• Please review our updated Terms and Rules here

Ti-990

Jim Hearne said:
Here is the real version of my programmers panel.
A couple of issues on the Issue 1 board, should be correct on the Issue 2 schematic which i've attached.
The circuit is an exact copy of the TI panel but it's laid out to no specific size as as the time i didn't have a chassis it needed to match.

Jim
Click to expand...
I forgot to attach the schematic i see.

Jim
 

Attachments

I have the Ti rack and one of Richards racks too. Agree the front panel does not fit Richards variant.

I saw those 990/5 cards for sale on Ebay too, I have a similar set and a later /10 set.
No experience of using a floppy emulator with the TFDC.
 
I seem to have acquired an X-Ten "X-PORTER/2" 990 coprocessor for running DX10 on a microchannel PS/2.

What are the odds this software still exists somewhere? Is there an active user community somewhere I might pester about it?
 
bear said:
I seem to have acquired an X-Ten "X-PORTER/2" 990 coprocessor for running DX10 on a microchannel PS/2.

What are the odds this software still exists somewhere? Is there an active user community somewhere I might pester about it?
Click to expand...
I have the software and a running image in my basement somewhere.
 
bear said:
I seem to have acquired an X-Ten "X-PORTER/2" 990 coprocessor for running DX10 on a microchannel PS/2.

What are the odds this software still exists somewhere? Is there an active user community somewhere I might pester about it?
Click to expand...
pics would be nice. i've never seen that x-ten product. great to hear someone has the software
 
I have several of the special serial cables too. Nothing special. Mini-Din to 9 pin. I'll try to make some time on Sunday to unearth it.
 
Any software for either the PC or PS/2 versions of the X-Porter or 7-XP boards, or the associated X-Link serial network and file accelerator boards, would absolutely be welcomed by me. I have some literature scans for these boards and will see if I can dig them up later, and maybe a pic of a 512K, 2 serial port 7-XP board. And I'm assuming X-Ten is what I know as Ten X, I guess. Not to be confused with the real estate people.

Ten X also made a TILINE COBOL accelerator for the 990, but that's another kettle of fish.
 
Last edited:
Ok, better late than never, but here's all the Ten X stuff I have ATM. This is a smallish Nextcloud server that will need to be moved to something bigger at a later date, and the URL will change, but have at it. I had issues with the uploads bogging things down, but downloads should go better. Just please be gentle. Board photos, ROM dumps, ftp and rsync links, etc. as part of a larger plan, soon.

dragonsweb.org

Ten X

Nextcloud - testing...
dragonsweb.org dragonsweb.org
 
Last edited:
kb2syd said:
I have the software and a running image in my basement somewhere.
Click to expand...
Any luck finding these?

ps. if anyone had trouble finding files in the link I posted above, please try again. Share links weren't set, but fixed now.
 
Last edited:
I've photographed most of the MDS990 manual in the TI Records archive at SMU (Dallas, Texas). This is the massive guide to microcoding for the 990/12.

Jim and I are independently writing disassemblers.

MDS990 has a listing of the Definitions file for MICASM (installed under XMCA command in SCI.) This defines the microcode formats and the macros to set all the bits. MICASM is generic and we have similar Definitions files for TMS7000.

Jim shared the microcode that DX10 loads to writable control store at boot (512 words of 64 bits.) From Dave Pitts.

I've got simple C code working to match the bit fields back to the macros. (Finding minimum set-- macros call other macros) Mine so far covers 40% of the DX10 runtime microcode.

Too soon to interpret what the code does, but XOP15 invokes something. DX10 accelerates the SVC call with microcode!
 
Has anybody yet dumped the microcode ROM of a 990/12? These are a bank of ROMs forming 2K x 64 bits, with other static RAM nearby (WCS). ROM is soldered not socketed.

I have an FX990/12 model CPU. What are the variations? The Depot Repair manual describes a 990/12 LR.
 
Actually, that was a dump from a DNOS run that Dave sent. (Also there was one from the MAP12 diagnostic program.) DNOS uses the WCS to implement enhanced I/O.

I'll have to see if DX10 uses the WCS; I didn't look at that. Dave made the dumps using the romtowcs function of sim990. The dump is enabled by the debugging flag, iirc.

Not aware of any variations of the /12 microcode roms. Good idea to check, though. Don't have any here regrettably.
 
I gather that source was an option.

