From: Bill Stewart <bill.stewart@pobox.com>
Date: Tue, 10 Feb 1998 08:43:07 +0800
To: Jim Choate <cypherpunks@ssz.com (Cypherpunks Distributed Remailer)
Subject: Re: Laptop TEMPEST (fwd)
In-Reply-To: <199802091430.IAA20623@einstein.ssz.com>
Message-ID: <3.0.5.32.19980209163616.0088a200@popd.ix.netcom.com>
MIME-Version: 1.0
Content-Type: text/plain



At 08:30 AM 2/9/98 -0600, Jim Choate wrote:
>> FCC RF/EMC testing is well nigh useless for TEMPEST protection.
>
>I disagree, it would give you a gross baseline on the total emissions
>between monitors and laptops. That field strenght measurement would at least
>allow you to calculate radiuses of equal strength to calculate approximately
>how far the emissions are from each class of device for equal probabilities
>of detection. One of the specific goals is to measure how effective the
>device is at effecting other co-located devices (such as seeing ghost images
>on other monitors or causing static in paging equipment).

FCC specs aren't real tight.

My old laptop*, which passed FCC specs (at least when it was new,
before being hauled around on airplanes and trains, dropped a few times,
having cooked parts replaced, and generally getting treated like a laptop),
broadcast enough emissions that its video would show up on my parents' TV
when used 2-4 meters away.  The video sync wasn't right, so there were
several hard-to-read images of the screen on the TV, but it was obviously
emitting enough that a properly tuned receiver could read it.
On the other hand, it doesn't bother my TV from 8-10 meters away,
but I've got cable, while my parents use Real Rabbit-Ear Antennas and 
some kind of antenna-booster widget.

I have heard that passive-matrix emits less, but I don't know
if that's true from a TEMPEST standpoint or only noise output.

It is a bit easier to put a laptop into an RF-shielded enclosure,
but enclosures that worked fine for harmonics of 4.77 MHz machines
don't necessarily block harmonics of 477 MHz machines :-)
especially critical are the penetrations used to get air and
fiber-optic connections through, which are basically waveguides -
once the wavelength of the signal is short enough, they lose.
___
* AT&T Globalyst 250P, which is a NEC Versa Pentium-75 with a Death-Star
painted on it.  It has the expensive active-matrix 24-bit-color 640x480
screen (sigh - I'd rather have had the cheaper 800x600 8-bit-color :-).
A couple years old, and the case was a bit dented so perhaps it had been 
more radio-tight when it was new.
				Thanks! 
					Bill
Bill Stewart, bill.stewart@pobox.com
PGP Fingerprint D454 E202 CBC8 40BF  3C85 B884 0ABE 4639