as well).
..
Metal working Tools
---------------------- (Likely needed for the building or retrofitting a existing case)
[] A drill and metal cutting drill bits.
[] A hand held hell grinder/cutter
[] A Hand held Jig Saw with metal blades.
[] Access to a vacuum cleaner that has a hose
[] An old school Hax saw still might come in handy
[] The attachment for a drill to cut round holes could also be handy for new fan vent holes.
[] A way to bend metals : A large Vice attached to a long counter plus two old wooded fence boards is good
enough to make up replacement drive mounting brackets and drive mounting plates.
** Or access to an actual metal bender.
[] Safety tools: goggles, face mask, Get A Tentis short is recommended when working with rusty metals.
-------------- (Quick access to first aid kit in necessary).
..
Enclosures/Case Parts:
---------------------------
Old PC cases parts – Panels, and other parts like drive bays and there covers, old power supplies, Fans
and fan gards, screws, back bay ISA slot plate, and spare recovered rails and ISA slot covers.
Electrical: Power connectors, extra cables ends, recovered leads, Cables and cable ends, Etc.
Weather striping tape comes in handy to fill in small gaps (to stop air/dust flow).
Old Vinal floor cover chunks/parts can work as a installation device (so I note it note as well).
Self sticking Wall Paper/Internal clubber liner roll/tape comes in handy as well.
..
Type of Case normally still available
--------------------------------------------
Modified PC cases for use with 8-inch Floppy disk drives
Any larger so called fat tower case that has at least 5 half height 5.25 inch drive bays with a 10mm gap
near there top and bottom is great (or case with 6 bays for half height 5.35 inch drives is better yet) .
Hopefully the case comes with the bays face plates included, if not that might cost a bit more.
If you are going to use the existing system 23 power supply as well make sure the case has the room
for that as well , so no short slime line cases.
To waste an real still working IBM PC Server 320 case to just reinstall an System 23 mother board into
it some would question that, but if you can find a intact case with power supply and its bay covers,
that’s should be OK to use with less work involved.
..
A note about mounting 8-inch diskette drives into a bank of 5,25 inch drive bays.
When you mount 2 half height 8-inch diskette drives in the UP/down wards position in the
Hor-position(/across ways) 5 drive 5.25 inch half height drive bays length wise there will be about a
35mm gap to be filed in. A Scrap plastic pice the right length will work to fill this gap. Of course you will
have to find a way to mount it (via : super glue or another method). As for power conversion I will
leave that omitted for now, +12 volt to +24 volt conversion, -24Volt to +24 Volt conversion circuits are
quite simple. As for the drives AC power connectors I have already posted a circuit for ANSI/ECMA type
8-inch diskette drives.
..
A lager none PC computer case would be a IBM 3274 type 52 or 52 Communication controller. This
unit is the about the size on an old 1970's microwave oven, but it has lots of space internally to work with.
Stuff like a motherboard mount and a hole cut for the systems ISA slots and the etc., still has to be done
with it as well
..
Using Other computer Cases with drive bays enclosed within it.
In a previous message I sent some indirect links to what other cases look like and there seller’s information.
So we will just continue from there but the most easiest case to be found would likely be a fat PC tower like
case with externally accessible 5 bays with gap spacing to 6 bays for half height 5.25 inch drives should be
able to make work, of courses with enough internal space for the power supply.
..
Older IBM System Cases that can possibly be modified (depending on modifiers skill level)
This operation is going to a little more complex that modifying a Fat PC type of full height tower case.
> The System 36 model 5363 an IBM AS/400 look like system – Has room internally but new front
bay covers will have to be made up as well as other unlisted modifications. The external and internal
case parts will have to be cut for access to the 8-inch drive that is to be mounted internally.
> The System 36 Model 5364 is a smaller IBM PCAT like case so doing all internal system 23 parts will be hard.
> If you have worked with tight spaces and only a signal full height 8-inch diskette drive then a
IBM PS/2 model 95 case can be made to work if you plan to use its newer power supply.
> A similar harder fitting thing is if it is the only affordable choice is to use an old IBM PS/2 model 80-386 case
and you plan to reuse its internal power supply to save space and power. Not much spare space in the case
for this application.
..
For external Diskette drive enclosures
In my previous (direct) message I hinted about using recovered external drive enclosures/cases,
Like: IOMega 8-inch Bernoulli Box/unit cases, IBM 6360 Units, A Xerox 820 computer systems external dual
8-inch full height floppy disk drive case.
> External Full height Commodore PET class External deal 5.25 inch floppy disk units still show up for sale
on Ebay.com. These cases need a slight modification to mount two half height 8-inch floppy disk drives
and of course you will have to make a fill in plate for the front on the case. Cut a connector hole on the
back of the case for the 50-pin standard drive connector.
..
Well that is about all the guidance I can give today (with this short guide line).
..
I wish you good luck in hunting down the parts you need to do this job at a realistic price.
So hopefully we will see a future update in a mater of time.
..

