Searching for IBM S/23 Datamaster users

[RetroAND]

The American unit is being diagnosed further. It has four bad system ROMs and three more of which I am suspicious. The character generator is gone for good too. I have no adapters for them, so I have to wait for some. I have designed one new that is able to replace up to 8 ROMs with two ICs, so it will come handy.

The French one seems to be suffering from a handshake issue so the DMA device keeps waiting for an event that will never occur. Strangely, the /SACK signal on the 8203 ceases to pulse after DMA is granted access to the bus. Does anybody know what could be happening to this component to stop handshaking?

 

[RetroAND]

I correct my previous message. Six ROMs, including the character generator are bad, and three more could also be in trouble.

 

[RetroAND]

The French Datamaster from Lucas has left the test 05h and is ready to have its ROMs restored. After that I will need to wait for a few spares and I think it will be ready to be returned.

 

[RetroAND]

Today I found the I/O ports for the NEC 765 and the 8255 in the FDC card. As always, I have documented them in my article.

 

[Al Kossow]

Is there a copy of the OEM drive manual anywhere?
The drive has two boards, one contains an Intel microcontroller and the other is the read/write channel
I've never come across FE manuals for the 5362. The 5360 is much better documented in the files on bitsavers.
There are some pictures of the 21ED drive under dasd/21ed there as well.

Thanks to a friend inside IBM with connections to the IBM Archive, IBM 5247 Disk Unit Service Manual SY34-0243-0
has been scanned which contains a lot of information on the 21ED. That gives me revived motivation to put a breakout
board on the drive attached to my 5362 to understand the interface protocol a bit better.

 

[RetroAND]

Thanks to a friend inside IBM with connections to the IBM Archive, IBM 5247 Disk Unit Service Manual SY34-0243-0
has been scanned which contains a lot of information on the 21ED. That gives me revived motivation to put a breakout
board on the drive attached to my 5362 to understand the interface protocol a bit better.

I established contact with some owners with a 5247. I might have the system dumps soon for them to be archived. The processor seems to be a 8085, same as the Datamaster and it has 32KB ROM and 64KB DRAM.
The motherboard does not have comms per se, it is done through cards in a shared bus. If we could understand the 5247 motherboard, we could be able to provide Datamasters with a way to store contents without passing through the old floppy drives.

At the same time, a collaborator of mine has developed an intelligent adapter to connect the floppy disk controller of the Datamaster with a gotek. The prototype should arrive soon for it to be tested with my American unit.

I have also acquired a word-processing card which will be studied in order to learn how it works. Expect more developments in the emulator side. Also, it would help rewrite the rules on how the computer adds RAM.

Also to note, I sketched a design to replace the 32KB DRAM board. I will construct a prototype to test it as soon as my American unit is repaired.

 

[N. Brown]

I established contact with some owners with a 5247. I might have the system dumps soon for them to be archived. The processor seems to be a 8085, same as the Datamaster and it has 32KB ROM and 64KB DRAM.
The motherboard does not have comms per se, it is done through cards in a shared bus. If we could understand the 5247 motherboard, we could be able to provide Datamasters with a way to store contents without passing through the old floppy drives.

At the same time, a collaborator of mine has developed an intelligent adapter to connect the floppy disk controller of the Datamaster with a gotek. The prototype should arrive soon for it to be tested with my American unit.

I have also acquired a word-processing card which will be studied in order to learn how it works. Expect more developments in the emulator side. Also, it would help rewrite the rules on how the computer adds RAM.

Also to note, I sketched a design to replace the 32KB DRAM board. I will construct a prototype to test it as soon as my American unit is repaired.

Well some of the secrets of the IBM 5247 external (kind of) networkable hard drive are starting to be reviled now is right on.
Hopefully the collaborator of of yours that has developed an intelligent adapter to connect the floppy disk controller of the Datamaster with a gotek will soon get posted. (As the ones we used at work are not publicly available).
It cool to read that you have acquired a word-processing card, but your short entry did not state wither it has the video through connectors on the front of the card or not, so which one you have is currently unknown.
..
Your replace design sketched of a 32KB DRAM board might come in usefully to others with an incomplete system or a system that has one memory board missing.
..
There are a lot of IBM 51TD drives that still show up on Ebay at a cost of course, so by now I thought some body would have done up a adapter design that is publicly accessible.
The one listed at https://github.com/inmbolmie/FloppyRider is a lot more complex then the our own inhouse ones that we used at work connected to an IBM compatible PC.
..
If needed (an old design file that doesn't use a MCU, so it is another so called Old school) has been attached.
I have attached a older you could call it a Automatic 8-icch diskette drive power (and disk saver) motor control circuit. To modify this circuit for the IBM 2 phase drives like the 51TD should be quite simple.
You will still have to add a IBM drive selection signal (+24Volt) circuit to this design. But that should be a rather simple circuit to do up. Convert the standard /Driveselect signal to the +24V Dc that the drive requires.
The other IBM signals are fairly easy to convert using an inverter. If necessary I can dig up an Index pulse to drive ready circuit as well (this is normally built in on the Floppy disk controller when used with 51Td drives).
Some on the circuitry can be updated and some of the components used can be replaced by optocouplers as well.

