New TMS9995 development kit

[jbdigriz]

Wichit Sirichote in Thailand has released a 9995 microprocessor kit based on his previous CDP1802 board. Interesting in that all I/O appears to be memory-mapped, not CRU, so, no 9901, uses the NS8250 UART for the serial port, and a GAL22V10 for memory and I/O address decoding. 40-pin expansion header, 8 digit LED display, flat panel keypad and add-on LCD display.  Schematics, sources, etc at https://www.kswichit.com/9995/9995.html . He has completed kits on Ebay, less the LCD and power supply, for $150. or DIY kits including the LCD for the same price. We all know how to search Ebay, right? Anyway, it's interesting to compare it to the original TI TMS9995 EVM, which seems to be scarcer than hen's teeth these days, or to the very appreciated designs of Stuart, pnr, and other folks here.

Edited June 1, 2021 by jbdigriz

[+dhe]

Thanks for sharing. Back when I was in college in my first EE class, we did assembler on a similar one board computer with a 8080.

 

I then later leaned of TI's "University Board" - I've tried for 30-40 years to find one, I did about 10 years ago find one, that I could afford for about $200, when I got it, it looked like someone had stored in an open shed, next to the ocean, it was eaten up with corrosion. I ended up giving it away to someone that wanted to try and restore it, and never heard back from him again.

 

So this is a really nice find.

 

Thanks,

[jbdigriz]

1 hour ago, dhe said:

Thanks for sharing. Back when I was in college in my first EE class, we did assembler on a similar one board computer with a 8080.

 

I then later leaned of TI's "University Board" - I've tried for 30-40 years to find one, I did about 10 years ago find one, that I could afford for about $200, when I got it, it looked like someone had stored in an open shed, next to the ocean, it was eaten up with corrosion. I ended up giving it away to someone that wanted to try and restore it, and never heard back from him again.

 

So this is a really nice find.

 

Thanks,

 

He teaches instrumentation at a university there. There's a bunch more interesting projects that he and looks like maybe some of his students have done, on his main page. Some device programmers, for example.  Other uP kits, microcontrollers, sensor interfaces, etc. Thoroughly documented, schematics, etc. Neat stuff. Trying to remember where I've seen his name before. Hackaday, maybe.

 

Yeah, I'd like to get one of those University Boards, too, and I keep running across them, but either don't have funds to spare or have something else ahead of it if I do. If I see another one I'll send you a heads up, though.

 

Speaking of rarities, a TI 955 workstation just popped up you-know-where. I don't see anything particularly remarkable about what looks like a bog-standard Taiwanese AT clone with a TI badge, but there it is. Not a TIPC,(now that would be interesting!) Collector material, though, maybe, or retrobattlestation.

 

If I had the $200 to spare the 955 seller wants, though, it's more likely I'd see if that offer I got from the seller of the ASC Positioning Arm Disc Controller manual might still be good. That badly needs to be archived. Unless there's a copy in the DeGolyer or other library, it may be the only one extant. The PAD  (IBM 3330 type disks) subsystem was only used on a couple of the ASCs, I think, but anyone attempting an ASC emulation would want to emulate it, too. And it would probably be helpful covering the gaps in what 980 documentation is available.

Edited June 2, 2021 by jbdigriz

[jbdigriz]

Ok, who all did I miss that's done a 9995 project that they've published and I failed to mention in my OP? It's so hard to keep track of all the development that's gone on over the years in the community. Stack overflow time. ? So, let's make a list?

[+dhe]

jb,

 I purchased one of these units. It's as close as I think I will get to a working University Board.

 FWIW - The creator offers for an additional $45 - the LCD screen, Transformer and Serial Cable.

 I have a fetish about keeping working cables with units - whether they be video/usb/serial - a known good labeled cable is to me worth a box of - "one of these should work" ?

 Much appreciation to you for pointing these out.

 

[jbdigriz]

3 hours ago, dhe said:

jb,

 I purchased one of these units. It's as close as I think I will get to a working University Board.

 FWIW - The creator offers for an additional $45 - the LCD screen, Transformer and Serial Cable.

 I have a fetish about keeping working cables with units - whether they be video/usb/serial - a known good labeled cable is to me worth a box of - "one of these should work" ?

 Much appreciation to you for pointing these out.

 

Cool, let us know how it turns out. Might be a good review post in it.

[+dhe]

Received today from Thailand. Construction looks good.

[+FarmerPotato]

Sweet.

[+dhe]

I can't help but compare, learning 9995 assembler on this board vs a minimem.

 

[+FarmerPotato]

On 6/2/2021 at 8:02 AM, jbdigriz said:

Speaking of rarities, a TI 955 workstation just popped up you-know-where. I don't see anything particularly remarkable about what looks like a bog-standard Taiwanese AT clone with a TI badge, but there it is. Not a TIPC,(now that would be interesting!) Collector material, though, maybe, or retrobattlestation.