 

[ggs]

I should note that during the development of S/23, we had plexiglass/acrylic cases that the IBM model shop made for all of the prototypes (yes, we could see all the guts). They weren’t exactly pretty but they were roughly the shape of the final enclosure. None of the joints were hidden. They were either screwed or glued.

 

[RetroAND]

I received a word-processing card for the System/23 and on close inspection it seems to be correct. However, I lack the cables to connect it to the video connector of my American motherboard (I can't place it on the other one due to the crystal frequency). Please, could anybody with the card and the wire harness make a pinout diagram of it so I can reproduce it on my own? That would enable me to test the hardware.

 

[N. Brown]

If you get no other reply, you could try to reach out to any museum that has a sys23 and inquire in they have a system with the Word processor card installed.
Maybe if you are lucky, they will send a couple of pictures of both sides of that cable. (But if not then the rest of this message will likely apply).
..
Most of the signal pins including the undocumented ones run right to the card's other connectors equitant pin. (I have attached a photo down below.)
..
The other hidden and unlisted signal pins are for the two cards to cooperation with each more or less without the use of the systems ISA bus like card slots.
Signaling stuff like: Yes, there is a Word processing Video expansion card connected, Syn of the video signals, and Card video is Active or not active, etc.
Signal stuff like this. Using just a multimeter you should be able to figure out the other unlisted signal functions.
..
With a multimeter and some manual signal line tracing, you may discover that this unit has 3 unlisted Signal Grounds like lines.
> 2 Signal grounds have additional circuity for Monitor Type Detection used on a large display option (, likely with a different Word processing like card)
> One + 5Volt DC signal line
> One signal line reserved for Base Brightness output (that can override the Highlight signal)
> One signal line reserved for Base Contrast output on a graphics board option
> One Frame Ground / Ring System input
> Other video circuit cooperation control signal lines that connect between the unit's video expansion card options (that are not on the ISA like expansion bus).
..
The photo and file attached should help you if you must manually make up a cable, since you made up a modified (SCART) RGBtoHDMI project this should
prove to not be too much of an issue.
..
IF you decide to make up a cable and want to protect the systems components then just use some isolation diodes to protect it. (Old school isolation diodes
is just a commonly used power supply diode connected to a Zenon diode more or less).
..
Hopefully some additional information becomes available.
Well, that's about all for now.

 

[RetroAND]

Hello!

I have good news indeed. Yesterday I accomplished the task of having stable image through HDMI from my American Datamaster! This is a leap forward in the conservation of these machines, especially the units with broken tubes or the 5324 which have lost their monitors. Enjoy the view!



I am working towards a replacement keyboard, as my American 5322 lacks it. The first phase is to trick the motherboard to think it has a keyboard attached so the 35h code ceases to blink. I have started to work towards this goal by using a 8749 as the core of the design, as this way I can reuse the original code. I have, however a newer design prepared but it should require the coding and with the amount of things left to do, I cannot do it in the deadline of two months.

Regards!

 

[RetroAND]

I have more news for today:

In the first place, I have tested the video adapter with my Spanish 5322 and made the European profile, so now all the Datamasters can be used with an HDMI screen.

Last but not least, today I have finished the first phase of my keyboard, which consists in a small board containing the microcontroller wired into the connector, tricking the motherboard to think it has a keyboard attached. Nowm I have to receive more components in order to continue it, but once this is done, only the matrix will remain to be done, and with that a full keyboard shall be reproduced.

With both the video and keyboard developments, I have seen the PROC START from the American machine for the first time since it arrived home.

Regards!

 

[N. Brown]

I have more news for today:

In the first place, I have tested the video adapter with my Spanish 5322 and made the European profile, so now all the Datamasters can be used with an HDMI screen.

Last but not least, today I have finished the first phase of my keyboard, which consists in a small board containing the microcontroller wired into the connector, tricking the motherboard to think it has a keyboard attached. Nowm I have to receive more components in order to continue it, but once this is done, only the matrix will remain to be done, and with that a full keyboard shall be reproduced.