 

[N. Brown]

I established contact with some owners with a 5247. I might have the system dumps soon for them to be archived. The processor seems to be a 8085, same as the Datamaster and it has 32KB ROM and 64KB DRAM.
The motherboard does not have comms per se, it is done through cards in a shared bus. If we could understand the 5247 motherboard, we could be able to provide Datamasters with a way to store contents without passing through the old floppy drives.

At the same time, a collaborator of mine has developed an intelligent adapter to connect the floppy disk controller of the Datamaster with a gotek. The prototype should arrive soon for it to be tested with my American unit.

I have also acquired a word-processing card which will be studied in order to learn how it works. Expect more developments in the emulator side. Also, it would help rewrite the rules on how the computer adds RAM.

Also to note, I sketched a design to replace the 32KB DRAM board. I will construct a prototype to test it as soon as my American unit is repaired.

In theory if some one wants to use System23 Basic and do some custom hard interfacing you could build a IBM 72 MD Magazine Drive emulation unit using a MFM hard disk instead of a bunch of diskette.
The IBM 72 MD Magazine Drive has some additional signal lines that are omitted on the system/23 floppy disk controller, but most of the system 34 basic function call still get passed through the System 23 basic interrupter correctly.
I don't know if a forum reader wants to take this project on, but I have now posted this entry. So I guess we will eventually find out.

​

 

[N. Brown]

Today I found the I/O ports for the NEC 765 and the 8255 in the FDC card. As always, I have documented them in my article.

This short entry is basically just an quick outline on a possible standard PCAT floppy disk adaption for the IBM Sys 23 computer system

Cool, Have you check to see what pins are active on the NEC 765 Controller, if all the pins are active then it will be quite simple to connected a standard floppy disk drive to
the system through a custom adapter board that the NEC 765 Controller plugs into. Note : There are some design differences in the controllers depending on there part numbers.
The differences in the floppy disk drive controller boards will/can make the adaption of a standard PCAT type floppy disk drive to the System 23 more complex.
..
NEC UPD 765 FDD Ctrl
Pin 39 [] //RW/Seek ..................... When Seek is high (logic 1): Step, DIR, 2Side(ed), Trk0 input are active.
Pin 38 [] LCT/DIR .......................... LCT is basically the same as TG43 signal.
Pin 37 [] FR/STEP ..........................
Pin 36 [] HDL ................................. The //RW/Seek signal is normally And gated with FR/STEP signal to generate +Step signal for the floppy disk drive
Pin 35 [] Drv ready ..................... The //RW/seek signal is normally And gated with LCT/DIR signal to generate the DIR signal for the floppy disk drive
Pin 34 [] WRprot/2Side .............
Pin 33 [] FLT/TRK0 ...................... The RDY signal could be pulled up to +5vdc
Pin 32 [] PS0 ................................. The US0 and US1 and HDL signals might be not used on the version of the controller that you have.
Pin 31 [] PS1 ................................. The Side select circuity and/or the Head load logic could also be in other logics as well
Pin 27 [] Side select ...................
Pin 29 [] US0 Unit select 0
Pin 28 [] US1 Unit select 1
Pin 17 [] Disk index Input
..
Depending on the controller version not all of the signal pins are connected on the 765 controller chip, in that case other logic on the floppy disk controller are still doing that function.
For the 51Td drives if the IBm SwitchFilter signal is required an updown counter can be used to generated this signal when the head is stepped past track 64.
..
Hopefully these new entries help some one else out

 

[RetroAND]

