[Bero]

I'm sad because didn’t saw these thread earlier.
I've made the same project from Paul Stennings page a few years ago.

And the first results was like in your case - QUOTE "When the varcap is fully meshed I get a beautiful sine wave in the oscilloscope but when I rotate the varcap the amplitude decreases proportionally and somewhere after halfway the oscillation stops "
I think FET isn’t critical, you can use on this position anything suitable (BF244, BF245, BF247 - I've try to use all of them and thing oscillate as good as any of them), careful on pin configuration, and thing must work.
Like "skywave" suggest is, that the types of coils you are using are unsuitable.
I have the same results like's yours - after some point oscillation stops.
Probably you have used coils, which is intended to use in circuits for prevent "wild" oscillations.
Much better results will given by using coils from a scrapped set. Indeed, some of then you can't find, and adder you must wind up by yourself.
I've do it that way, and this part works fine.

But in this project - there are much more troubles then your first trouble is.

Next trouble you will find in interaction between coils.
That's good described in Paul's text. There is all manner of interaction inside RF devices.
I've solve this trouble with an elegant trick.
Instead using 1x12 position switch, I've used 2x12 position switch.
Do it that way, you can to effect that unused coils are in the air - there are not connected anywhere.
You can saw picture like I've sort elements and coils.
My unit cover a frequency range of 150 kHz to 30 MHz over 6 ranges.




But worst trouble which you would reached is mistake into Paul's schematic.
I've never solve that part to the ends, my thought is that Paul due an error when was redraw a circuit diagram. If he read this, maybe he can suggest something.
As you can read on introduction page, the output is not a pure sine wave, so the unit may not be suited for more exacting electronic development work.
Using unit for aligning old radios this won't be a trouble.
But stability will be very important thing, otherwise the unit will not be useful at all.
In my case, TR2 cause instability, and something around them isn’t good.
When you turn attenuator (VR1-level pot. into diagram), to any side, like you changing the level, oscillator frequency changing too.
My first thought was that level changing, cause change to the oscillator voltage, and that could be reason for frequency instability - but did not.
At finally, I've reached some improvements changing schematics this way.
(above is an original, lines marked red color are changed)