I've been piddling with a little solid state rig for a while now. Both the LO and VFO are synthesized and stable, and the mixers work *fantastically*, with IP3 performance greater than +37dB on the bands of interest. (H-mode mixer)
I've been using a bidirectional IF strip, switching the oscillators and IF amplifiers to change the direction. It works, but the IF amps are not great, only really showing +11dB IP3 performance. Also I'm told that I only need one IF amp, and I have no AGC happing in the IF. Also, the crystal filter matching is not great.
To get the really stellar performance out of the mixers, I am using a diplexer to present a stable 50R load on the IF ports. The problem comes when it's time to do the impedance transform to feed the filter-I need to transform that 50R load to a 1000R+25pf filter.
Currently, I am using a two-transistor amplifier that I am not sure does the job properly. Hopefully the silly black background image that Eagle produces is visible.
Now, this IF amp is SUPPOSED to present a 50R input impedance, and a 1000R output impedance, and be relatively insensitive to reflected impedances. Gain shouldn't be high-I only really want about 8dB-10dB, just enough to overcome the mixer loss (4dB), and the crystal filter loss (roughly 2.5dB when matched properly).
Can anyone recommend a better method, or maybe a better IF amplifier? One that implements AGC?