Thanks Jo and Anders for replying...
Yes after having seen your two posts, i was also changing my opinion from "it is completely borked" to "probably correctly assembled" and just toroid permeability / capacitor tolerances and winding style differences to blame for the strange things i see.
Let me try to attach the screenshots this time. Maybe that works better than inline pictures.
I'd still like to fix the 20m RF filter sweep.
All the others traces are fairly close to the pictures in the manual "shape wise", but i am sitting on the high end of the filter curve, so closer to the roll-off, which might reduce HF output, but 20m is pretty weird.
I am sure it is correctly assembled. Correct capacitors, correct number of turns on all toroids. Tightly wound and evenly spaced. So possible to increase inductivity by squeezing, but not lower it.
Therefore i haven't put it in it's case yet.
4W @ 12V output on all bands would be quite nice, but it seems the highband variant is a bit fiddly in that respect.
73 Tobias DL3MHT