 

If I had the $200 to spare the 955 seller wants, though, it's more likely I'd see if that offer I got from the seller of the ASC Positioning Arm Disc Controller manual might still be good. 

 

I stumbled on the TI 955 over-there today. At first I was intrigued, until I began to figure out the only big chips visible were EGA and some BIOS EPROM. The hint about 8 or 12 MHz startup speed reminded me of the clone era.

 

But it is local pickup for me...

 

The photos are also at https://wiki.preterhuman.net/Texas_Instruments_955_Workstation

 

I also stumbled on the ASC PAD manual, same seller as a $100 990 Pascal manual. I do not know if the PAD manual is in DeGolyer but I will be sure to check if/when I go there again.  I only read the one book about "History of TI's early computers" covering the ASC.

 

 

[jbdigriz]

On 8/30/2021 at 9:27 PM, FarmerPotato said:

 

I stumbled on the TI 955 over-there today. At first I was intrigued, until I began to figure out the only big chips visible were EGA and some BIOS EPROM. The hint about 8 or 12 MHz startup speed reminded me of the clone era.

 

But it is local pickup for me...

 

The photos are also at https://wiki.preterhuman.net/Texas_Instruments_955_Workstation

 

I also stumbled on the ASC PAD manual, same seller as a $100 990 Pascal manual. I do not know if the PAD manual is in DeGolyer but I will be sure to check if/when I go there again.  I only read the one book about "History of TI's early computers" covering the ASC.

 

 

The 955 has been up for a while; maybe they'd take less?

 

The PAD manual is crying out to be archived. Hopefully I'll be able to up the offer I made a while back that was ignored, or someone else will purchase and archive it.

 

Another application of 980s as storage controllers here, btw:  https://www.computer-history.info/Page4.dir/pages/Storage.dir/

 

 

[speccery]

This is off-topic, but I hope you don't mind. The same guy making the 9995 microprocessor kit also makes another one using the TMS320C25 DSP. I built a PC plugin board about 30 years go probably in the beginning of 1993, since I wanted to learn to program the TMS320C25. I also wrote an assembler for it (in Turbo C), and got it calculating Mandelbrot sets pretty fast. The GAL chips used by this project were programmed using a programmer I also built myself, based on a magazine article. 

This actually was my second TMS320C25 DSP project, I build first a standalone board. It no longer exists, I only had this one DSP chip and reused the components.

I was studying at the time. Telling about this hobby project landed me my first "real" job in the high-tech industry. Quite a lot of wiring this was one was... No cheap Chinese PCBs then.

[jbdigriz]

On 8/31/2021 at 11:48 PM, jbdigriz said:

The PAD manual is crying out to be archived. Hopefully I'll be able to up the offer I made a while back that was ignored, or someone else will purchase and archive it.

 

Forgot to mention the seller very graciously lowered the price a some weeks ago, so I snagged this. In the process of scanning it (and the 990/1 manuals I mentioned some time ago), but it's taking some time as I haven't figured out acceptable stitching and post-processing of the multiple scan passes of the many foldout schematics. May have to break down and get a wide-format scanner.

Edited October 11, 2021 by jbdigriz

[jbdigriz]

On 8/30/2021 at 9:27 PM, FarmerPotato said:

 

I stumbled on the TI 955 over-there today. At first I was intrigued, until I began to figure out the only big chips visible were EGA and some BIOS EPROM. The hint about 8 or 12 MHz startup speed reminded me of the clone era.

 

 

The 955 is listed in some places as a terminal. Unless there is an emulator in ROM on the board, since it looks like a non-HD box, I would assume the emulator(s) were supplied on floppy disk. Probably the same as the PC versions TI supplied.

[+FarmerPotato]

@speccery I like the wire wrap.

 

In 1989 I only had one color, too: pink! Wire-wrapped my own 9995 board.

maybe like you, I wrote to TI asking about the TMS340, but they only sent back glossy brochures. Did you get free sample chips? 
 

 

My 9995 board didn’t get me a job, but I did win a memorable trip from city-level  to state science fair, exhibiting “Udo: the 16-bit Microcomputer”. 

 

looks like I wrote a TLDR; here, feel free to not indulge my digression. 

 

TLDR;

My dad worked for TI and introduced me to several engineers. 

 

For the 9995, I robbed an irreplaceable Mike Read one-off bar code scanner. (TI engineer who went also worked on 9900-based HARM missile.) 

 

Wish I had that barcode reader now, cuz I just found hundreds of speech bar codes for it, buried in my dad’s basement.

 

Yamaha sent me a 9938 chip sample, unaware that it was just a teenager who wrote asking about it. 
 

For the 9938, I had, and never returned the 21.77 MHz crystal, which came out of Nick Hulbert’s HAM gear. (deceased TI engineer. Author of a TI technical book?) 

