Thank you Allison for the clarification. I read the thread about the mixing problem and understand where the problem is.
Maybe I will try to play around with bandpass filters, but this is no priority - output is not everything, I would rather have a clean signal.
By the way, I checked with my calibration station - and the RX seemed to be okay as soon as I fed a not changing/fading-signal.
Maybe it is / was a problem related to the AGC or a combination of both.
I've just another small question which I hope you will not crucify me for:
Is it possible that changing R23 is abolutly necessary, if Q20 is swapped to a BFR106?
I had not checked the transmitter with modulation yesterday as I only tested CW which was good. Today i connected a microsphone and noticed
that here was nearly no output at all with SSB modulation anymore. So I also changed R27 to 39 ohm but that did not doing anything at all.
R23 I thought would not be really important as it was only changed from 10 to 18 Ohm (okay, in percentage it is a lot) and i thought it would be pointless.
So I changed all BFR106s back to 3904s. Q20 was the transistor which I swapped last - and as soon as that one was in again modulation went up again.
So was this really an operating point problem? If not maybe my transistors are really of bad quality?