With both the video and keyboard developments, I have seen the PROC START from the American machine for the first time since it arrived home.

Regards!

With the keyboard if you have to make the keyboard circuit matrix for the keyboards microswitches from scratch and them solder in the key microswitches this will take some time to remake from scratch.
You could try to cheat a bit and reuse most of the keyboards parts on a old IBM PCXT keyboard, Other then the MCU on its existing MCU PCB module and of course some of the missing key cap lettering will have to be made up.
If you look up the keyboards scan codes for the original IBM PC keyboard (83-key) variety and then compare it to the one the system 23 keyboard uses you will see that they are very similar. Other then the Functions keys, cursor
keys and Etc. locations. If you can do this this would help save you some work.
..
In your above reply you did not state which keyboard model it is or wither it has the 40-pin MCU or the 28-pin MCU installed, (Some clones have no MCU install, but they have some additional logic).
The keyboard schematic from www.kbdbabel.org for the IBM 5291 Data terminal is close to the real thing but there is no MCU and there is no keyboard ID circuit in it. (So this could cause some issues with the systems keyboard
firmware rut-teens and operating system issues).
..
Hopefully you have access to a programmer that can read the existing MCU (Microcontrollers) internal programmed code, then it will be easy to copy onto another MCU that is compatible with it. (That is, Pinout and Instruction set Compatible with the original one that is used in the systems 23 keyboard).
..
AS far as using the original PC keyboard IBM Byte magazine December 1986 issue has BUILD THE GT180 COLOR GRAPHICS BOARD PART 2: HARDWARE has a simple IBM PC compatible keyboard Interface that is done in standard 74XXXX family logic gates, so this interface should be easy to adapt in the future. like just adding a prom to do character conversion. Schematic was on PDF pages 98 and 99 I think.
..
Well that is about all for now.

 

[RetroAND]

With the keyboard if you have to make the keyboard circuit matrix for the keyboards microswitches from scratch and them solder in the key microswitches this will take some time to remake from scratch.
You could try to cheat a bit and reuse most of the keyboards parts on a old IBM PCXT keyboard, Other then the MCU on its existing MCU PCB module and of course some of the missing key cap lettering will have to be made up.
If you look up the keyboards scan codes for the original IBM PC keyboard (83-key) variety and then compare it to the one the system 23 keyboard uses you will see that they are very similar. Other then the Functions keys, cursor
keys and Etc. locations. If you can do this this would help save you some work.
..
In your above reply you did not state which keyboard model it is or wither it has the 40-pin MCU or the 28-pin MCU installed, (Some clones have no MCU install, but they have some additional logic).
The keyboard schematic from www.kbdbabel.org for the IBM 5291 Data terminal is close to the real thing but there is no MCU and there is no keyboard ID circuit in it. (So this could cause some issues with the systems keyboard
firmware rut-teens and operating system issues).
..
Hopefully you have access to a programmer that can read the existing MCU (Microcontrollers) internal programmed code, then it will be easy to copy onto another MCU that is compatible with it. (That is, Pinout and Instruction set Compatible with the original one that is used in the systems 23 keyboard).
..
AS far as using the original PC keyboard IBM Byte magazine December 1986 issue has BUILD THE GT180 COLOR GRAPHICS BOARD PART 2: HARDWARE has a simple IBM PC compatible keyboard Interface that is done in standard 74XXXX family logic gates, so this interface should be easy to adapt in the future. like just adding a prom to do character conversion. Schematic was on PDF pages 98 and 99 I think.
..
Well that is about all for now.

Hello Nathan,

I considered that, but the microcontroller has to be aware of the topology of the matrix, and the matrix of the keyboard I studied and the ones from PC/PC-XT are different. It would involve more work instead.

I know the matrix since I had to discover it for the emulator. I have keys from old electronic typewriters and plan to wire the matrix with wires under the board they are soldered. The microcontroller is already programmed, and the keyboard is detected as is. The only things that I do require for continuing the assembly are boards to solder the keys on and a 8:1 multiplexer that mimics the function of that SLT/ MST the original keyboard had.

This is made using an equivalent of the original microcontroller. However, I also designed a controller board for it, just I don't have time to program it. If somebody wanted to help, I would pass the info in order to proceed with it.

Then the other thing I would need to end with the keyboard development would be the 5324 addons. With them I would be able to know what signals are altered between the motherboard and the keyboard.

That would be all I have to report on the keyboard for the moment. I have something else to talk about, and that is memory.