Well some of the secrets of the IBM 5247 external (kind of) networkable hard drive are starting to be reviled now is right on.
Hopefully the collaborator of of yours that has developed an intelligent adapter to connect the floppy disk controller of the Datamaster with a gotek will soon get posted. (As the ones we used at work are not publicly available).
It cool to read that you have acquired a word-processing card, but your short entry did not state wither it has the video through connectors on the front of the card or not, so which one you have is currently unknown.
..
Your replace design sketched of a 32KB DRAM board might come in usefully to others with an incomplete system or a system that has one memory board missing.
..
There are a lot of IBM 51TD drives that still show up on Ebay at a cost of course, so by now I thought some body would have done up a adapter design that is publicly accessible.
The one listed at https://github.com/inmbolmie/FloppyRider is a lot more complex then the our own inhouse ones that we used at work connected to an IBM compatible PC.
..
If needed (an old design file that doesn't use a MCU, so it is another so called Old school) has been attached.
I have attached a older you could call it a Automatic 8-icch diskette drive power (and disk saver) motor control circuit. To modify this circuit for the IBM 2 phase drives like the 51TD should be quite simple.
You will still have to add a IBM drive selection signal (+24Volt) circuit to this design. But that should be a rather simple circuit to do up. Convert the standard /Driveselect signal to the +24V Dc that the drive requires.
The other IBM signals are fairly easy to convert using an inverter. If necessary I can dig up an Index pulse to drive ready circuit as well (this is normally built in on the Floppy disk controller when used with 51Td drives).
Some on the circuitry can be updated and some of the components used can be replaced by optocouplers as well.

Precisely, my collaborator is inmbolmie. I asked him if he could help us in the task of creating a gotek adapter and he used his FloppyRider and an old PC to simulate the Datamaster floppy drive controller. He is a very kind person.
As stated, I should receive the prototype very soon. We will know then if the simulation was close enough to be compatible with the real unit. In any case, we all should be grateful for him for his work and dedication. He joined and made his our necessity and worked hard to achieve the result. So three hurrah for inmbolmie!

Regarding my replacement for the 32KB DRAM module, it is still a work in progress. I will be updating this in that regard, but I expect to have something very soon.

 

[N. Brown]

Precisely, my collaborator is inmbolmie. I asked him if he could help us in the task of creating a gotek adapter and he used his FloppyRider and an old PC to simulate the Datamaster floppy drive controller. He is a very kind person.
As stated, I should receive the prototype very soon. We will know then if the simulation was close enough to be compatible with the real unit. In any case, we all should be grateful for him for his work and dedication. He joined and made his our necessity and worked hard to achieve the result. So three hurrah for inmbolmie!

Regarding my replacement for the 32KB DRAM module, it is still a work in progress. I will be updating this in that regard, but I expect to have something very soon.

Right On - Thanks for the update reply
..
To save some time I won't post the Index signal to drive ready signal circuity then , as it might not be needed.
..
We are looking forward to your next update - until them good luck.

 

[RetroAND]

Further progress with @Lulu's Datamaster.

I restored the ROM section and got greeted by the diagnostics screen. I have a keyboard error, but after testing it with our own enclosure and keyboard we found out it is responsibility of the motherboard. By the way, the character generator also seems to be corrupted.



With small steps, but I progress enough to be able to repair it little by little. I think it will be able to boot by itself soon.

 

[RetroAND]

Right On - Thanks for the update reply
..
To save some time I won't post the Index signal to drive ready signal circuity then , as it might not be needed.
..
We are looking forward to your next update - until them good luck.

You can post it. We need as much information as we can assemble. The prototype arrived yesterday, but I was told it may not work as expected due to timing constraints.

I am a bit busy repairing two Datamasters at the same time, as well as designing the new 32KB memory cards and also making reverse-engineering tasks - as well as my daily tasks. However I am willing to listen to new projects and other things that could help the scene.

By the way, soon we will also start repairing an IBM 51TD.

 

[RetroAND]

Hello again folks,

I have great news. This may be the final update on @Lulu 's French Datamaster.

After changing the 82c55 again we found that all of our spares but one were also defective. When fitted with the working one, we were able to boot into BASIC! But the screen was still corrupt, so we had to replace the character ROM with an EPROM by means of an adapter. Now the screen is free of corruption, is responsive and runs BASIC without issues.









Considering the damage this unit had, it was seriously faulty. However I made another feat with this computer, as I smashed the irreparability myth.

Now, I hope more units follow this one. Counting this one, in the community two units have been repaired this year, and at least three more should follow soon. Maybe even some more.

For the ones that are to follow my steps, I wish you luck and I renew my promise that I will help you by any means that are on my possibilities.

Thank you very much for following this repair.

 

[Shadow Lord]

Excellent work there!