On Mon, 19 Feb 2007 01:09:52 GMT, james wrote
in :

On Sun, 18 Feb 2007 09:21:02 -0600, "Pete KE9OA"
wrote:

+++I understand what you are saying, but the RF amplifier should be conjugate
+++matched to 50 ohms anyway, in order to have uncondisional stability. I am
+++not sure what the noise figure of this system is, but it seems that the gain
+++distribution is such that most of the gain is in the 2nd I.F. strip anyway.
+++Even so, under 30MHz, in most areas the excess environmental noise is in the
+++15dB region so a receiver with a 12dB noise figure does just fine.
+++I remember the old Allied Model 2568 CB radio. This thing had quite a bit of
+++RF gain and relatively low I.F. gain. As soon as you connected an antenna,
+++it sounded like an FM unit. The problem with that design is that the AGC
+++voltage was derived from the RF stage with its relatively low selectivity,
+++in addition to the I.F. strip. Strong off channel signals would capture the
+++AGC loop and desense the whole system. Remember the old term "bleed over"?
+++You do have a good point about keeping the RF gain ahead of the mixer as low
+++as possible, since any gain ahead of the 1st mixer degrades the dynamic
+++range by that same amount. In my last contract with Motorola, we were using
+++mixers that had an IP3 of +40dBm so we were able to get away with having
+++some gain ahead of that mixer.
+++
***********

Conjugate match is needed for maximum power transfer.


IMPEDANCE match... for maximum power transfer. A 'conjugate' match is
when the impedances are complex, which isn't always the case.


Nuetralization
helps extend stability over various mismatch condistions.


Lots of things can improve stability, but unless the amp/mixer is
oscillating the point is moot.


In a receiver RF front end it is preferable to match for best noise
figure and accept the gain. The less gain before the mixer the better.
The RF front end sets the noise figure for the whole receiver. The
gain of the RF Front end need only be high enough to overcome the
noise figure of the next stage and any losses it may present if any.


All true. But the point I was trying to make (which I incorrectly
assumed was already understood) is that any impedance matching device
or network between the antenna and the 1st RF can cause more noise
from IMD than the noise from the 1st RF or mixer, -especially- if that
matching device/circuit uses a ferrite core or solid dielectric, which
includes almost all CB radios. That's why strong signals can sometimes
be heard even when the source is several MHz away (often mistaken for
receiver overload).

The concept here is to reduce (or, ideally, eliminate) that impedance
transformation stage. A long time ago I built a common base (voltage
follower) RF preamp using eight transistors in parallel followed by
the impedance transformation stage (transformer). The input impedance
directly to the transistors is about 100 ohms, but I feed it directly
anyway. The difference is like night and day. I use it for lowfer work
these days.


More often than not,CB calibur radios places far to much gain in the
first stages so that more simple IF stages can be used. Thereby
lowering costs.


Like you said before, the first RF is all that matters.