Powertran Cortex

[tms9995]

So Tony, can you give us a little history about the Cortex as a project? It's a pretty serious piece of kit for a big company to put all that effort into just for a magazine project. It surely had to be an abandoned commercial home/business computer product? We know about the TI99/8 that was dropped at the last minute, was this something similar? Do tell

 

[Tony Rowell]

how the cortex came to be...

how the cortex came to be...

The cortex came about as an extension of a group of us at TI trying to build ourselves a home computer out of "scrap", remember that at the time a home computer was somewhere between non-existant and insanely expensive. We started of by taking faulty viewdata boards (sort of teletext over a dial-up modem) based on the TMS9980. When the TMS9995 came out we started looking at building something around that. At that time TI had a minicomputer division (990/12 etc) and a home computer division (99/4A) with lots of internal politics to ensure that the home computer didn't tread on the minicomputer dept toes so to speak, we however worked for the semiconductor division an just wanted to build something and management just wanted to sell chips. The semiconductor division building a home computer was a big no-no, it would have been stomped on by both the home computer division & the minicomputer division so we had unofficial encouragement to go ahead with the design and the way to "get it out" was ETI (as long as there was no direct connection to TI), Powertran simply made the kits & sold it.

Jim Gill and myself did the BASIC (based on one from TI but mostly rewritten), Colin Hinson did the debug monitor, Neil Quarmby did the schematics and floppy boot code and Ian White did hardware prototyping and "resourcing".

Originally the computer was named "Synapse" and the first prototype photos showed that name but it was renamed at the last moment because of a conflict with another product with that name.

 

[tms9995]

Some fantastic information there Tony - thanks! It must have been very cool to have all the resources available in-house and just be let go to do your own thing. Talk about kids in a candy store!

 

[Tony Rowell]

This is the schematic of the ebus arbiter from the design manual, probably the LS2001 was intended to have this inside. Should be doable with a simple PAL these days.

For those people looking at the memory mapper chip I seem to remember that we could use eithe a LS610 or a LS612 and that one of those was used in old XT class PCs so if you can track down an old old PC motherboard you might be able to reclaim the chip. The memory mapper remapped the top 4 address lines from the processor, thats why there are 4 links to bypass the mapper if it isn't fitted.

 

[Tony Rowell]

resources were available but not officially, lots of favours and being resourceful with all the work being done out of hours but it was a lot of fun too.

 

[tms9995]

So once it was out the door to ETI/Powertran was that it or did you remain involved with support/further development? I only know Neil was involved later on because he was selling CDOS for about 50 pounds per copy in 1985. Now that makes sense, since you said he was the one who did the floppy code! When I was scanning the ROM listing, I noticed that the comments for the floppy boot are in lower-case, whereas the rest of it is in upper-case. I guess he had his own style...

 

[Tony Rowell]

I wasn't with TI shortly after the magazine article came out although I did go back to work for them years later. Part of the deal with the cortex was that we got one of the kits each as "payment" and it my macine and Jim Gill's that I have in my loft. None of the group are still with TI but with the exception of Ian White I know where they still are so fortunately still accessible for anything I've forgotten.

 

[tms9995]

I guess it's only fair to ask, but as I plan on making my Cortex Emulator available to anyone on here that may be interested (like 3 people!), I will obviously be posting the ROM image which I guess is your work! Do you have any objections? Also, I would like to put up CDOS. Do you think Neil would mind? (Perhaps you could ask him). I would feel better knowing. Thanks.

 

[Tony Rowell]

I don't think anyone would care, if push came to shove then TI would probably claim that they owned everything (naturally) but this is all so far in the past and of no commercial interest to anyone.

 

[tms9995]

Thanks. Like I said, I feel better for asking! If it was my stuff, I'd be honoured 26 years later!

I don't suppose you have the ROM source on disk anywhere? As I've said, I have the SDSMAC listing output, but would love to be able to rebuild the ROM from source files.

 

[Tony Rowell]

Maybe processing the listing will give you a machine readable copy of the source code. As a matter of interest how did you come by the listing?

 

[tms9995]

Yeah I guess the next step would be to try and put it through OCR to try and extract the text.

