Installation Instructions

Hardware Installation
RVP8 User’s Manual
September 2005
2–14
When two amplifiers are cascaded so that the output of the first drives the input of the second,
the overall gain is the product of the two linear gains
G
1
lin
and G
2
lin
, and the overall noise figure is
computed from the two noise factors
F
1
lin
and F
2
lin
as:
NoiseFigure + 10 log
10
ƪ
F
1
lin
) ǒ
F
2
lin
* 1
G
1
lin
Ǔ
ƫ
where the two noise factors are simply the linear representations of the noise figures that were
expressed in deciBels:
NoiseFigure + 10 log
10
[
NoiseFactor
]
.
Suppose that our first amplifier is an LNA/Preamp with a 2dB noise figure (noise factor 1.58),
and we want to know what gain it must have such that, when cascaded into the RVP8/IFD, the
overall noise figure will be 3dB. The 32dB noise figure of the IFD is equivalent to a noise
factor of 1585, hence we have:
3dB + 10 log
10
ƪ
1.58 ) ǒ
1585 * 1
G
1
lin
Ǔ
ƫ
from which we solve G
1
lin
+ 3771, i.e., 35.8dB. This agrees with the 36dB of gain that was
computed in the example of the previous section for the same RF/IF components and desired
overall performance.
2.2.10 Choice of Intermediate Frequency
The RVP8 does not assume any particular relationship between the A/D sample clock and the
receiver’s intermediate frequency. You may operate at any IF that is at least 2MHz away from
any multiple of half the 35.9751MHz sampling rate (nominally 18, 36, 54, 72 MHz). The valid
frequency bands are thus:
6-16MHz, 20-34 MHz, 38-52 MHz, 56-70 MHz
There are many reasons for staying clear of the Nyquist frequency multiples. Most of these
considerations would apply to all types of digital processors, and are not specific to the RVP8.
As an example of what can go wrong at the Nyquist frequencies, suppose that an intermediate
frequency of 35MHz was used. This is only 1MHz away from the (approximately) 36MHz
sampling rate. The external anti-alias filter must now be designed much more carefully since a
spurious input signal at 37MHz would be aliased into the valid 35MHz band. If the valid signal
bandwidth were 2MHz, then the anti-alias filter would have the difficult task of passing
34–36MHz free of distortion while rejecting everything above 36MHz. The filter’s transition
zone would have to be very sharp, and this is difficult to achieve.
Another problem that would arise with a 35MHz IF on a magnetron system would be the RVP8’s
computation of AFC. If the processor can not distinguish 37MHz from 35MHz, then it can not
tell the difference between the STALO being correctly on frequency, versus being 2MHz too
high. The symmetric AFC tracking range would be reduced to the very small interval
34–36MHz.