LNA at 2 GHz using ADS - need feedback, help

[mw_rookie]

ads+lna+feedback

I am designing an LNA at 2 GHz and the S-parameters are:
S11 = 0.972 ang(-21)
S21 = 4.436 ang(153.9)
S12= 0.026 ang(73.8)
S22 = 0.612 ang(-16.6)

and K<1 so I calculated Γs and ΓL and also got the input and out stability circles using smith chart utility in ADS. But, both Γs and ΓL are almost on the circle and its difficult to judge whether they lie in the stable or unstable region. the question is since K<1 shouldnt there be a value of Γs that lies in the stable region and one in the unstable region rather than both in the same region and likewise for ΓL and the worst part is the magnitude of the relection co-eff.s are about 1 in magnitude using these S-parameters though the angles vary...can anybody help??? thanx

 

[Canosa]

feedback lna using ads

You have to use techniques to improve stability of your LNA. To get k>1 at 2 GHz and out of that. It's alway good to improve k outside the working frequency because you get a minor chance to have your LNA oscillating in other frequencies.
You can use some techniques to do that:

1)Series feedback: inductor at source (source leads) improve stability but if is large can cause peak gain at high frequencies so u can have the device unstable at high frequencies.

2)Parallel feedback: you can use a resistor as feedback element between gate and drain portion of the device. The NF can be increase by a litle bit but in that way it's possible to achieve k>1 for almost all frequencies.

3) Resistor at the drain portion of the device can help to improve estability at all frequencies but it's value has to be small to do not kill the gain of amplifier.

The combination of 1) and 3) works fine!

Parallel feedback is good too, but if can pay by some figure to achieve estability this is the best way...

Best Regards

 

[kspalla]

lna oscillating

k is no more good to test the stability. µ is the trend.