Re: Manufacturing New 650x - Need help

Oneironaut · Post by **Oneironaut** » Wed Dec 21, 2016 7:09 pm

No doubt, you have already memorized these...

http://www.downloads.reactivemicro.com/ ... onics/CPU/

Brad

BigDumbDinosaur · Post by **BigDumbDinosaur** » Wed Dec 21, 2016 7:28 pm

Oneironaut wrote:

No doubt, you have already memorized these...

http://www.downloads.reactivemicro.com/ ... onics/CPU/

Brad

Note that the WDC datasheets listed there are woefully out of date.

Tor · Post by **Tor** » Thu Dec 22, 2016 8:16 am

But if it is to be exact you'll actually need the original masks.. can't see how you could get the exact side-effects (e.g. all the non-documented opcodes, which weren't designed in as such) otherwise.

Arlet · Post by **Arlet** » Thu Dec 22, 2016 9:56 am

Even the exact mask layout is no guarantee that you'll get the identical effects of non-documented opcodes, some of which depend on subtle analog effects inside the chip, which could be different on a 70's process compared to a modern one. That applies even more to the SID characteristics.

sark02 · Post by **sark02** » Thu Dec 22, 2016 11:37 pm

Giebels2609 wrote:

I researched a lot in the past days and figured, that it is at this point almost impossible to find equipment from the 70s and 80s...

It's a real disappointment to hear that. The project sounds fascinating, and I'm curious how you went about searching for the old integrated circuit fabrication equipment. It seems a shame to leave stones unturned, as it were. Can you tell us what sources of information you used to find out if there was any equipment out there? Maybe you missed something and/or maybe others can help.

Dr Jefyll · Post by **Dr Jefyll** » Fri Dec 30, 2016 6:07 pm

Speaking of producing vintage CPU's, I remember reading a related forum post some while ago. I don't think it's been mentioned in this thread.

Today I happened to find that post. On our sister forum, Anycpu.org, see Production restart of "old" CPUs. The article it refers to is dated November, 2014 -- not exactly current, but still interesting to read.

ETA: I just checked their web site, and the supplier (Rochester Electronics) claims that "More than 60 of the world's leading semiconductor manufacturers trust Rochester to extend the life of devices that they choose to no longer have in production."

-- Jeff

whartung · Post by **whartung** » Fri Dec 30, 2016 6:47 pm

It's an interesting phenomenon.

You would like to think that capability of creating a 40 year old chip, using a 40 year old process, would be straightforward and even cheap today.

But, instead, it may well just be impossible.

There was a story, somewhere, where I saw mention that we can no longer create the F1 motors of the Saturn V rocket, because the procedure and techniques of manufacture that were documented simply don't match what was really done. If you "follow the instructions", you won't get a viable F1 motor. And the people that actually did it, the ones with the institutional knowledge of how it was done, are scattered or dead today, with that folklore lost.

The early chip technology upon whom's shoulders the industry stood upon, may simply be gone today. Replaced with the newer, better tech. The amount of re-engineering involved to recreate it is simply not worth the time and money.

Good luck on your quest!

Tor · Post by **Tor** » Fri Dec 30, 2016 10:10 pm

Re. lost technology - somewhen between 1975 and 1979 (I don't remember the exact year anymore) I visited a research site which had a number of gigantic antennas used for atmospheric research, including northern light research. One was a rectangular mirror some 160 metres long, you could walk upright inside the feed. Everything was very big. Anyway, another antenna was used to literally heat the atmosphere. For that they needed an enormous Klystron vacuum tube. But the factory initially couldn't produce one that worked. They finally managed to do it by engaging the help of two very old, retired guys, and the research facility got their working tube, after a considerable delay.
Unless someone took careful notes during this process I suspect that doing it again must be impossible, I don't expect the old fellows to still be around by now. ..

