LNA circuit not working

[BigBoss]

promach..

Your MOS Transistors have L=150um, W=0.18um and you cannot never make those transistors into saturation mode so you will get a never positive gain..!!Therefore Ids is around nA level..
Why don't you want to understand ? You don't know the fundamentals of MOS transistors but you try to design an LNA @ 2.4GHz..I suggest you to learn the fundamentals of MOS transistors, how they work, how they are biased, what is saturation mode what is linear mode ...You cannot never learn about'em by asking many question here, never..
Engineering starts to read-write, then read-write, then read-write again.. When you have a concrete fundamental, you may start to play with simulators.Otherwise you effort cannot be more than digging a deep hole with a needle..

 

[promach]

See my previous post screenshot.

I have already realized I mixed up W and L.

Your comment came after I have solved this problem of W, L mixed up

- - - Updated - - -

I have modified the circuit to have a positive gain of 6.

But the gain is not enough.

Any advise ?

 

[volker@muehlhaus]

What's the idea to have 1pF value for the DC-blocks? That's quite a large impedance at 2.5GHz.

 

[frankrose]

Great! I am sure it helped a lot, however you still run transient analysis. Run AC to find the resonance frequency, with 2.25GHz and transient analysis you won't excite the LC tank on the exact resonance frequency. Parasitics of the transistors will shift the resonance.
I don't get what is the purpose of C3, seems bad idea. Lbias is still very low. Do all recommendations above, then we can go further.

 

[promach]

@frankrose

I am doing this at 2.4GHz

See the following snippet about L1 and C1 parallel network. I am not sure about C3 though.

 

[promach]

What I do not understand is the strange AC response in the previous post just above with a sharp magnitude dip and abrupt 180 degrees phase shift at 2.9GHz


The transient response looks normal though.

 

[FvM]

What I do not understand is the strange AC response in the previous post just above with a sharp magnitude dip and abrupt 180 degrees phase shift at 2.9GHz

The behavior is just by design. L1/C1 is creating a real transmission zero. I guess you did neither fully understand the purpose of series L respectively the parallel LC circuit nor its limitations.

It's a small band matching circuit, the wanted effect is the peak at 2.4 GHz, the dip at 2.9 GHz is its side effect.

 

[promach]

The behavior is just by design. L1/C1 is creating a real transmission zero. I guess you did neither fully understand the purpose of series L respectively the parallel LC circuit nor its limitations.

It's a small band matching circuit, the wanted effect is the peak at 2.4 GHz, the dip at 2.9 GHz is its side effect.

I could not find such narrowband matching circuit topology from section 3.4.3 of book RF circuit design 2nd edition by Richard LI

- - - Updated - - -

Ok, I see it now.

L1 and C1 make an equivalent inductor using less inductance for the purpose of RFIC implementation

As for the narrowband matching circuit, see the screenshot below ( taken from section 3.3.1 of the same book )

How is the dip at 2.9 GHz its side effect ?

 

[FvM]

Calculate L1/C1 resonance frequency...

 

[promach]

I have put everything under https://github.com/promach/LNA

Any advices or comments on the TODO list within the github README ?

 

[frankrose]

....noise analysis?

 

[promach]

@frankrose

Any comment on the .noise analysis simulation result below ?

 

[frankrose]

Hmm... nice curves....
If it is an IC LNA be cautious with the output termination, it is not necessary to use 50 Ohm resistor.

 

[promach]

Are you implying that I should not include any load matching resistor RL at all for all type of circuit spice analysis (.tran, .ac and .noise) ?