[Elecraft] 2.8 Khz vs 2.7 Khz roof

[Joe Subich, W4TV]

Ken, 

> I am looking for something somewhere that will help
> me decide between the two.

I went back and forth several times before I finally decided 
to go with the 8-pole filters.  

1) The 2.8 Khz filter is actually a bit tighter than the 
   2.7 Khz filter.  As W4ZV writes:  

> 1.  6 dB bandwidth and 6/60 dB shape factor:
> 
> Filter  BW(-6dB)     Shape Factor
> 
> 2700    2910*            2.9
> 2800    2888             1.6
> 
> *The 2.7k may have wider BW than the 2.8k.  This is
> Elecraft's measurement and my 2.7k closely agrees with it.

2) If you extend the analysis to the skirts, the 2.8 KHz filter 
   shows some significant bandwidth advantage at -30 dB.  Using 
   a simple linear skirt approach I first used for comparing 
   the CW filters we see: 

>               2700      2800 
> -------------------------------------------------
>     - 6dB     2910      2888  Hz
>     -60dB     8329      4620  Hz 
>      slope    50.18     16.03 Hz/dB  
> 
>     -10dB     3311      3016  Hz
>     -20dB     4315      3337  Hz
>     -30dB     5319      3657  Hz
>     -40dB     6322      3978  Hz
>     -50dB     7326      4299  Hz

While the shape/width of the roofing filter is not critical, 
the added selectivity can't help but be some benefit when the 
strong station fires up 2.5 KHz away.   

3) Eric indicates that the 8-pole filters are tested to a 
   higher standard than the 5-pole filters.  They will have 
   less filter IMD than the 5-pole filters. 

For me, the little things added up.  I decided "better safe 
than sorry" was the best course of action ... I ordered the  
8 pole filter up front rather than possibly regret not doing 
so every time I used the radio.  However, the second receiver 
will have the 2.7 KHz filter.  

73, 

   ... Joe, W4TV