It was July 1986 Engineer's Toolbox. (Thirty years ago now!) I don't want to publish it myself on-line for copyright reasons but I think you can download it here:
https://ia802709.us.archive.org/35/items/byte-magazine-1986-07/1986_07_BYTE_11-07_Engineers_Toolbox.pdf
I used to also have a linear circuit analysis program that ran on the BBC micro - it generated conductance matrices for linear, proportional to omega and proportional to 1/omega. It worked well except you couldn't have time delays (except as approximations) which made transmission lines difficult.
If you're interested in SPICE I highly recommend getting yourself a copy of Lawrence Nagel's original 1975 PhD thesis ERL-520. I ordered myself a paper copy from his University years ago (it cost something like $11) but I think the pdf is also available on line.
http://www.eecs.berkeley.edu/Pubs/TechRpts/1975/ERL-520.pdf
Thanks!
I've just read the BYTE article, and it was very interesting and educational. The details, length and quality of his article, really rub in what has been said, earlier in this thread. Which is that as well as becoming, rare (electronics/computer magazines), especially on paper. The technical detail/strength and quality, has been significantly watered down, over the years.
It is rather sad, really.
As I hinted, in my earlier post, the lack of floating point hardware, the relative slowness of the processor (6502), and the limited memory size, DID impact on his SPICE project. But on the other hand, he has done a marvelous job, of (highly) optimizing his program.
He went to amazing lengths, including writing some of it in machine code. Directly calling the floating point routines in software, to save time. All sorts of other optimizations.
Thanks for the PDF SPICE file, I have downloaded it, ready to read, later.
I think these days, the furthest a magazine could go, is to say "SPICE". Then they can expect huge complaints, because so many people have NOT heard of it, or don't know what it is.
Those early days of electronics and computing, were such FUN!
I'm feeling SAD now, thinking, what I am missing now.
In another part of that BYTE magazine, it mentions a HUGELY powerful, 65536? node, super-computer. Which could then exceed a billion instructions per second, and cost a huge amount of money. These days, a raspberry PI, is probably considerably more powerful than that machine. Which might have been one of the most powerful computers, available then.
Anyway, thanks again for the links!