ABEL support for legacy chips

That's because no one makes the chips anymore.....

That said, I see you CAN still actually buy the PLS100, and I also see my PLD programmer still shows a PLS100 on the menu...

Nope - Xilinx's response will be to tell you to take your ABEL source, and recompile for a CoolRunner CPLD. ( but that does not retrofit too easily into a PLS100 socket.)

The Xilinx ABEL has a certain amount of inbuilt ability, and it looks like, if you really want to, you can get PLA and BLIF files from the current Xilinx ABEL flows.

What you then need is the relevent Signetics PLS100 fitter, or you could look for the Signetics software - IIRC their SW called Amaze supported the PLS100, and it had a successor called SLICE, but both are ancient.

Or, you could look for a device similar to the PLS100, but still in production, such as the ICT PA7536 or PA7540 ?

ICT offer current software to support these devices. See

The 22V10 is still quite active (Atmel, ICT, Lattice), but has fewer total pins than a PLS100.

-jg

I would design a small PCB for the new CPLD that would plug in to the old device socket.

Leon

Thanks for answering. The 82S100 is now in production reintroduced to the marked by QP Semi. Anyway, this is not an issue for me as my organisation has 82S100's in-house in adequate numbers. My project consist of replacing an older version of the programmed chip with a newer one on a limited number of circuit cards. I have already considered the PA7536, but I find it way easyer to get an older version of ABEL, compile the code for 82S100 and burn the chips.

Eirik

Wow, a PLS100 data sheet with June 2004 on it :) Google did find what looks like (most of?) Amaze, here

- in case you have problems finding an old ABEL with the 82S100/PLS100. You could also try direct fuse editing :)

-jg

Eirik,

Have you located it? I'm not familiar with the 82S100. I have an ancient version of ABEL, circa 1988. Is that too ancient?

Mike

Not sure. I've found Abel 2.0 and Abel 4.0 on the net and tried to compile the source, but in both versions I got the message "fatal error: out of memory". It worked when I tried to compile only parts of the source.

Eirik

Thanks, but I've found and downloaded it when you gave me the name. Haven't tried it though, as disk 2 was missing and I later on found Abel on a web site.

Eirik

Some (most?) ABEL's needed keys, this may be a security artifact ? It is hard to believe the 82S100, which my info shows has just 1928 fuses (smaller than a 16V8), would tax memory ? CUPL probably also supported the 82S100.

-jg

Could be, but I suppose it would come up with an error message asking for valid keys or something like that. When I cutted down the source it would compile. I haven't really set up an old dos environment yet so I think I will try this as soon as I get the time to do it. AFAIK CUPL is a different HDL than Abel, which means I have to rewrite the source to CUPL syntax. My HDL programming experience is somewhat limited. I've written some PLD-cicuits in PALASM and another GAL-assembler called GALASM. Very basic, no high-level coding. Although I probably understand most of the code I'm really not so keen on translating it. The equipment using the FPLA also incorporates a serious amout of other PLD devices, mostly 22V10's and EP600's, all documented with Abel source code.

Eirik

Sounds close - at this point I would try something trivial like

8 NAND gates in a test file ( uses every pin ), and confirm you get what looks like a valid JED file.

source to

CUPL was suggested, should you hit a brick wall with ABEL, right now it sounds more like a speed-bump..... :)

-jg

Finally got the time to set up a box with DOS. Didn't help much though. I guess you were right about the need for a licence key. It seems like it worked for smaller designs, perhaps for evaluation purposes.

Eirik