Talk:LM3000
From BenningtonWiki
Wow getting this machine working was a real accomplishment, I worked on a homebuilt processor for a couple of years in the mid nineties so I'm one of the few who have some idea how difficult it must have been. I never got mine working but I was working by myself, and in the end simply didn’t have enough time to finish everything. Ok it was a really different design because I was working backwards. I had a complete ALU and register unit, the main data bus interface and about half the control unit. The most difficult part of the design for me was the control sequencer, and I also had a few problems with the instruction set. But that’s mainly because I built it to test new instruction set ideas - so the whole thing was pretty experimental.
Ok a few specs - 16 bit data and address bus, Complex ALU that could add, and subtract, and multiply and divide - using a simple barrel shifter. Originally it was going to use simple pipelining so it could run semi-asynchronously - so the machine could do other things while multiplying (very very slow), but this wasn’t practical. It had about 8 registers and used 1 as the program counter. Its instruction set decoded directly through 3 or 4, 4 bit demultiplexer chips, with a set of safety interlocks Physically the whole thing was wired on four strip boards.- I used chip carriers and cut all the stips, then used a special prototyping wire to put everything together. This left an amazingly small finished board but was also amazingly fiddly to do, you burned your fingers constantly - and if you made a mistake could be a nightmare. If I was doing it now I would have drawn all the designs out carefully in reverse first. Mine might sound a lot more complex but I tinkered around with it for over two years, and in the end I never did quite finish it. Sadly I’ve lost everything now but seeing your project I might just try again sometime.
The most important thing though is that it taught me more about computing than almost anything else I’d ever done, and the knowledge helped me a great amount when I worked on a whole series of theoretical computer projects. What I am saying is that this is a really great way to teach people about the heart of computing - I think it should be promoted for all computing, IT students.
- Sorry for being so long Robert Lucien 217.42.232.208 01:19, 2 February 2007 (EST)
