Robert G. Brown
rgb at phy.duke.edu
Tue Jan 21 16:21:34 PST 2003
On Tue, 21 Jan 2003, Jack Wathey wrote:
> Aside from the risk of fire from a non-OTS beowulf enclosure, are there
> any risks to life and property if the thing leaks RF as discussed above?
> Can a shelf of radiating motherboards slowly cook (like a microwave oven)
> the poor nerd standing in front of it? Can it interfere with
> communication/navigation equipment in low-flying aircraft? My hope and
> expectation is that the power of such RF signals is MUCH too low to do
> such damage, but it would be comforting to hear an authoritative answer
> from someone who knows more physics than I.
I would say no, no, no. And no.
Remember, the TOTAL power going into the thing (per CPU) is order of 100
Watts these days. Most of that power is dissipated making flips flop on
the many itty bitty switches, and in making disks dance this way and
that, and in creating signals that run down this line or that.
Engineers view power going anywhere that they don't need it to go
(correctly!) as LOSS. So radiated power is bad, not just because it can
cause all sorts of malfunctions by generating voltages where you don't
want them but because they have to replace that power and run everything
hotter as a consequence.
I'd guess that total radiated power even from an unshielded unit is less
than a watt (possibly much less) and radiates out in all sorts of
directions and in frequencies that aren't absorbed "suddenly" in human
tissue. That may be plenty for a radio to pick up, might even be enough
for a nearby circuit to resonate to, but will heat your nearby hand up
far less than the warm air convecting off of the hot surfaces, far less
than the blackbody energy radiating from the hot surfaces, and even then
far less than your body can easily dispose of.
You're also well below the frequencies and intensities where quantum
transitions (of, say, a biological molecule) were at all likely. So
neither average power (heating of tissue) nor quantum transition
(resonantly breaking a key molecular bond in e.g. DNA) is likely to be
SO I wouldn't worry. I don't worry much about cell phones, either.
Microwave ovens, I worry about. TV sets and CRT's in general I worry
about (they produce soft X rays and other ionizing radiation with some
probability, especially close to the tube).
> Along those same lines... Is there a handy-dandy (and not too expensive)
> gadget for measuring RF emissions from an electonics enclosure in a
> standard, quantitative way? I'm hoping for something that can tell me
> whether or not a prototype or DIY enclosure is sufficiently RF-tight that
> it would pass the official certification tests. Are those tests simple,
> or way to complex and expensive for an amateur to duplicate?
I'll let some others answer that. A Famous List Person and I were
kicking around this question as a possible interesting topic for EE or
Computer engineer student types. I don't think that they equipment is
particularly cheap or DIY, though...
Robert G. Brown http://www.phy.duke.edu/~rgb/
Duke University Dept. of Physics, Box 90305
Durham, N.C. 27708-0305
Phone: 1-919-660-2567 Fax: 919-660-2525 email:rgb at phy.duke.edu
More information about the Beowulf