LNA Stability due to Wide Impedance Differences

[infinite_gbps]

So I had a narrowband antenna, Q approx equal to 20 and my frequency is 300 MHz. My impedance at 300MHz is 60.95 - j45.75. BUT, my impedance changes significantly when I move off 300MHz.

My issue is that the LNA will see high kOhms when I open a switch on the antenna (receive antenna).

What should I be concerned about when it comes to LNA stability (or lack there of) when the antenna is open (due to switch) and the impedance is high around that frequency.

Can I do anything about it? I was thinking about plotting K-factor of the amplifier with the s2p files by terminating the output with 50Ohms but putting S11 of the antenna when it is open and see how the k-factor changes with frequency.

This makes sense. Would like to do it in ADS but I have to setup the analysis properly.

Any feedback would be well appreciated.

 

[G0HZU]

Use an absorptive switch?

 

[infinite_gbps]

Use an absorptive switch?

I added some stuff to the first post.

I wish I could use a absorptive switch. So we need really fast switching speed and the switch we have has <1ns switch time. Bad news is that the document that I have has no information on it being reflective or absorptive. Just looking at it I am assuming that it is reflective. Until they get back to me then I won't know.

So if it is reflective then .....?

---------- Post added at 16:14 ---------- Previous post was at 15:43 ----------

Question about the LNA, if the LNA is unconditionally stable (k-factor greater than 1 from 0-20GHz) then I don't have to worry about a reflective switch, correct?

For an LNA, how low does S12 have to be to have a stability? Esspecially if the LNA has a NF < 0.5 dB.

Thanks.

 

[Joel-Tang]

Using an absorptive switch will only help when switched away from the LNA, not when you are connected to the narrow band antenna.
Definitely plot the K-factor, if that is above 1, then you have verified stability in all source/load conditions. Your goal should be stability at all frequencies at any source impedance. This will inevitably take a toll on the noise figure, but this will ensure stability. You could make your life more complicated by characterizing the impedances you expect to be presented to the input of the LNA in all circumstances and design for stability in these regions of the smith chart. Plotting the input and output stability circles on a smith chart is needed in this process, but nothing beats an unconditionally stable LNA when you're faced with the concerns you've described.