Xref: utzoo unix-pc.general:7227 comp.sys.att:11477 Path: utzoo!utgpu!watserv1!watmath!uunet!zaphod.mps.ohio-state.edu!usc!apple!portal!cup.portal.com!thad From: thad@cup.portal.com (Thad P Floryan) Newsgroups: unix-pc.general,comp.sys.att Subject: Re: microwaved people (was Brownouts, shorts, ...) Message-ID: <37925@cup.portal.com> Date: 12 Jan 91 14:12:27 GMT References: <37881@cup.portal.com> <1761@mitisft.Convergent.COM> Organization: The Portal System (TM) Lines: 48 dold@mitisft.Convergent.COM (Clarence Dold) in <1761@mitisft.Convergent.COM> writes: [...] When the radar was dismantled in 1976, we found flies inside the waveguide. They were dead, but I imagine they starved. They didn't look burned, even after 20 years of fairly constant high level radar exposure. [...] Flies are smaller than 1/4 or 1/2 wavelength at "most" the frequencies at which high power RADAR operates, so would not have been "cooked" resonantly. If you think back of that chain-link fence which "sang" when hit by RADAR, think of the length of each side of each "square" formed by the wire and you can deduce the frequency (or harmonic thereof) at which the RADAR operated. My reference to human eyeballs and testicles was based on an approx. 1GHz frequency whose wavelenth "likes" objects that size. Reason I bring up the above points is that once I had a "problem" with a computer I designed in that one trace on the motherboard tended to resonate and caused all sorts of weird problems that didn't occur on the prototype which I hand-made; the guy who did the PC layout for Teflon-fiberglass simply wasn't familiar with high-frequency considerations for layout. Computers using ECL or GaAs components are also susceptible to poor layout at the higher frequencies. I'll have to show-'n'-tell (via photos) some of my previous designs at even higher freqs; the cases, traces, connectors, etc were all gold with ceramic, sapphire or even diamond substrates, and, except for the occasional chip capacitor, one wouldn't recognize ANY component on the microstrip "boards". Designing with clocks far in excess of 1GHz often required empirical "tricks" that mimic black magic and voodoo! And the completed circuitry was nicknamed "Tenement Housing" (picture that in your mind). When people "boast" of 50 MHz-clocked computers today, I still think of that as practically DC. My lab operated with frequencies from "DC to daylight", and the designator "THz" wasn't that uncommon. And a short anecdote in closing: back when I was doing all this stuff, the term "GHz" didn't even exist, we used "GC" or "GCPS" (Giga Cycles [Per Second]) . When the standards committee wanted to honor Hertz with "Hz", we all protested with a petition claiming they should instead honor Charles Proteus Steinmetz and we'd then be able to continue using "CPS"! :-) Thad Floryan [ thad@cup.portal.com ]