In 1994/5 I had built a SCSI interface for the Cortex and had more or less finished what I called HDOS. For the life of me I can't remember what for but I needed some info so I went through all the User Group letters I had (which was extinct by about 6 years plus!) and I tried contacting some of the people involved. I ended up meeting with one of the guys who had run the Group and he said if I gave him the hard drive/SCSI design and driver code, I could have everything he still had. One of those things was the listing! I don't know how the User Group got it though. I also got some extra manuals for the MDEX O/S plus a couple of other compilers & tools for MDEX. There was a lot of passing stuff along as each of the bigger companies involved no longer wanted to support the machine any longer. I always hung on to everything I got as I wanted to get back into it 'one day'...

 

[Ksarul]

I do have a bunch of the 74LS612 Memory Mappers, so that chip isn't a problem either--the problem was with the pinout of the 74LS2001, since that is the one chip I don't have sufficient info on. Many thanks for the extract of the bus arbiter logic.

Seeing all of the information that is actually out there for these has me in a major state of anticipation to read all of it! I plan to put direct scans of all of the ETI articles I have original copies of into the Cortex folder on WHTech this weekend, including a single high res scan of the component side of the board instead of in halves like the other copies I've seen.

One interesting question, the component mask numbering on mine bears absolutely no resemblance to any version described so far--almost everything is different. . .but the component positioning is the same. Any idea how that one came to be, and are there schematics that reflect this other number scheme (note it isn't the minor changes noted in either the ETI articles or the Powertran construction manual that reprinted them with some corrections)?

Once again, many thanks!

 

[Tony Rowell]

tms9995 - was the info in my PM of use?

 

[tms9995]

I didn't know I had a PM. I'll check it out - thanks

 

[tms9995]

I do have a bunch of the 74LS612 Memory Mappers, so that chip isn't a problem either--the problem was with the pinout of the 74LS2001, since that is the one chip I don't have sufficient info on. Many thanks for the extract of the bus arbiter logic.

Seeing all of the information that is actually out there for these has me in a major state of anticipation to read all of it! I plan to put direct scans of all of the ETI articles I have original copies of into the Cortex folder on WHTech this weekend, including a single high res scan of the component side of the board instead of in halves like the other copies I've seen.

One interesting question, the component mask numbering on mine bears absolutely no resemblance to any version described so far--almost everything is different. . .but the component positioning is the same. Any idea how that one came to be, and are there schematics that reflect this other number scheme (note it isn't the minor changes noted in either the ETI articles or the Powertran construction manual that reprinted them with some corrections)?

Once again, many thanks!

If you get the 31 page download from old-computers.com, there is an article called "Parallel I/O for the Cortex" (page Cortex_20.jpg). That shows the plug-in header for the 74LS2001. It enables the E-Bus hardware on the Cortex PCB to work.

As far as the component numbering, I believe I recall that when the kits came out, there was kind of a Rev B to the PCB artwork and hence it differed from the original ETI articles.

I have finished scanning the bulk of my documentation and have PM'd Ksarul for his help in getting it up on the FTP site. The Emulator is good enough for sharing too.

 

[Ksarul]

I just finished a major documentation upload to WHTech at:

ftp://ftp.whtech.com/Powertran Cortex/

Many thanks to TMS9995 for scanning all of this stuff in and making it available to the rest of us.

 

[tms9995]

Wow, the cover of the first ETI brings back some memories. I was in the computer room at school and my friend came in (he was an ETI subscriber, whereas I had the sister publication Hobby Electronics) and he said, "This is it!" and threw down the magazine in front of me. After he explained all it's attributes, I was sold. Three of us ended up with one. My second machine is the one I bought from my friend when he went off to university.

 

[tms9995]

74ls2001

74ls2001

I think what is interesting is that if you zoom in on the picture of the ETI cover, there appears to be a chip fitted in the location of the 74LS2001! Despite what Tony said and from my experience that it never existed, something obviously did!

 

[Ksarul]

I've got the complete internal logic for the 74LS2001--just not a bullet-proof identification of which pin each of those signals need to reside on. Some of it isn't used on the Cortex, which is why I was hoping someone had the signal listing for what belonged on which pin on the chip. With the signal name to actually designated pin position, I can design and program a true replacement for it. . .and then it will be around again. As a side note, there appear to be about 300 of them somewhere in Poland, assuming that they really are the same IC. . .I haven't gotten a response yet from the company that says they have them though, so they may be vaporware.

I'm glad you liked the inclusion of the cover. I have the Aug 1984 issue as well, I just need to finish the scans for it. It contains the rest of the Jun 1984 article along with the board layout for the parallel board.