|
Posted by Hairy One Kenobi on December 15, 2005, 1:27 pm
If you were Registered and logged in, you could reply and use other advanced thread options
>
<snip>
> > You'd be surprised.. after all, what's actually inside the box is the
> > same as the semi-shielded version (I say "semi", because EM shielding
> > has been a requirement for a not inconsiderable time). All that's
> > changed is the degree of shielding, and materials technology is
> > roughly constant throughout th universe.
>
> What could be more trivial than a hard-disk? An electric motor that
> spins an aluminum disk covered with rust particles, a pivoting-arm read-
> write head reminiscent of that on an ancient phono turntable, and ten
> cents worth of electronics. And yet, silly me, I choose to buy them
> rather than make my own. And so with Tempest equipment.
Hang on a minute.. as *I* remember, that's what I've ben saying all along.
And you've been saying that the specifications have some absurdly high
classification, and that the devices can't be reverse engineered..?
> > I'm not entirely sure who you're arguing with at this point. After
> > all, I've already pointed out that it's been the general rule to
> > shield buildings and particlar enclosures, rather than buy specialised
> > expensive kit. Maybe I should have put a timeframe in there? How about
> > "two decades" (true for the UK, no idea about the US).
>
>
> Not so much arguing as clarifying. I believe that Tempest is a very low
> risk threat for most people (in large part because much easier
> alternatives are usually available to our adversaries) and that scarce
> security money is usually best spent elsewhere. Specific situations may
> differ, of course.
>
> But for those who DO need Tempest equipment, rolling one's own is a
> clumsy makeshift and an invitation to disaster. I wouldn't do my own
> appendectomy either - even if I had read all about it, seen one done,
> and had a surgical-instrument catalogue. Nope, I would go to a specialist
> who does such things for a living and get it done right with all the
> followup support required!
See above.
> > Forgive me for saying, but that sounds like some reading from a set of
> > tables, rather than doing any calculations. You *do* know that dB is
> > an exponential unit, rather than linear? That each 3dB indicates a
> > /halving/ in signal level?
> >
> > Measure your signal drop-off by taking your distance and applying the
> > inverse-square rule (as defined by Newton in the seventeenth century),
> > then take the base-10 logarithm and multiply by -10. Simple.
> >
> > And, of course, largely irrelevant in the electronic Real World (which
> > tends to have variable permittivity, most of us not living in a
> > complete vacuum and all ;o)
>
> If you check I think you will find that my math is impeccable. In a free-
> field intensity drops as the square of the distance: 20 dB for a tenfold
> increase in distance. Exactly as expressed above.
>
> And, as the paper by Kuhn I cited in my last post shows, once you're
> outside any buildings you are in a near-perfect free field. Morever, the
> difference in permittivity of atmospheric air and a vacuum is extremely
> small - negligibly small many times over if we're talking in terms of dB!
>
> IOW my math is directly applicable to real-world situations!
(Cough) chain link fence (cough).
Let alone the surrounding buildings ;o)
> > Or is someone about to argue that taking a signal below
> > instrument-detectable levels and then dropping it by another order of
> > magnitude is somehow useful?
>
> I have repeatedly explained to you (supported by cites) that this is not
> only possible, it is commonplace.
Okaaayyyy... <snip comment about length of shirt sleeves>
> > Remember, with a simple 400m distance (not a magic number - just
> > thinking of a particular building) you're talking a 52dB attenuation
> > (assuming that pesky vacuum) - i.e. you're signal has dropped to
> > 0.000625% of what you were previously looking at. If you also assume
> > that no establishment with guards that aren't utterly brain-dead will
> > let you lurk within 100m of the wire with a suitcaseful of dodgy
> > electronics, then that drops still further, to 4 tenths of a
> > thousandth of a percent.
>
> Clearly the lot your house sits on is more commodious than mine - your
> Real Estate licence must really have paid off for you :-)
I've never actually lived at a government establishment (although it
certainly felt like it, some weeks!)
> In my house the distance from my computer to the nearest property line is
> a little over ten meters. An observation van parked on the street would
> be less than 20 m away. I believe my numbers are representative of
> houses in many cities
I guess we must differ on the need for Tempested kit, then - my experience
is limited to the main users (near 100%) in government, or the commercial
shielding equivilent in trading floors, etc. I've never, ever seen Tempested
kit outside of those two areas. And rarely, even then.
> It is widely acknowledged that full-bore Tempest specs (AMSG 720B
> equivalence) require 100 dB suppression.
These would be the ultra-secret specs that noone knows about? Just
checking...
> Moreover, I believe your
> attempts to dazzle folks with the "smallness" of numbers that correspond
> to dB attentuation are specious.
But accurate, given they use the self same formulae (and assumptions) of
your own figures; people tend to forget that there is a difference in, say,
the logarithmic sensitivity of the human ear and the more-or-less linear
sensitivity of a straightforward electronic transducer.
> For instance, it is quite possible to
> recive signals on earth from satellites several AUs away which transmit
> with a power of only a few watts. I'll leave you to do the math but,
> without logs and dB, prefixes like pico, atto and femto regularly appear!
Indeed it is, and indeed they do. In the latter case, much more than dB, in
fact (never come across a capacitance measured in deci-Farads), and in the
former - it would take a rather large van to hold a mobile version of
Arecibo.
> > P.S. Almost forgot - if you're not willing to order leaded glass from
> > someone like Pilkington, then it's perfectly possible to buy
> > self-adhesive gold film off-the-shelf and make you're own, as in the
> > A6 Queer (apologies to anyone of that disposition - can't remember the
> > official aircraft designation!)
>
>
> Leaded glass? Gimme a break! We're not making crystal wineglasses here.
> And films and embedded wire meshes are clumsy makeshifts (and I defy you
> to cite any with documented RF attentuation numbers). No what is needed
> is the real-meal-deal: ultra-thin vapour-deposited coatings of things
Yes - leaded glass, not lead crystal. Suggest you look up A6 types - they
*will* have figures, although whether the USN would let you see them is
something quite, quite different.
Anyway, to summarise - we now both believe that the best way to get
Tempested kit is to buy it from a manufacturer, we now both believe that the
actual shielding characteristics are reasonably well known, or, at least,
easily definable.
As far as I can tell, that really only leaves us with questions as to the
proportion of AOL users that routinely use Tempested kit to read their email
(probably measured on a logarithmic scale), and whether Grumman made their
aircraft from whiskey tumblers?
H1K
| 
XML site map