Page 2 of 2

Re: Learning CPLD and testing GALs

Posted: 21 May 2020 23:58
by tzok
PhilC wrote: 21 May 2020 21:29So I've got WinCPLD loaded up on my Windoze laptop and am going to learn CPLD programming.
If you've meant WinCUPL, then throw it away. It is horrible. If you want to use more modern CPLDs, then go for Quartus, if you need GALs, then I recommend you to use Opal Jr. It is DOS based, but runs fine under DOS-Box, and is much more stable and reliable than WinCUPL. In WinCUPL you can never know if you've got a faulty chip, made a mistake in code, or is it just some memory leak in the WinCUPL... you could write a code, compile it, then close WinCUPL, reopen it, compile again, and you will end up with two different JEDEC files.

Re: Learning CPLD and testing GALs

Posted: 22 May 2020 08:29
by JezC
exxos wrote: 21 May 2020 23:01 I cannot remember if it was tested, but any drive should work on the H4 without needing to change the jumper...
Ok, great - if it's not already tested then I will check it as soon as I can finish off the first H4. That will probably only be in DD mode rather than HD mode... :roll:

Re: Learning CPLD and testing GALs

Posted: 22 May 2020 10:58
by PhilC
tzok wrote: 21 May 2020 23:58
PhilC wrote: 21 May 2020 21:29So I've got WinCPLD loaded up on my Windoze laptop and am going to learn CPLD programming.
If you've meant WinCUPL, then throw it away. It is horrible. If you want to use more modern CPLDs, then go for Quartus, if you need GALs, then I recommend you to use Opal Jr. It is DOS based, but runs fine under DOS-Box, and is much more stable and reliable than WinCUPL. In WinCUPL you can never know if you've got a faulty chip, made a mistake in code, or is it just some memory leak in the WinCUPL... you could write a code, compile it, then close WinCUPL, reopen it, compile again, and you will end up with two different JEDEC files.
Lol, yes thats what I meant. I'll have a look at the Opal program over the weekend.

Thanks

Re: Learning CPLD and testing GALs

Posted: 22 May 2020 11:20
by exxos
You don't need to open and close CUPL to get different jed files, every recompile you will get a different JED because of the time stamp in the file..

Capture.JPG

Re: Learning CPLD and testing GALs

Posted: 22 May 2020 11:35
by czietz
BTW: It seems like the most modern software to still support GALs (Lattice ispLEVER Classic) has paid-for licenses (590 USD per year) only. Didn't this use to be free-of-charge? Does Lattice want to "milk" the customers stuck on legacy devices? This also makes their 5V-tolerant CPLDs (mach 4A) very unattractive.

EDIT: The Internet Archive shows that I did not remember it wrong: Until at least end of 2019, the license for ispLEVER Classic was free. Now the link to request a free license has vanished from the Lattice website and has been replaced by a link to their online store.

Re: Learning CPLD and testing GALs

Posted: 22 May 2020 21:47
by tzok
I didn't mean the time stamps, but completely different fuse map. If you mess too much with the code in editor, and compile just after editing, it is likely, that compiler will output garbage.

Another advantage of Opal Jr is, that it allows to convert JED back to EQN (decompile).

Re: Learning CPLD and testing GALs

Posted: 22 May 2020 21:56
by exxos
I've not really had any issues with it. Just it doesn't create a jed (and gives no warnings) if you do not set a clock on a flipflop for example. If you move code about it will create a different fuse map as well.