Doubts regarding on linear, non linear simulation and layout generation in ADS

[raju_kambar]

Dear Sir,
I have taken Avago ATF54143 transistor to my design of low noise amplifier (LNA). I have the following doubts regarding on linear simulation, nonlinear simulation , and layout generation.

(1) For linear simulation, I have taken S-parameter model of the transistor and connected the input and output matching circuit . I have measured the S- Parameters (S11,S12, S21, and S22), gain , noise figure and stability of the low noise amplifier. I didn't use the DC- simulation.

(2) I used the nonlinear device model of ATF54143 ( which is downloaded from the Avago datasheet .Zap file)and connected the input and out matching circuit and the DC biasing model and measured the IIP3 ,OIP3, and P1 dB compression point.

(3) Whatever exactly similar input and output circuit I used for both linear and nonlinear circuit and simulated. I didn't get the similar results for linear and non linear simulation.

(4) For practical realizable, we must connect DC source to measure fabrication results.We must bias the non linear device.

(5) What is my prime doubt is, what ever the input and output matching and DC biasing circuit (Suppose DC biasing used Vds =3 v, Ids =60 mA)with Vds =3V and Ids =60 mA with MLIN'S ,MTEE'S ,CORN'S , MSABND, and MSOBND for NON LINEAR ckt. The same input and output matching and DC biasing made it zero (Vds =0 Volts) with MLIN'S , MTEE'S ,CORN'S,MSABND,and MSOBND. What ever I used for Nonlinear with biasing, same thing used for linear only difference DC biasing made it zero.
Above said methodology work it for layout generation and later for fabrication measurement results.For crosschecking the linear and nonlinear device model will it work properly.

(6) As we know, non linear device we are using same input and output matching what ever we used for linear device along with DC bias. But we must not use the DC -bias for the linear simulation. Without connecting the DC bias we can't practically realizable the low noise amplifier. How to cross check this non linear with biasing and input and output matching and how to consider this effect of DC -biasing in the linear circuit for practical realizable for layout generation and fabrication results.
I have attached word document files of both LINEAR and NONLINEAR files same input and output and biasing zero made for LINEAR and biasing 3 volts made it for NONLINEAR device.

Can you anybody clarify above queries.

 

[BigBoss]

You wrote too much things and I have confused..
You find ome differences between the simulations whic uses linear model and nonlinear model ??

 

[raju_kambar]

Sir,
There is no confusion in this. What ever I had the doubts, just expelled it out. Presently what ever the doubts I am facing, I have pointed it out.

 

[raju_kambar]

You wrote too much things and I have confused..
You find ome differences between the simulations whic uses linear model and nonlinear model ??

Sir,
Yes, I found difference in the simulation of linear and nonlinear circuit model with same circuit for both.

 

[BigBoss]

Sir,
Yes, I found difference in the simulation of linear and nonlinear circuit model with same circuit for both.

If the differences are smal, it's normal.Because nonlinear moel is "an extracted" equivalent circuit and it represents the circuit behaviour upto certain level.Their accuracy is totally depended on extraction method which has been used and model itself.
But linear model ( s-parameters) is measured one and they are valid under small signal driven conditions.That's why, DC or HB (large signal driven) simulations don't have any sense at all.They are just used to predict small signal response of the circuit such as Gain, Reflection, Noise etc.
But nevertheless the differences should be small and negligible.

 

[raju_kambar]

Dear Sir,
what ever the input and output matching and DC biasing circuit (Suppose DC biasing used Vds =3 v, Ids =60 mA)with Vds =3V and Ids =60 mA with MLIN'S ,MTEE'S ,CORN'S , MSABND, and MSOBND for NON LINEAR ckt. The same input and output matching and DC biasing made it zero (Vds =0 Volts) with MLIN'S , MTEE'S ,CORN'S,MSABND,and MSOBND. What ever I used for Nonlinear with biasing, same thing used for linear only difference DC biasing made it zero.
Above said methodology work it for layout generation and later for fabrication measurement results.For crosschecking the linear and nonlinear device model will it work properly.
What ever I said above, same thing I have done in the attached files of posting # 1 linear device circuit and nonlinear circuit with same input and output matching circuit. The only one difference in the two circuit is, Linear device ckt DC bias made it zero and nonlinear ckt DC bias is given.
What is my doubt is , will the above method is correct for linear device with circuit DC bias made it zero and nonlinear device with circuit DC bias is given.

 

[RealAEL]

During linear (s-parameter) simulation DC voltages, or currents, are not calculated. Any V_DC source is therefore not used in that simulation. According to the manuals when it is not used it is treated as a short circuit. The voltage parameter is therefore not relevant. It could be set to zero, 3V or 1000V, it does not matter as the simulation results will always be the same.

 

[raju_kambar]

RealAEL Sir,
Finally we have to take the nonlinear device model with input and output matching circuit and along with the DC biasing circuit along with MLIN'S, MTEE'S, MBEND, MCORN'S etc ....., as I have shown in the Nonlinear circuit with input and output matching circuit and DC biasing with 3 volts, which is shown in the posting #1 , and SMT Capacitors and SMD resistors.
Finally we have to generate the layout for this,and give it for fabricator manufacturer and latter placing the SMD components for the fabricated piece.
For cross checking this nonlinear layout generated circuit. We can connect same things to linear device , what ever we used in the layout generation for nonlinear circuit with input and output matching,DC biasing circuit along with MLIN'S, MTEE'S , MCORN'S, MBEND etc......... we can check the gain , S-parameters, noise figure and stability of the linear (s-parameter model)circuit.
Is it the correct way of cross checking for linear and nonlinear circuits, because nonlinear and linear along with same circuits will give the different results, may be closer results. which one I have to trust . Linear device along with ckt is used for measuring gain, S-parameter, noise figure , and stability. Nonlinear device with ckt is used for measuring IIP3,OIP3 , 1 dB compression point for simulation purpose. I need to consider parasitic effects of MLIN's MTEE'S, MCORNS, MBENDS, etc...... So that I have connected the what ever things used in the nonlinear ckt, same thing I used in the Linear ckt and simulated.
Is it both circuits must be same for linear and nonlinear.Can you clarify it.