[Elecraft] K2 ultimate rejection

[Geoffrey Mackenzie-Kennedy]

Martin,

On paper it would not be too difficult to "build"  an IF crystal ladder 
filter for the K2 whose stopband attenuation is an honest 100db or better, 
with a passband ripple of something less than 0.5db - probably less than 
0.25db. But the filter would require good quality high Q crystals, 
reasonably well matched.

In practice to use such a filter in the K2 would, I believe, require quite a 
lot of  surgery so that the filter's performance is not ruined by unwanted 
coupling between the input and output, "blow-by" is one example of this. 
Consider all those control lines as a first thought! If a KSB2 is used, it 
would probably be necessary to reposition it, and as a guess some shielding 
might be required in other areas as well.

If different and selectable IF bandwidths are required, then I am certain 
from past work that a filter for each bandwidth would be required, selected 
by high quality relays that have the necessary inter-contact isolation. 
There is no place for diodes of any sort in this 100db+ playground.

However, even if a better filter was installed and worked properly, the very 
close in and far out phase noise of the K2's LO seen at the3 TUF-3 mixer LO 
port would have to be reduced, if anything is to be gained from the new 
filter's stopband attenuation. Kevin, ZL1UJG  noted this problem. I cannot 
comment on the effect of the existing BFO phase noise at this time.

73,

Geoff
GM4ESD

----------------------------------------------------------------------

Martin, OK1RR, wrote:


> Looking at the Rob Sherwood's receiver data table
>
> http://www.sherweng.com/table.html
>
> I discovered a remarkable fact: the K2's weakest point is the FILTER 
> ULTIMATE REJECTION. Sherwood says it is 80 dB only (and there are two K2s) 
> and it was confirmed by one of our top grade contesters and DXpeditioners.
>
> A closer look onto schematic awake an idea to reuse KSB2 crystal filter 
> for receiving on both CW and SSB. No doubt the additional 7 crystals will 
> improve this figure significantly. I had no opportunity to track the 
> rather complex diode switching scheme but Sherwood's figures saying it 
> clearly – there are 5+2 crystals in CW receive path only. So my question: 
> is there any mod to improve the ultimate rejection using KSB2 crystal 
> filter? (of course, G3RXQ's filter mods are here but I don't plan to 
> destroy a KSB2 board and I am not interested in SSB). Would any 
> experienced K2 builder willing to play around this?
>
> And a QSK improvement to listen between elements up to 40 wpm would made 
> from the K2 a dream radio :-)
>
> I don't own my K2 yet but a 100 dB ultimate rejection would ensure me that 
> I should't lose my time with thinking if to buy or not :-)
>
> 73,
> Martin, OK1RR
>