I am working on a replacement for the 32KB modules, as I have found a user who was tricked and his card was missing when he bought the machine. In order to continue with the diags, he needs the 32KB modules as minimum.

I am going the fake DRAM route with SRAM and a latch, for the time being I haven't succeed, but I think my design wasn't correct. Now I made a much more primitive and unoptimized design and I hope it will work.

The developed modules will also hep when we start to have failing TMS4132 memories. I test them stealing the signals from the bus through a ribbon which goes to the prototyping board.

Well, I hope that's enough for now. I am excited for this summer, I hope we discover many, many things.

Regards!

 

[N. Brown]

Hello Nathan,

I considered that, but the microcontroller has to be aware of the topology of the matrix, and the matrix of the keyboard I studied and the ones from PC/PC-XT are different. It would involve more work instead.

I know the matrix since I had to discover it for the emulator. I have keys from old electronic typewriters and plan to wire the matrix with wires under the board they are soldered. The microcontroller is already programmed, and the keyboard is detected as is. The only things that I do require for continuing the assembly are boards to solder the keys on and a 8:1 multiplexer that mimics the function of that SLT/ MST the original keyboard had.

This is made using an equivalent of the original microcontroller. However, I also designed a controller board for it, just I don't have time to program it. If somebody wanted to help, I would pass the info in order to proceed with it.

Then the other thing I would need to end with the keyboard development would be the 5324 addons. With them I would be able to know what signals are altered between the motherboard and the keyboard.

That would be all I have to report on the keyboard for the moment. I have something else to talk about, and that is memory.

I am working on a replacement for the 32KB modules, as I have found a user who was tricked and his card was missing when he bought the machine. In order to continue with the diags, he needs the 32KB modules as minimum.

I am going the fake DRAM route with SRAM and a latch, for the time being I haven't succeed, but I think my design wasn't correct. Now I made a much more primitive and unoptimized design and I hope it will work.

The developed modules will also hep when we start to have failing TMS4132 memories. I test them stealing the signals from the bus through a ribbon which goes to the prototyping board.

Well, I hope that's enough for now. I am excited for this summer, I hope we discover many, many things.

Regards!

Thanks for the updated status on your project
..
As far as the keyboard suggestion if I had known how far you where with your replacement unit then I would have not likely suggested it. The only reason I did suggest it was that most of the companies that are still servicing IBM equipment of that age under contract just replace the keyboard control and its PCB module in a newer existing IBM compatible keyboard.
..
In your message from above you noted that you just I don't have time to program it. If somebody wanted to help, I would pass the info in order to proceed with it. OK maybe you should post for help with that on this forum as my spare time is quite limited.
At one time there use to be firms that made replacement microcontroller for keyboards that could also provide the program services to clone them. (I don't know if any of them are still in business and if I still have their contact information in the archive of paperwork)
..
I have read your above note above about the DRAM memory board, but I have no additional information to share on that subject and my
internet access time is limited today. (To much to do again)

 

[ggs]

I would use a keyboard designer (there’s a website that aids in the layout/design). I believe it provides the design for the matrix and you can choose the key switches. I also believe that there are somewhat off the shelf micros that do the keyscan but they probably interface using USB.

 

[RetroAND]

Hello,

It's been a while since I last wrote here.
I was mainly busy with the keyboard stuff...
However I have been in a trip in the later two days and as a result I am now the proud owner of an IBM 5324, the tower model of the Datamaster. It has no keyboard or monitor, however I am working to solve these issues by the end of the summer.
I also plan on document it as much as possible given the short amount of time.

Regards

 

[N. Brown]

Hello,

It's been a while since I last wrote here.
I was mainly busy with the keyboard stuff...
However I have been in a trip in the later two days and as a result I am now the proud owner of an IBM 5324, the tower model of the Datamaster. It has no keyboard or monitor, however I am working to solve these issues by the end of the summer.
I also plan on document it as much as possible given the short amount of time.

Regards

Thanks' for the status update.
I known what it is like to do a keyboard matrix manually by hand (using wire rapping by hand) and then bread board the rest of the I/O logics.
I also know what other work issues and other unrelated job timing issues can do to that planed rough schedule.
So hopefully by the end of the summer, you do get it done if not sooner.
..
Well that's about all for now.

 

[RetroAND]

After a few days searching through IBM docs, I found interesting information about the Datamaster, specifically most of its important dates!

