Problem with ADS simulation on LNA

[stevie-g]

Hi guys!
Got a problem here.
My simulation result for S11 is ~-0.1dB, but S22 seems normal with -18dB. The gain is nice too, 20dB.
But in actual, the demoboard shows that S11 should get ~ -8dB.
What's wrong with my simulation? It's just a simple S-Param simulation and the Ids seems okay.
Oh ya, I'm measure in 2.5GHz.

 

[BigBoss]

Have you simulated your board in Momentum ???The difference might come from this..

 

[stevie-g]

I'm measure in around 2.5GHz.
Do you think microstrip line has major effect on my S11?

 

[vfone]

Check if the LNA bias current is the same as in the S-parameters file used in the simulation.

Also for better accuracy, the simulation shall contain all the parasitics of the components and S-parameters of each passive component. For example SRF of C and L could dramatically affect a high frequency simulation.

 

[rfreak]

0.1dB vs 8dB is a big different....please check & verify the input side connection...there should be a significant mistakes (i.e...open, short)

 

[stevie-g]

Hi vfone,
The bias current is the same as what I get in actual board. If components is a matter, why I can get S22, S21, S12 but not S11?

Check if the LNA bias current is the same as in the S-parameters file used in the simulation.

Also for better accuracy, the simulation shall contain all the parasitics of the components and S-parameters of each passive component. For example SRF of C and L could dramatically affect a high frequency simulation.

Added after 1 minutes:

Hi rfreak,
If the input side got mistakes, I won't get 20dB gain, right?

0.1dB vs 8dB is a big different....please check & verify the input side connection...there should be a significant mistakes (i.e...open, short)

 

[vfone]

I agree, if you have a short or open at the input, the gain (or S21) will be low.
But S11 could be different from the simulation when other S-parameters are relative OK, and the reason could be real-life parasitics, losess, or resonances of components, which may not be taken in consideration by the simulation.
Also sometimes the reason could be a bad S-parameters model released by the manufacturer.

 

[stevie-g]

I'm using components from the same manufacturer. If you say got parasitics or losses problem, why it didn't affect S22?
Sorry, I'm still no idea what happening here...

I agree, if you have a short or open at the input, the gain (or S21) will be low.
But S11 could be different from the simulation when other S-parameters are relative OK, and the reason could be real-life parasitics, losess, or resonances of components, which may not be taken in consideration by the simulation.
Also sometimes the reason could be a bad S-parameters model released by the manufacturer.

Added after 1 hours 13 minutes:

Guys! I know what's the problem. It's just because of my common source MOSFET problem.
Thanks for help!