VK3XL wrote:I never had any doubt that mast head mounting the pre amps was the best way to go....
There is NO DOUBT that the system Noise Figure is lowest when the coax loss before a given properly tuned LNA is lowest, ie that it is located nearest to the antenna connector.
A more intelligent analysis is to quantify HOW MUCH BETTER, and to weigh that BENEFIT against the DISADVANTAGES of location of the preamp AND its input filtering at the masthead.
The analysis will depend on the band (amongst other things).
Since almost all the responses so far have been anecdotal, and I don't recall a single one offering a numeric value for S/N improvement (which has to be the principal reason for using an LNA), allow me an anecdote.
A Canberra 2m station was located about 7km from the TV transmitter site of Black Mountain. His rx lineupt was antenna, feed line to shack, LNA, coaxial resonator (filter), transceiver. The coaxial resonator (most people will call that a cavity resonator, but anyone who has looked inside them knows it isn't a resonant cavity, but a TEM mode coaxial resonator) was to reduce IMD effects mainly due to the TV transmitters.
I asked him why the filter was located between the LNA and transceiver and his response was one of astonishment "where else would you put it", apparently not for a moment thinking of putting it in front of the LNA. When I asked if he had compared the measured S/N on a consistent weak signal for both configurations, he hadn't and was confident that the insertion loss of the filter placed in front of the LNA would seriously degrade system NF. He did agree to try the configuration, I felt more for the opportunity to demonstrate that I was talking nonsense... because his report of the difference conveyed more than genuine surprise that it was better, disbelief even... "who would of thought...".
The reason for this outcome was that although the filter in front might have introduced some tenths of a dB of Insertion Loss (which as you know degrades system NF dB for dB), the reduction in mixing of out of band signals to produce in-band noise was significantly greater giving an overall benefit. The solution and the benefit were specific to the scenario, and whilst the optimal configuration in that scenario is an option to be explored in another places, it might not be the optimimal solution in that other place.
Shortly after, I heard this chap advising another ham closer to the TV transmitters to try the same thing, but the latter could not get his mind around it as he KNEW that the BEST location for an LNA was at the masthead, and he sure as hell wasn't going to put a coaxial resonator up there with it. Of course, he was correct in recognising the disadvantage of putting a 200kHz wide coaxial resonator where temperature might vary from -10° to 60° and where it would need protection from rain, but his failure was in KNOWING that the BEST location for the LNA was simply and always at the masthead... that is when ROT (Rules of Thumb) become rot.
Why the filter discussion? Because many if not most LNAs do not give you an filter response curve, and the filter can be very important in the real world. Some designs work very poorly on some antennas, depending on the narrow bandpass effect of very high gain arrays... which is all right so long as that is what you are using.
One of the references earlier in the thread is for a source that includes in its 'product list', an LNA for 2m with NF=0.04dB. This cannot have a very effective input filter with such a low NF... but you could be lured by the low NF than cannot be realised in many practical antenna systems (eg 4 bay configurations).
The challenge here is to find a solution which is a reasonable compromise of competing factors for your own situation, and the greatest variation from station to station is the external noise environment yet almost no one quantifies it, and it is NEVER reflected in device specifications that are measured in a shielded room.
Owen
) that having the pre-amp ON makes a good improvement in the signal to noise ratio figures of decoded signals.