Clarification on phase shift and S-parameters, phase shift of DRO and transistor.

first briefly:
1. Transistor phase shift is always 180deg, or it is equal to transistors S21 phase shift? Or it is different things?
2. How much is phase shift of DR (in parallel feedback DRO)?
3. How to connect two FET gates with 180deg microstrip, it would be simple 180deg microstrip or (180 - angle(S11))deg?

indepth:
Transistor amplifier provides 180deg phase shift (S21 magnitude=1 angle=180, or S11=-1?). But from S-parameters, if transistor have S21 coefficient magnitude=2 angle=50, does that mean phase shift is actually 50deg and we must forget about that 180deg?

In Avago's "Application Note S005" we can see parallel feedback oscillator with DRO based on MSA-0870 amplifier. On the fig.4 there is miscrostrip scheme with lenghts. Question is about:
I understand, that signal must come in-phase for oscillation, but why DR and gap between Dielectric Resonator and microstrip lines does not exists in this formula? I think it must be like:

The 180° phase shift through the DR provides the correct current direction for the oscillator.

Click to expand...

if this "angle(S21) + 2*L3 + 2*L3 = 360deg at LO" povides 360deg phase shift, then how oscillator will even work, if DR adds 180deg, making signal in anti-phase?

The last question i asked here before, still trying to understand:
If i have two equal active devices and want connect their S11 port with 180deg phase shift microstrip, how to do it correctly? For example, S11 angle is -15deg. Does it mean that line lenght must be 180+15deg or it still be 180deg? (for example, push-push oscillator with FET gates connected with 180deg line case. Is it 180 or corrected according to S11?).

still have this question

Terminator3 said:

still have this question
1. Transistor phase shift is always 180deg, or it is equal to transistors S21 phase shift? Or it is different things?

Click to expand...

No, transistor circulate phase all time, especially FET at hi frequency above 10-15GHz it turn around easily up to 360 and more degree . For low frequencies BJT they approximate with h hybrid, Y matrix and here are basically simple RC models. And as result phase shift should be within 90* but really more and as result it can generate with complex load as IF stages at simple receivers.

Terminator3 said:

2. How much is phase shift of DR (in parallel feedback DRO)?

Click to expand...

from -180 to +180

So what does they mean by saying "angle(S21) + 2*L3 + 2*L3 = 360deg at LO", that DRO phase shift is 0 deg?

I prefer something like that. And it is on the right way.