| |
VK3XX > TECH 08.04.06 06:11l 64 Lines 3384 Bytes #999 (0) @ WW
BID : 511442VK3XX
Read: DL1LCA GUEST
Subj: Re: Noise figure query
Path: DB0FHN<DB0FOR<DB0MRW<DB0PV<OE7XLR<DB0ZKA<DB0GPP<DB0OFI<DB0PRT<DB0LHR<
DB0GE<LX0PAC<ON0LGE<SR1BSZ<SP7MGD<VK7AX<VK6HGR<VK4TRS<VK3FBD
Sent: 060408/0351Z @:VK3FBD.#MEL.VIC.AUS.OC #:2559 [Lilydale] FBB7.00g
From: VK3XX@VK3FBD.#MEL.VIC.AUS.OC
To : TECH@WW
In response to VK3ZRG's query here are my thoughts.
If you have a 3DB loss between the antenna and the radio that is almost
predominantly caused by the loss in the coax feedline and connectors and
will determine the minimum system noise figure however good the rf stage
of the receiver might be. It is also very much physical length and
frequency dependent. One way ahead is to use much better cable. Heliax
provides for example a much better RX noise figure compared to a typical
high grade coax and I dont mean RG58!!!. It is more important on RX than
its loss of power on TX.
This also leads to the practice of putting the initial RF stage at the
antenna rather than in the RX or on the sack table. On transmit only the
radiated power from the TX is lost in the feedline losses and that can be
compensated for by increasing power output.
So in your hypothetical case of being able to cool the antenna noise
which, apart from local electrical noise is really of galactic origin,
then in practical terms it really doesnt mean much in defining the noise
figure of a practical receiving system. Ultimately galactic noise becomes
the limit.
When setting out your program you should really treat the feedline as a
stage of amplification having negative gain. Apart from the first stage of
rf amplification, because feedline loss is real, the only point in adding
further RF stages in cascade does nothing to the noise figure of the
system. Only the gain distribution in the RX is affected such that the
noise component ahead of the first mixer is sufficient such that the mixer
noise is itself exceeded. If the initial stage of rf amplification is at
the antenna with virtually no feedline, then short of packaging the
antenna, the feeder and the rf stage in liquid helium. for example, to
reduce the thermal noise from those components, do you come close to a
noise figure that is the ultimate for a given receiving system. Also the
gain of an antenna system just affects the overall system gain
distribution without doing anything for the signal/noise ratio (noise
figure) of the system.
This also suggests why, having bought a super 0.6 DB(or less these days)
noise figure pre-amp for example, and locating it in the shack, the
signal/noise is in most cases no better than it was without it. Your
program should reflect this fact. Many good VHF/UHF radios have a front
end which has a good noise figure (eg a GASFET) often sound very lacking
in gain (S meter reading). This is because they have been designed for a
gain distribution that allows for a 20DB pre-amp without compromising the
large signal handling capability (Dynamic range) which gives rise to so
many spurious signals being heard from localised noise sources (pagers
etc).
At the end of the day the noise figure of the RX does not play such a
significant part on the ability to dig deeply for weak signals. It
certainly did when vacuum tubes were the front end devices in a radio or a
pre-amp. Keeping the losses ahead of the radio is the only way to improve.
The biggest culprit will always be the antenna to radio loss which
determines the practical outcome. All the super numbers should be kept for
the adverisers and then treated with a pinch of salt!
I hope that these thoughts are helpful to you.
Cheers and 73 from Gordon VK3XX@VK3FBD.#MEL,VIC.AUS.OC
Read previous mail | Read next mail
| |
