IC-746 FAQ
4.9
Can an IF filter be too narrow?
Yes, an IF filter can be too narrow, but sometimes, that can actually work to your advantage.
We'll describe only a few specific situations as examples.
When the IF filter's bandwidth is less than the bandwidth of the desired signal, then part of
the signal will be rejected. This degrades the fidelity (quality) of the received signal. In
phone modes, the fidelity loss is often mistakenly perceived or described by operators as
"distortion".
Fortunately, the inherent information redundancy of human speech allows it to be readable
even when an overly narrow filter has rejected part of the desired signal. That's where our
advantage comes from.
One form of interference is when a nearby signal overlaps your desired signal; as in an
overcrowded band. Rejecting the interfering signal necessarily rejects part of the desired
signal as well. Thus, the filter is too narrow to allow all of the desired signal to come
through, but it is narrow enough to reject the interference. While the result is "distorted" in
comparison to a clean signal, it is still more readable than if you left the interference there.
Another example would involve CW or PSK31. The more narrow the filter is, the less the
interference and noise. In these modes, it's almost impossible for a filter to be narrow
enough to actually reject part of the desired signal, as in the previous phone example.
Instead, when a CW/PSK31 filter gets too narrow, you start to experience the auditory
equivalent of tunnel vision. This can make it difficult to scan around for signals.
One of the handy things about TPBT is that you can narrow your effective IF filtering
bandwidth at any time.
4.10
What's the difference between IF filters from Inrad vs. Icom?
Inrad filters tend to have a superior shape factor to Icom's (smaller numbers are better) as
well as deeper stopbands. They also tend to be cheaper.
When dealing with adjacent channel interference, the filter shape (depth and slope of the
attenuation curve) is generally the prime factor. Inrad filters tend to outperform the Icom
filters for this application.
For example, the stock Icom filter in the 455Khz IF has a -6db bandwidth of 2.4Khz and a -
60db bandwidth of about 5Khz. That gives it a shape factor of approximately 2.1. The
comparable filter from Inrad has a shape factor of 1.6, a -6db bandwidth of 2.4Khz and a -
60db bandwidth of 3.8Khz.
This all means that the Inrad filter has a steeper attenuation curve, resulting in better
extraneous signal rejection and improved selectivity.
The Inrad filters tend to have greater ultimate attenuation than the Icom counterparts. Thus,
Inrad filters tend to do a better job of adjacent signal rejection than Icom's filters.
However, while the 455Khz Inrad filters are drop-in replacements for the Icom counterparts,
the 9Mhz filters are not. They need a short piece of coax and a connector to wired up to the
filter first. The adapting connector plugs into the filter slot, thus alleviating any mods to the
Page 54 of 113
2003-02-08
