Yes, that's the first of the two articles from G3VA's Technical Topics
column in RADCOM
(Including a reference to Peter G3RZP, who frequents these boards.) The second was a month or
two later and included comments from those who tried to build it.
I think there are likely some better loop antenna designs available, as well as other methods for
achieving a null (phased dipoles, ZL Special, Moxon beam, etc.)
The standard WB2HOL "tape measure yagi" is designed for a very sharp rear null at 146.56 MHz, at
the expense of a bit of forward gain. Hunting the null is capable of very sharp bearings, and can
be especially useful when you don't have a good attenuator, as it is easier to see the difference of
a 30dB null on a standard FM HT than a 3dB peak in the pattern. However, getting a reliable null
that holds over frequency is not trivial: a pair of phased dipoles fed with equal-length cables will
work, as the null is not frequency-specific, as will a standard loop, but both are still bidirectional.
I hunt the peak of the pattern on 2m rather than the null, as it is less affected by reflections
and common mode current, and tends to be much more stable over frequency. But that requires
a good attenuator (preferably the "offset" or "active" type) and a receiver that can show small
variations in signal strength. Fortunately the S-meter in many FM rigs has only about a 12dB
range, which works well enough with "eyeball averaging". For serious work, the VK3YNG "sniffer"
has an audio S-meter that allows me to detect a 1/4dB difference by the pitch of the tone,
which makes bearings off the peak much more accurate.
Meanwhile, here is G3ZOI's modified construction of the WB2HOL "tape measure" yagi using
commonly-available British parts: http://open-circuit.co.uk/wp/wb-yagi/
And here are my modified dimensions if you want to optimize it for gain/forward pattern
rather than the rear null: http://www.ardf-r2.org/equipment/wb6byu_2m_yagi.shtml