 

Nick also lent me his EPROM programmer for burning 2532s on the 4A. He kept my Apple II floppies, but still hardly a fair trade. (I feel weird that he passed away while I had his stuff.)

 

 

Everyone I met was so generous in time, knowledge, and parts.

 

Can’t remember the name of the engineer who loaned the wire wrap gun, fixed up the 9938, guided me in how to draw a proper schematic, and taught me to use his oscilloscope. (Which scope I had to return before the state science fair, alas.) 

 

To use the shrink-DIP 9938, he proposed a machine pin 64-pin socket a la  9900, a chunk of wood, and very short wire wrap from the 9938 soldered to each of the socket pins. That took hours to make. 
 

I never understood video out. I connected composite out, directly to a monitor. No way could that ever sync up. 

9995 from the bar code reader: Better than stealing the 9995 from a 99/8, made available by another TI engineer. Now I know that the >F000 RAM is disabled on the 9995 from the 99/8. Drove me crazy trying to start it up until I tried the “normal” 9995 from the bar code reader. 

 

still have what’s left of my board, esp the 9938 adapter monster, and the 2532 EPROMs. Alas, I later started to tear it down, to start an all new one, so it can’t run. 
 

Yours looks complete and very tidy. 
 

 

 

 

 

 

 

 

 

 

 

[speccery]

On 10/11/2021 at 9:48 PM, FarmerPotato said:

@speccery I like the wire wrap.

 

In 1989 I only had one color, too: pink! Wire-wrapped my own 9995 board.

maybe like you, I wrote to TI asking about the TMS340, but they only sent back glossy brochures. Did you get free sample chips? 
 

 

My 9995 board didn’t get me a job, but I did win a memorable trip from city-level  to state science fair, exhibiting “Udo: the 16-bit Microcomputer”. 

 

It's actually not a wire wrap - I did use wire wrap wire, but soldered the ends, no wire wrapping ?

 

Thanks for sharing your memories, that was an interesting read! I have actually never gotten free chips - I guess I did not know better to ask for them in a good way at the time. It sounds your dad has really brought you good connections at the time

 

You've really had access to rare chips, as you had the TMS9995 version for the TI-99/8, without the RAM! I find it interesting they did not go with the regular TMS9995, as that would have been faster with it's internal 16-bit wide RAM. They really have held compatibility at a very high value.

 

I just remembered my own TMS9995 project from 2016, you might find it interesting: GitHub and some pictures here. In this project I use a Xilinx FPGA with TMS9995, by dropping the operating voltage of the TMS9995 to 3.7 volts, thus avoiding level converters ?

[+mizapf]

3 hours ago, speccery said:

You've really had access to rare chips, as you had the TMS9995 version for the TI-99/8, without the RAM! I find it interesting they did not go with the regular TMS9995, as that would have been faster with it's internal 16-bit wide RAM. They really have held compatibility at a very high value.

I imagine they considered it difficult to have a mapper for the normal RAM on one side, and the on-chip RAM shadowing the normal RAM on the other side. We know these effects from the Geneve.

[speccery]

9 hours ago, mizapf said:

I imagine they considered it difficult to have a mapper for the normal RAM on one side, and the on-chip RAM shadowing the normal RAM on the other side. We know these effects from the Geneve.

Yes, good point, I have never used a Geneve so keep forgetting it has the memory paging as standard.

[+mizapf]

And on the Geneve, the issue is almost negligible. You know which page you map into E000-FFFF, and that the on-chip space will spoil the contents of those addresses in the RAM page.

 

(Apart from the fact that the mapper of the 99/8 is way too complicated, compared to the Geneve mapper.)

[Tom Frerichs]

On 6/2/2021 at 5:41 AM, dhe said:

Thanks for sharing. Back when I was in college in my first EE class, we did assembler on a similar one board computer with a 8080.

 

I then later leaned of TI's "University Board" - I've tried for 30-40 years to find one, I did about 10 years ago find one, that I could afford for about $200, when I got it, it looked like someone had stored in an open shed, next to the ocean, it was eaten up with corrosion. I ended up giving it away to someone that wanted to try and restore it, and never heard back from him again.

 

So this is a really nice find.

 

Thanks,

So, this is what you were looking for?  

 

I bought this new in the late 70s, early 80s--don't remember when exactly.  Had to build a power supply, but once I did, I powered up the board and remembered how miserably hard it is to read alpha characters on basically a 7-segment display.  Now I've got to rebuild the 2K static RAM expansion. And I need to find the correct MOLEX connector for the tape interface.

I'm looking forward to playing with the board again, although I guarantee the I'm going to use the serial interface (with the approved 1200 baud modification) to write programs in. That TI calculator keyboard and display are just TOO hard to use.

 

Tom