BigDumbDinosaur · Post by **BigDumbDinosaur** » Fri Dec 30, 2016 10:52 pm

Tor wrote:

They finally managed to do it by engaging the help of two very old, retired guys...I don't expect the old fellows to still be around by now. ..

Hey! Watch it with that "old fellow" stuff!

Your anecdote reminds me of all the activity that was going on in 1999 as the year 2000 approached and everyone was fearful of computer crashes, elevator crashes, plane crashes, etc. One of my clients was still running a lot of COBOL software on their system and that code was full of two digit years. They had quite the time trying to round up some competent COBOL programmers to fix their code, as most of it was software inherited from the days of IBM System 360 dinosaurs running CICS. Most of the programmers who wrote that stuff had retired or passed away. They tried to get me involved, but I firmly resisted, especially since I hadn't done anything with COBOL since the early 1980s and never was into it that much. Eventually, they found a husband-and-wife team, both retired COBOL programmers and in their seventies, who had come out of retirement to cash in on the Year 2000 excitement. I do recall that a large amount of money changed hands to get everything up to snuff.

Tor · Post by **Tor** » Fri Dec 30, 2016 11:16 pm

Heh, I actually did a lot of y2k work myself (no COBOL though!), as did many programmers, and I still hear "what was all the fuzz about? Everything worked, didn't it?"
Yes, exactly.

BigDumbDinosaur · Post by **BigDumbDinosaur** » Fri Dec 30, 2016 11:54 pm

Tor wrote:

Heh, I actually did a lot of y2k work myself (no COBOL though!), as did many programmers, and I still hear "what was all the fuzz about? Everything worked, didn't it?"
Yes, exactly.

It was amazing just how much irrational conjecture was prompted by the Year 2000. Just to prove to my wife that the Year 2000 wasn't going to be an apocalypse I took her to the Sears Tower (now the Willis Tower) in Chicago on the afternoon of January 1, 2000 and we went up to the observation floor 103 stories up. The elevator worked fine.

Tor · Post by **Tor** » Sat Dec 31, 2016 12:21 am

Elevators don't need timers, at least not for going up and down.. and they have a completely mechanical failsafe that prevents them from falling down even if almost everything else fail. And I don't know what kind of y2k error would make planes fall out of the sky. So the run-for-the-hills hysteria was uncalled for. But there was still a huge amount of software which couldn't handle y2k, including banking and, in my area of work, satellite data processing (but we were used to it - one protocol used a 16-bit counter for satellite orbits, not accounting for the satellite lasting much longer than its expected lifetime, and orbiting more than 65535 times around the earth).
Many things would have failed after 1999, if it weren't for all the work done. And still some people don't see the connection between the focus on y2k and the fact that it all went well, instead concluding that the focus was uncalled for, because everything went well..

BigDumbDinosaur · Post by **BigDumbDinosaur** » Sat Dec 31, 2016 1:01 am

Tor wrote:

Elevators don't need timers, at least not for going up and down.. and they have a completely mechanical failsafe that prevents them from falling down even if almost everything else fail.

I know all that, although many elevator controls do have date/time reference functions for alerting building operators when maintenance is due or mandatory inspections are required. A friend, who is now retired, worked on elevators his entire career and said at the time that none of the elevators he worked on would quit operating due to any date references adversely affected by Y2K.

My wife didn't at the time understand all this technology mish-mash and for a while, bought into the hysteria over Y2K. I had a hard time explaining to her and others that almost all of the unresolved Y2K problems would be, at worst, annoying, not dangerous.

EugeneNine · Post by **EugeneNine** » Tue Jan 03, 2017 1:24 pm

I was on a forum at the time where a guy insisted his Mercedes would stop at Y2k because his radio display only had a two digit date.

Doesn't WDC own the 6502 IP so you would still have to license it from them?

GARTHWILSON · Post by **GARTHWILSON** » Tue Jan 03, 2017 6:04 pm

WDC didn't exist yet when the NMOS 6502 was designed (at MOS technology); but they do own the 65c02 IP.