Using an LNB as an IF and feeding an LNB


Gareth G4XAT
 

Could anyone advise me on the following please?
I'm hoping to receive 3.4GHz DATV using a Titanium C band LNB (3.4-4.2GHz).
I have a 4x4 flat plate antenna (ex WiMax link) which I have added a SMA connector to. I want to feed this signal INTO the round waveguide of the LNB, which has at present two probes, at 90 degrees to each other, set at approximately 1/4 and 1/2 wavelength from the back of the waveguide cavity (60mm ID). Could I fit a third probe at 3/4 wavelength into the LNB side, aligned with one or the other of the probes?
I also have a small Log Periodic yagi (6dBi nominal 2-10 GHz) which I fitted to a BSB small dish (not the squarial) in the same place as its original LNB. Using this sideways (greater dish capture) shows it has a little more gain than the 4x4 flat plate.
Both of these solutions provide for easier mounting than the large C band LNB on a suitable dish but would I hope offer similar/more gain than the LNB on a small dish.
Which leads me to my next idea....if this works, I hope to feed the IF from a mixer (via a pipe cap filter on 3.4) from a 5.66GHz flat plate (Yagi.pl 24HZ). My intention is to use the gain and NF of the C band LNB to down-convert further within the range of my DATV receiver (144-2.45GHz so plenty of options).
I have built a cantenna for 3.4 that fits in the end of the C band LNB, but there are clearly some interesting things going on as signals vary wildly on rotation (somewhat expected) but also when I adjust the (movable) back-plate of the cantenna. So I'm hoping the concept of adding a third 1/4 wave on a sma will work.
Guidance sought ...thanks

Gareth G4XAT