Hi,

The I.F. is stagger tuned, with swamping resistors to kill the Q. It is
running at 60mhz with a 6db bandwidth of 8mhz. It is part of a 'short pulse'
marine radar unit. The pulses were like maybe .06microseconds at about 1khz
rep rate.

The transmitter now squirts out 'long pulses' of about .8microseconds. I
want to raise the value of the swamp resistors, and 'un-stagger' the tuning
to get a 1.5mhz bandwidth at the 6db point.

So the original question was "If I reduce the bandwidth from 8mhz to 1.5mhz,
and everything else remains the same, what will happen to the signal to
noise ratio in terms of db snr?

I don't pretend to be able to answer this question completely as the subject
is more complicated than it would seem. However, all things being equal, SNR (
Ps/Pn) is inversely proportional to bandwidth so you would expect it to
increase by 10 x log(8/1.5) ~ 7dB for the values you give.

It is important though that bandwidth and pulse length match for optimum
results and in "Threshold Signals", MIT Rad Labs Vol. 24, they show
experimental evidence that this is indeed the case (B x Tau = 1). However you
intend to reduce the bandwidth by a factor of ~5 and to increase the pulse
width by ~13 (a B x Tau increase from 0.5 to 2.5) and the graphs say that this
should decrease the visible threshold by ~3dB.

There are a few other things I would mention. An increase in pulse-width
will increase the power dissipation in the transmitter by the same factor,
should reduce target discrimination (1us is 300m and so particularly important
close-in) and that a wider bandwidth is helpful in simplifying tuning or
reducing constraints on the system AFC.



Cheers - Joe, G3LLV