Continue to Site

Welcome to EDAboard.com

Welcome to our site! EDAboard.com is an international Electronics Discussion Forum focused on EDA software, circuits, schematics, books, theory, papers, asic, pld, 8051, DSP, Network, RF, Analog Design, PCB, Service Manuals... and a whole lot more! To participate you need to register. Registration is free. Click here to register now.

Register Log in

How to measure the input reactance and output reactance of a MOSFET/gain stage

Status
Not open for further replies.
C

circuitking

Full Member level 5
Joined
Jan 8, 2018
Messages
291
Helped
1
Reputation
2
Reaction score
1
Trophy points
18
Activity points
2,503
Hi, I found relative answers to the question but I want a little clarity.
https://www.edaboard.com/showthread...apacitance-and-output-capacitance-of-a-MOSFET
https://www.edaboard.com/showthread...-Z11-ZM-and-ZP-in-calculating-input-impedance
https://www.edaboard.com/showthread.php?168947-Plotting-inductance-from-Z(im)-Cadence-calculator
https://www.edaboard.com/showthread...-plot-of-L-vs-freq-of-an-inductor-(SpectreRF)

I mentioned it as reactance as a general case but I know it is mostly capacitance. How do I find input/output capacitance of a gain stage/MOSFET/Network?. Usually it is the imaginary part of the total input/output impedance. So If I do SP simulation and plot the imaginary part of ZP from direct plot, It will give me the input/output capacitance of a gain stage. Correct? But why do some people use YP from direct plot for this purpose. I just want to know which parameter to use (Say ZP) and what modifications I should make if I use the other parameters(Say YP).

P.S I know if the network has series connection it is easier to use Z parameters and If it is parallel connections it is easier to use Y parameters. But If we consider the network as a block box we wouldn't know how the connections are inside.
 

circuitslave said:
So If I do SP simulation
Click to expand...
You should not use S-parameters, since you can not understand RF.

circuitslave said:
and plot the imaginary part of ZP from direct plot,
It will give me the input/output capacitance of a gain stage.
Correct?
Click to expand...
Open state.

circuitslave said:
But why do some people use YP from direct plot for this purpose.
Click to expand...
Short state.

Consider a definition of Z11 and Y11.

circuitslave said:
I just want to know which parameter to use (Say ZP)
and what modifications I should make if I use the other parameters(Say YP).
Click to expand...
Surely learn a definition of Z and Y parameters.

circuitslave said:
But If we consider the network as a block box we wouldn't know how the connections are inside.
Click to expand...
It is very easy linear algebra whatever you use as 2-port parameters.
 
Last edited:

circuitking said:
Hi, I found relative answers to the question but I want a little clarity.
https://www.edaboard.com/showthread...apacitance-and-output-capacitance-of-a-MOSFET
https://www.edaboard.com/showthread...-Z11-ZM-and-ZP-in-calculating-input-impedance
https://www.edaboard.com/showthread.php?168947-Plotting-inductance-from-Z(im)-Cadence-calculator
https://www.edaboard.com/showthread...-plot-of-L-vs-freq-of-an-inductor-(SpectreRF)

I mentioned it as reactance as a general case but I know it is mostly capacitance. How do I find input/output capacitance of a gain stage/MOSFET/Network?. Usually it is the imaginary part of the total input/output impedance. So If I do SP simulation and plot the imaginary part of ZP from direct plot
Click to expand...
No.. If you are interested in the input impedance, look at ZM1 and for output impedance, look at ZM2. Not Z parameters. V1 = Z11*I1 + Z12*I2. If you only look at Z11, then you don't see the effect of Z12 which is the effect of the feedforward capacitance (aka Cgd). ZM1 captures both of them.
 

Hi,

In my eyes the miller effect makes the system unlinear, you see this in the voltage plateau of V_gs.
Thus you can't apply S parameters.

