I was quite surprised during my time at Sun Microsystems that the address and data buses on our Sparc Systems were not terminated. In fact there was no consideration at all given to them being transmission lines, even though they were running at a few GHz. The systems worked very well, with no problems. These days high-speed buses in computers are serial. It turns out that running a few hundred parallel lines that are the same electrical length (and thus have the same propagation timing) is a very, very difficult problem to solve. Much easier to have just a few serial lanes that carry the data at super-high speeds.
Sometimes we can overthink things when simple is what is really needed. Wayne did a really good job with the K2… 73, Jack, W6FB > On Jul 31, 2025, at 2:47 PM, Fred Jensen via Elecraft > <elecraft@mailman.qth.net> wrote: > > "It is amazing how we were able to build radios with only 2 sided PCBs for 50 > ohm micro or strip lines." > > In the later 70's we took delivery of a Data General Nova 3 mini-computer > that came with a FORTRAN compiler. The circuit boards were 15" square-ish, > 2-layer. We quickly discovered that periodically but unpredictably, > execution of the compiler would quit dead on a DO statement in the source. > It just stopped. No other statement, only a DO and not all the time, but if > it stopped, it did so on every recompile of that source. > <mailto:elecraft@mailman.qth.net> > > After several DG visits they finally believed us and showed up with their > Guru in flip-flops, ragged board shorts, a torn T-shirt, and fairly long-ish > uncombed hair. He said almost nothing and the other 3 DG's [in "computer > expert" business attire] were very attentive to him. > <mailto:elecraft@mailman.qth.net> > > He spent most of the morning with the 'scope making notes in a beat-up > notebook in indecipherable script and diagrams. Finally muttered something > to his minions who raced to get him an Xacto knife, some very light insulated > wire, and the soldering station. He cut both ends of several parallel traces > and replaced them with twisted pair jumpers. The compiler never stalled on a > DO statement again. > > Problem likely would never have occurred had the board had a 3rd ground layer > between the two outer ones but they they had yet to be invented. > > 73, > > Fred ["Skip"] K6DGW > Sparks NV DM09dn > Washoe County > > > > -- > This email has been checked for viruses by AVG antivirus software. > www.avg.com > ______________________________________________________________ > Elecraft mailing list > Home: http://mailman.qth.net/mailman/listinfo/elecraft > Help: http://mailman.qth.net/mmfaq.htm > Post: mailto:Elecraft@mailman.qth.net > > This list hosted by: http://www.qsl.net > Please help support this email list: http://www.qsl.net/donate.html > Message delivered to jackbrin...@me.com ______________________________________________________________ Elecraft mailing list Home: http://mailman.qth.net/mailman/listinfo/elecraft Help: http://mailman.qth.net/mmfaq.htm Post: mailto:Elecraft@mailman.qth.net This list hosted by: http://www.qsl.net Please help support this email list: http://www.qsl.net/donate.html Message delivered to arch...@mail-archive.com
