notch filter
amstel78,
I think I can explain - at least some of the apparent "strangeness" you are seeing.
When you tried to "notch out" the interference at 118.5MHz you expected that tuning the filter to 118.5MHz would reduce the level there. As you say, you understand that you should tune the filter to the interferer's frequency but you thought that you could test it by tuning it to any frequency wherein you receive interference and see the level drop. Unfortunately, it will not work that way. The noise/interference you here at 118.5MHz or 108MHz is not really there - it is on some other frequency or number of frequencies. It is "bleeding over" - that is, in this case, appearing on these other frequencies (118.5MHz and 108MHz, to use your examples) because of weaknesses in the receiver's frequency conversion networks when faced with certain high level out-of-band signals. So by tuning the notch filter to your "desired" frequencies (118.5MHz and 108MHz) you will hardly at all affect the apparent level of noise/interference because it isn't really there anyway (but the scanner is being "tricked by brute force" into making it appear there). The only way you can really test the notch filter is to actually try and notch out a real signal that is really on the frequency you want to test it at. I suggest, if you want to try this, to try your local weather frequency at 162.XXX MHz and tune the filter there while you are listening to it and see how much it drops the signal. Any other desired signal, such as a pager signal at 152.XXXMHz or taxi base station or some LMR strong signal that you know belongs on some frequency within the notch filter's tune range (80MHz to 190MHz) would also be a good candidate for testing; I mention the weather frequency because it is continuous and therefor easy to test. If the filter will not work on real known on-channel signals within its specified range then it may, indeed, be defective. If it does work, and despite all your best efforts, cannot reduce or eliminate your problem then it may be that:
1) You have more than two interferers that are causing the problem ("intermod" as described by most scanner users, is usually caused by the interaction of two interferers which internally mix to cause the the result to match the receiver's internal IF; when this happens, you hear a mixture of those two signals being heard all across the band between the two interferer frequencies - if you can notch out one then you should be able to solve the problem. Unfortunately, if there are multiple interferer "pairs" causing such problems then you would need multiple "notches" to cover them all).
2) You have a severe overload situation wherein the interferer is swamping the receiver by either the fundamental or a harmonic of the fundamental simply overwhelming some point in the receiver's frequency conversion chain. Depending on how the signal is really getting into the receiver and how strong it is, it may not be easily dealt with.
3) The interferer's frequency (or frequencies) is/are located outside the range of the notch filter's "notchable" bandwidth.
4) The notch filter is just not attenuating the interferer enough due to your proximity, etc., as we've already discussed.
Or some combination of the above.
There are other avenues of typical receiver interference such as images, 1/2IF, and basic adjacent channel interference. But all those are specific to a desired frequency (only affecting a certain specific frequency) - it sounds like you are getting a "splatter" type interference which covers a large range of frequencies so I think it more likely you are suffering from the possibilities I noted above or some combination thereof (the last two are actually sub-categories of the first two, in a sense).
I honestly wish you luck! And I honestly sympathize! But I would test the notch filter first by trying it on a well known constant (NOAA weather broadcasts) or almost constant (pager, taxi base) signal within the filter's range and see what it does - if it has little or no effect on those then it may indeed be defective.
-Mike