[Elecraft] K3 2.7 k Filter vs. 2.8 k

[Don Rasmussen]

Now we are comparing apples to apples. The original
question was using the 2.7k 5 pole filter or going to
2.8 8 pole filter. In either case a VERY robust AGC is
waiting on the receiving end of the filter, followed
by a very robust DSP. 

How much did the Elecraft guys pay for those fancy
no-wear buttons on the radio's front panel? For the
(2) hearty VFO encoders, or to remap and repaint the
front panel for that matter. If there was any real
advantage in going to 8 poles at 2.8 I think it would
have been included as stock with the base K3.

I think of some of the blatent cost cutting you find
on other transceivers that would never happen with the
guys. Also, when I see those QST ads of the other more
expensive offshore brands and they are showing this
tuned device or that (internal) tuned mechanism and I
figure what it cost to put it in there and the
relativly small advantage they get in the specs. I
wonder who is getting fooled. ;-)



[Elecraft] K3 2.7 k Filter vs. 2.8 k
Vic K2VCO vic at rak 
Tue Sep 18 17:06:02 EDT 2007 

K4IA at aol.com wrote:

> Well, sorta kinda.  If the unwanted signal gets in 
and pumps the AGC, then 
> the DSP might knock down the signal but won't be
able to  do anything about the 
> AGC pumping.  Then, you won't hear the signal if it 
is weak.  That is the 
> whole issue of the DSP being in or out of the AGC 
loop and it generated a 
> mountain of controversy with the Orion when TT moved
the  DSP to within the loop.

The K3 has two AGC loops. One is part of the DSP, so
as long as the DSP 
does not overload, a signal outside the DSP passband
but within the 
crystal filter passband will not cause the AGC to
pump.

But, if the signal is greater than, I think, S9+30 (30
dB above 50 uv), 
then it might overload the DSP. So in that case, the
analog AGC which is 
outside of the DSP kicks in to prevent that, and you
would notice a gain 
reduction on the desired signal.

This is where you could use a narrower roofing filter.
I ordered 2.8, 
1.0, and 0.4 KHz filters. I noticed this effect with
huge signals by 
listening with the bandwidth set to 450 Hz, which uses
the 1 KHz crystal 
filter. Moving it to 400 Hz causes the narrower filter
to be activated, 
which eliminates the pumping.
-- 
73,
Vic, K2VCO
Fresno CA
K3 no. 00007 ('James')
http://www.qsl.net/k2vco