If you look at the entire radio spectrum with a spectrum analyzer off a wide band antenna you will see potentially hundreds or more frequencies in use all at the same time. Some will be very weak like less than -100dBm and many will be very strong. Strong signals might be at levels of -50dBm or higher and they can even approach -10dBm and higher.
One of the Nooelec LNAs has a 1dB compression point varying from -10dBm to 5dBm depending on frequency. That means if you feed it a very strong signal at -10dBm the amplifier is now compressed and cannot put out hardly any more signal, you hit its limit on just that one frequency. Testing the LNA in a lab, feeding in only one frequency is very predictable. Feeding two different but equal level signals to the same amplifier will create Intermodulation Distortion or IMD at some specific level that will be way below where the amplifier saturates. I'll take a guess that two equal signals in the -80 to -70dBm range will start to produce IMD in this particular Nooelec LNA in a lab setup that you will hear in your receiver.
Intermodulation distortion will produce sum and difference frequencies of the two different frequencies in the lab setup above. Once IMD starts, if you increase the levels of the signals that are causing it, the level of IMD will grow at a 2:1 ratio, meaning if you double the level of the original two signals, the IMD will go up by 4X and so on.
With all that out of the way, connecting the little Nooelec LNA to a broadband antenna will most certainly spank it with gobs of signals that are above the -80 or -70dBm range and the more frequencies that are well above the level that will create IMD, the more sum and difference frequencies will be produced. What happens in the end is, so many IMD sum and difference frequencies or copies of the original signals but each containing modulation from all contributing signals will be so dense that it will simply become a huge noise floor as they will be all over the place and tightly packed together.
So you may have started out with a noise floor at the end of your coax in the -130dBm range allowing you to hear a -120dBm signal ok, but adding the low level rated Nooelec LNA has created tons of IMD or ghost signal copies of other signals and that now looks like a -120dBm or higher noise floor. The IMD is now covering up weaker signals that you could hear without the LNA and the noise floor will be constantly changing based on the amount of signals hitting the LNA and their specific levels.
The fix is to get an LNA with high enough IP1 and IP3 ratings so it can survive in your RF environment and not create internal IMD that will raise the noise floor and degrade your reception. It will be a combination of IP1, IP3 and the noise figure rating that will determine how well it will play for you and the better those numbers get, the higher the cost. Just to throw some numbers out there, an LNA with a 1dB compression point of 30dBm or 1 watt, an IP3 of 50dBm or more and a noise figure of .5dB over the 100MHz to 1GHz range would be the Holy Grail of LNAs for our hobby and would serve most people well on a wide band antenna with no problems.
I have not been able to find this Holy Grail and have seen LNAs with and IP1 of 26dBm and IP3 of 46dBm but the noise figure is high at 3.5dB. Or you can get a lower noise figure of .5dB but the IP1 is only 22dBm and the IP3 is only 39dBm. We haven't even mentioned gain yet and for feeding a single receiver you would want to keep that in the 10-15dB range at the very most. Maybe more if feeding a divider and many receivers and the Nooelec LNA is something like 20dB gain, which is just too much to feed a single receiver.
Ok, so an interesting observation... and one I can't quite explain. Insertion of the Nooelec LaNA into the signal chain caused a degradation of signals that doesn't appear to have been a result of overloading. When the LNA was in the signal chain, powered on, there was a noticeable weakening of signal integrity. Signals were received with lower RSSI... the precise opposite of what one would expect. It definitely wasn't overloading, because strong signals from aircraft were still coming in loud, clear and with full quieting, but weak signals were noticeably weaker. Almost as if the LNA resulted in signal loss rather than gain. Order of the LNA/Filter made no difference.
Perhaps someone can explain this to me.
Edit: Before someone asks, it was definitely installed in the correct orientation, unless it is mislabled.