The secondary of the IF is probably designed as an impedance match to the base of a standard BJT transistor. Which maybe has an input impedance of 5 to 10k. If you could some how directly couple from the resonant part of the IF to the gate of the following FET you could get more gain.
It does give more gain, almost too much, especially when another stage is added. Also I learned that any type of resistor in the gate adds noise, so decided to stick with the link coupling. I tried about every configuration this one seemed to give the best results overall. The yellow can transformers also have a third tap in the middle , which I didnt show in the schematic. I found no difference when I tied moved the collectors there instead of on the "hot" end, which I figured should also be a better match.
I built this circuit with three IF stages and the gain was crazy, and background noise went up dramatically from two stages. But the AGC action was phenomomenal. I guess it is normal to get a lot of background hiss when that much gain is used. I tried everything to quiet it down, but settled on just two stages to keep the speaker quiet.
An experiment with the single stage above would be to try adding two more IF transformers by coupling the low impedance links together and using the high impedance sides to drive the gates. Then bandwidth coud be adjusted by small coupling capacitors between the low impedance pins.
Here is the same circuit, but with added IF transformers and another 2N3904 for a meter driver. AGC is OK for only one IF stage.
https://vimeo.com/203763287
https://vimeo.com/204075926
The secondary of the IF is probably designed as an impedance match to the base of a standard BJT transistor. Which maybe has an input impedance of 5 to 10k. If you could some how directly couple from the resonant part of the IF to the gate of the following FET you could get more gain.
Sorry it is a file, Im trying to learn how to convert it to a picture.