With the HT tuned -4MHz, I heard nothing and the meter was zero as expected.
I'm assuming that the battery was good, the unit was turned, and properly
connected in line between antenna and receiver. And nobody had
transmitted through it by mistake...
In that case, I'd expect it to work if you have adjusted the attenuation
control through the range. (The pot might be wired so attenuation is
maximum at either end of the scale.)
When I've measured the meter range of mobile FM rigs it often is
about 12 dB, so I try to adjust my attenuators to have about 10 dB
of loss: minimum overlap means I get more attenuation out of it
at the other end of the range. But attenuation varies with signal
level and some other factors, so sometimes I can't hear the signal
(or can't get a meter reading) until the meter is full scale over at
least 180 degrees of azimuth turning the beam. In that case, I
have to "eyeball average" where the signal hits full scale on either
side of the main lobe and split the difference. Some day I'll
adjust it for a bit more overlap...
So the next question is, how close did you have to get to the
transmitter in that case to hear it using the attenuator? That
gives a better indicator whether the through attenuation is
just a bit too high, or the box isn't working at all.
Actually there is another option: many of the simplistic
single-diode attenuators can be used both in THROUGH
mode (receiver tuned to transmitter frequency) once the
signal gets too strong, with the signal level varying with
the knob setting. Then, when the signal gets too strong
in that mode, you retune the radio to use it in OFFSET
mode.
If that is the case, then the OFFSET mode may be
set for stronger signals, which could explain why you
can't hear the signal at that distance: try inserting it
and using it
without changing the receiver frequencyand see if that works.
The difference between the two modes is whether RF
or DC is applied to the mixer diode. Simplistic circuits
apply both at the same time, so no switching is required
between modes. But I've found this can cause issues
with other off-frequency signals, so I build my attenuators
with two separate modes, and can adjust the overlap
between them.
If you have a schematic you can send me, I can look at
the circuit and see how it is intended to operate.
Note that performance also depends on the design of the
receiver you are using it with. There have been complaints
that the Baofang and similar receivers don't have enough
filtering in the front end, so can be overloaded by the
fundamental signal even when tuned 4 MHz off frequency.
In that case, the signal can disappear when the transmitter
is too strong. I don't have such an HT to test with: my HTs
are older (and larger), and they have better filtering built in.
I guess I need to figure out how to use the NULL on the yagi better.
That certainly is one approach, though the null is not always an
accurate method of taking bearings. It tends to be frequency
sensitive: the classic WB2HOL tape measure yagi has a very
deep rear null for several hundred kHz around the design
frequency, but I can't count on using it on one of our typical
hunts where we may have transmitters in the 146 and 147 MHz
ranges. Some yagi designs may actually have a peak in the
rear direction with nulls on either side of it. And reflections,
or signal pickup on the coax, can shift the apparent direction
of nulls in the pattern.
Rather than a sharp rear null, I design my DF antennas so that
the rear attenuation is better than about 15 dB over the whole rear
quadrant of the antenna over the whole 2m band. That doesn't
give me as precise of a reading if I have to use the null, but,
given the limited range of most meters, if I can see a lobe in
the rear I know it must be due to a reflection, not the antenna
pattern.
So the first step would be to see if there is an intermediate
THROUGH mode that allows you to attenuate the signal
without retuning the receiver: if so, that is the simplest
solution. If not, then there could be something wrong with
the unit, or the minimum attenuation may be too high for
your equipment. Often that can be fixed by varying one
resistor inside the unit.