Coax Loss in Non-LOS applications

[seligman]

I have a question I'm not sure how to answer.

We're in the process of installing a permanent 900 MHz antenna at a friend's house. He tested my 7.2 dBi gain 900 MHz whip with a magmount base on his rooftop, and we were wowed by the performance -- especially for non line of sight reception. Various radio propagation modelling programs claim this shouldn't be possible due to miles of obstructions (hills) in between.

I explained to him the perils of coax loss over long lengths. Running 30 to 40 feet of LMR-400 would introduce an additional 0.5 - 1.0 dB of loss compared to the short length of LMR-195 on the magmount. The question I'm having a hard time answering is what effect this will have practically speaking with the distant non-LOS locations, especially at 900 MHz.

Would another 1 dB of loss be the difference between hearing and not hearing a distant non-LOS transmit site? In other words, is non-LOS reception more sensitive to lossy coax and assembly loss?

 

[tech020]

Coax loss

Loss is loss, regardless of length or type of coax. Having an antenna in a desirable location, you need to calculate what the line loss would be for your coax. Usually the gain from height offsets the cable loss unless too long a cable is used. If the same antenna at ground level doesn't receive the desired signal, but will receive at a given height, then one should use a feed cable with the lowest practical loss. Usually for receiving purposes, a line loss of 6db is passable. For transmitting, 3db is better. I worked on a trunking radio system. We used tower top amps to increase receive performance. On transmit, our 70 watt signal was reduced to 20 watt due to combiner losses and then suffered 3db (50%) loss in the coax going up the tower. Still with gain antennas, the 10 watts were actually 100 watts effective radiated power.

 

[seligman]

He put the magmount (with 7 dBi gain whip) on top of the chimney for testing purposes. That's where he wants to mount it permanently, on a different mount of course, but the coax run will need to be 20-30 feet longer. I calculated the loss with LMR-400 to be an additional 0.5 - 1.0 dB compared to what he tested with, excluding any assembly loss.

I don't know how to answer the question whether an additional 0.5 - 1.0 dB loss will have a meaningful, noticeable difference for him. He's not a radio guy so quoting him loss figures isn't helpful. How do I explain this in practical terms?

 

[prcguy]

1dB of loss is generally not noticeable by most people unless you can A-B the change in real time. If a signal is already weak a scratchy it will be slightly more weak and scratchy with an addational 1dB of loss.