[Elecraft] K2 Ripple in SSB Filter

[Don Wilhelm]

Michael,

That kind of ripple is normal when the variable xtal filter is operated 
at wide bandwidths, but it still provides an intelligible signal 
(surprisingly).

Since you are using Spectrogram, check the actual width of those wide 
filters.  Typically on newer K2s the actual width is much wider than 
that indicated by the K2 - there is not much sense in having an FL2 that 
is actually 2800 Hz wide when the FL1 (OP1) filter is 2400 Hz wide.  
Narrow the actual width to something usable (less than the OP1 filter) 
and the ripple will be reduced as the width is reduced.   The easiest 
way is to use LSB - first set the BFO to place the low frequency slope 
of the passband (near 300 Hz) and then change the width and you will see 
the high frequency end of the passband move a lot more than the low 
frequency end.  When you have the actual width correct, record the width 
number on the K2 display.  With a 2400 Hz OP1 filter, I usually 
recommend actual widths of FL2=2200 Hz, FL3=2000 Hz and FL4=1800 Hz.

BTW - this behavoir is the fault of the varactors - the varactors being 
shipped with newer K2s have a greater capacity range than those used 
when the firmware was developed, and it also makes the filter frequency 
charts in the manual incorrect in this respect - use Spectrogram and 
measure the actual width is the only known 'cure' if the firmware is to 
support older K2s as well.

There is not much ripple when the actual width is somewhere between 1200 
and 1600 Hz, and the actual width will more closely track with the K2 
displayed width at settings of 1.00 and less.

73,
Don W3FPR

Michael van Hauten wrote:
> Hello elecrafters,
> after rechecking  all my filters with sprectrogram, i noticed that there is a ripple of about 10 db in all SSB filter settings without the OP1 settings. Is this normal for this kind of filters. Is it possible to reduce the ripple?
>
> vy 73 Michael, DC0ZO
>
> k2# 6614, k3#1999
>   
>