"Source Installation" appears in the titles of about 100 manuals in the TI Records, at DeGolyer Library, SMU. They correspond to "Object Installation" manuals. Here's what I found for operating systems:

DX10 Operating System Source Installation, 1 September 1982. (Box 52)
DX5 Operating System Source Installation, 1 September 1981. (Box 52)
DNOS Source Installation Guide, 1 August 1981. (Box 48)
RX990 Operating System Source Kit Installation Guide. 15 December 1978. (Box 20)
TXDS Source Installation, 15 February 1978. (Box 31)
DNOS Online Diagnostics and System Log Analysis Tasks Source Installation, 15 December 1982. (Box 35. DX10 in Box 36)

Most other titles were for one of:

DBMS-990 , QUERY-990, TIFORM, BASIC, FORTRAN, COBOL, PASCAL, TIPE, 3270 Emulator, HDLC, AMPL, RTS, DXP, RPG-II, Sort/Merge, Rifle.
(Multiply by a number of operating systems and revision dates. Compiler Source and Runtime Source sold separately.)

Examples:

Model 990 Computer DX10 COBOL Compiler Source Installation, 1 April 1980. 939435-9701*B
Model 990 Computer DX10 COBOL Runtime Source Installation, 1 April 1980. 939436-9701*B


Citation:
Accession 94-08, Boxes 20, 31, 36, 48, 52. Texas Instruments Records, DeGolyer Library, Southern Methodist University. Dallas, TX.
 
The DX10 and DNOS System Design documents on bitsavers describe the contents and directory structures of the corresponding source disks. Comparing this info with available tape and disk images is in order.

Appendix B, http://bitsavers.org/pdf/ti/990/dx10/0939153-9701C_3.4_SystemDesignDocument_Oct81.pdf
Section 21, http://bitsavers.org/pdf/ti/990/dnos/2270512B_DNOS_System_Design_Document_Nov83.pdf

Somewhere on my daily driver desktop that is down for the count with a flaky mb atm, there is some document that describes the DX10 source distribution kit. IIRC, the sources were distributed on something like a DS25 or DS31 disk pack, or larger. Probably on 1/2" tape as well, but don't hold me to that. The source kit included listings on microfiche as well, I believe. I will try to dig reference out though, and get PNs for the source kits and contents if I can. I'm thinking it was a price list of some sort. These numbers, though, will also be referenced in the documents @FarmerPotato listed above.

Here is our problem: Any distribution media seems lost to time. The source kits don't seem to have been widely distributed outside of TI, and media doesn't seem to have been a priority, or even on the radar, for archivists, very unfortunately. It would be good for someone to triple-check the SMU archive just to be sure. Possibly one of the 990 OEMs like Timberline or Moore Business Systems would have gotten source kits, or possibly one of the companies which bought TI divisions with 990 relevance, but at this point anything 990 from those sources is likely only going to be found at estate sales, something someone took home with them as it was headed out the door to the dumpster, or from an employee sale of obsolete items from an upgrade, etc.

Which brings me to my point, I apologize if this is a horse I have beat to death elsewhere, but the 990/12 in the possesion of the Centre for Computing History in Cambridge, UK came from a development lab at TI Bedford. Most 990's were sold for production use. 990s had nowhere near the penetration in academic environments that DEC equipment did. What 990s TI did sell, for instance to the Georgia University System, was primarily, if not exclusively, used in business and administration departments, and my searching over the years indicates most machines were scrapped, documents pulped, and media recycled. The Blanden machine at the museum in Cambridge came out of a development environment, though, and possibly includes the sources for one or more operating system on disk or tape. Really, really needs to be checked out.

Other than that, our best hope is that one of the other computer museums has something lurking in a back room or warehouse that hasn't been cataloged and/or imaged yet. And if anyone here has anything they could share, that would be great, too.
 
Last edited:
That's an intriguing possibility. If I gain enough experience with the 990 at home, I would consider it a privilege to be able to work on restoring the Bedford 990.


While there's no substitute for commented source, I just took a crack at disassembling an object module library from one of Dave Pitts' installation tapes. It wasn't hard to recognize records of the object library, but is there documentation of the "tap" file structure?

Much less difficult, and higher up on my list, would be to disassemble the missing half of RX, a small (6K) operating system for TM990.

The MPP (Pascal) floppy archive has the commented Source disk 2 for RX, aka Pascal Runtime Library MPX. It is missing disk 1, but the object libraries are complete, and there is extensive documentation on its design. This came from RKN's scan of the Real-time Executive manual, and the MPP manuals.

I see no MPP manuals in DeGolyer. I bet they were not in Dallas or Houston to be archived with the rest in 1983.
 
Last edited:
You must log in or register to reply here.
Back
Top