Release of the 5322: 28 July 1981
Release of the 5324: 18 May 1982
Discontinuation of the 5322: ongoing, around 1983
Discontinuation of the 5324: 7 May 1985

This may seem very little, but until now we didn't have a proper chronology of the lifetime of the series of computers. I also found some dates about the peripherals.
These may be complemented with further dates as my investigation continues.

Regards!

 

[RetroAND]

Yesterday, I found more information about the withdrawal of the 5322!

Discontinuation of the lower-spec models: 18 May 1982
Definitive discontinuation: 1 June 1983

The 5322 wasn't even two years in the market! Discontinuation began after less than a year...

If you want to take a peek, look at the following document.

 

[RetroAND]

I have dumped the firmware of the 5247 communications card. They will be available soon at bitsavers.

 

[Shadow Lord]

bg

I have dumped the firmware of the 5247 communications card. They will be available soon at bitsavers.

Excellent. I ma glad the card was helpful. Next step a DM23-IDE card....

 

[RetroAND]

bg

Excellent. I ma glad the card was helpful. Next step a DM23-IDE card....

I don't think a Datamaster-IDE card is possible, but on the other hand a Raspberry Pi identifying itself as a hard disk drive is possible. The original disk drive was a computer itself, so we should regard it as some kind of shared NAS.

I have to scale down the operations, but could try to smuggle the card to where I will be being for the next two years and study it before my studies begin. However, that would be a hardware study only. I don't know how to disassemble those three ROMs because they seem to contain some kind of header. We need to disassemble the whole set of ROMs in the motherboard first in order to know where the entry points are. Unfortunately, I don't have time for this task anymore... but if some volunteers could create a joint effort to do the software task, I could provide all the data I find in my hardware and get the job done between all of us.

Once everything is documented, we would need to design the program of the Pi, and decide if we go with a Linux application or a baremetal solution, and design a hat. I may be able to obtain the firmware of the 5247 itself through a contact, so I think that might help too. When designing the hat, I think we should add the full four connectors to it, since the original disk drive unit was capable of being shared between four Datamasters.

If anybody is open to help, I would be glad to share my data.

 

[N. Brown]

bg

Excellent. I ma glad the card was helpful. Next step a DM23-IDE card....

Ya an IDE interface card for the IBM Datamaster (System/23) would take quite a bit of work to make up (due to the fact that the systems hardware emulation is based of an IBM mini computers compatible storage units).
The original documents for any 3RD party devices that where once made for this system are now nonexistent. An SASI bus interface was once done by 3Rd party, but that data is also now nonexistent. The hardware part would be simply, but the device driver will be a big issue to recreate. Knowledge of the IBM Systems 34 combination Floppy and hard disk controller boards would be recommended if you plan to real IBM system 34 type of drives. .

 

[N. Brown]

I don't think a Datamaster-IDE card is possible, but on the other hand a Raspberry Pi identifying itself as a hard disk drive is possible. The original disk drive was a computer itself, so we should regard it as some kind of shared NAS.

I have to scale down the operations, but could try to smuggle the card to where I will be being for the next two years and study it before my studies begin. However, that would be a hardware study only. I don't know how to disassemble those three ROMs because they seem to contain some kind of header. We need to disassemble the whole set of ROMs in the motherboard first in order to know where the entry points are. Unfortunately, I don't have time for this task anymore... but if some volunteers could create a joint effort to do the software task, I could provide all the data I find in my hardware and get the job done between all of us.

Once everything is documented, we would need to design the program of the Pi, and decide if we go with a Linux application or a baremetal solution, and design a hat. I may be able to obtain the firmware of the 5247 itself through a contact, so I think that might help too. When designing the hat, I think we should add the full four connectors to it, since the original disk drive unit was capable of being shared between four Datamasters.

If anybody is open to help, I would be glad to share my data.

A Datamaster-IDE card is possible but the amount of work to get it done will be quite large due to the firmware Routens that would have to be done and the emulation of the original
IBM system/34 actual Storage controller modules controllers' registers (to be fully compatible with the original system 23 software). Of course you could always use a newer MCU to emulate this hardware registers but the time and resources to that could be an issue.
..
Emulation of the external 5247 unit itself should be a lot easier to re implement as this device uses the standard IBM external SNA like network storage device protocol. AS far as I know the data for the clone devices are not available public-ally online. This process will take more that dissembling the 5247 firmware but also part of the hardware as well. But if you can find another firm drive that uses a protocol that is close the same as on the IBM 5247 that would save you a whole lot of work.