You need to input power to get the gate charged. ... but you don't get the whole power back, because all the power for the miller effect is dissipated within the Mosfet.
It is a lossy unlinear system, depending on V_DS...and can't be simulated with an R C circuit.

Klaus
 

vivekroy said:
No.. If you are interested in the input impedance, look at ZM1 and for output impedance, look at ZM2. Not Z parameters. V1 = Z11*I1 + Z12*I2. If you only look at Z11, then you don't see the effect of Z12 which is the effect of the feedforward capacitance (aka Cgd). ZM1 captures both of them.
Click to expand...

Yeah sure.
I used this
1.CinfromY=(1 / (abs(imag((1 / ypm('sp 1 1)))) * 6.28 * xval(ypm('sp 1 1))))
2.Cin=(abs(imag(zm(1 ?result "sp"))) / (2 * 3.14 * xval(zm(1 ?result "sp"))))

There is a lot of difference between two of the results and using 2. the Cin value is not constant over the frequency range. That is the my confusion as to which one to use and the Cgd effect would so much ?

Cin.JPG
 

circuitking said:
Yeah sure.
I used this
1.CinfromY=(1 / (abs(imag((1 / ypm('sp 1 1)))) * 6.28 * xval(ypm('sp 1 1))))
2.Cin=(abs(imag(zm(1 ?result "sp"))) / (2 * 3.14 * xval(zm(1 ?result "sp"))))

There is a lot of difference between two of the results and using 2. the Cin value is not constant over the frequency range. That is the my confusion as to which one to use and the Cgd effect would so much ?

View attachment 154599
Click to expand...

The expression for Cin is incorrect.
Xc = abs(1/imag(1/zm(1 ?result "sp")) )
Cin = 1 / (Xc*(2*3.14159*xval(zm(1 ?result "sp"))))

Xc = 1/jwC ==> C = 1/(j*w*Xc)

This Cin is Cgs + Av*Cgd


You can check this value at very low frequencies. Your Cin is in nF which is really high.
 

circuitslave said:
Yeah sure.
Click to expand...
Are you truely sure ?

circuitslave said:
1.CinfromY=(1 / (abs(imag((1 / ypm('sp 1 1)))) * 6.28 * xval(ypm('sp 1 1))))
Click to expand...
Very wrong. Can you understand imag(1/admittance) ?

circuitslave said:
2.Cin=(abs(imag(zm(1 ?result "sp"))) / (2 * 3.14 * xval(zm(1 ?result "sp"))))
Click to expand...
Very wrong. Can you understand imag(impedance) ?

circuitslave said:
There is a lot of difference between two of the results and using 2.
Click to expand...
It is a very natural result.

circuitslave said:
the Cin value is not constant over the frequency range.
Click to expand...
It is a very natural result.

circuitslave said:
That is the my confusion as to which one to use and the Cgd effect would so much ?
Click to expand...
Both are very very very very wrong.
Simply you can not understand impedance and admittance at all.

Surely learn a definition of Z and Y parameters.

Y11 is a short state admittance.
Z11 is a open state impedance
ZM1 is a loaded impedance where load impedance value is reference impedance.

vivekroy said:
If you only look at Z11,
then you don't see the effect of Z12 which is the effect of the feedforward capacitance (aka Cgd).
Click to expand...
Wrong.

Z11 surely includes the effect of Cgd due to intrinsic load.

vivekroy said:
The expression for Cin is incorrect.
Xc = abs(1/imag(1/zm(1 ?result "sp")) )
Cin = 1 / (Xc*(2*3.14159*xval(zm(1 ?result "sp"))))
Click to expand...
I don't think Yin=1/zm(1 ?result "sp") is not appropriate, since circuitslave can not understand S-parameters at all.

I don't think reference impedance is matched to actual load resistance.

circuitslave said:
So If I do SP simulation
Click to expand...
Show me reference impedance and load impedance.
 
Last edited:

Z11 surely includes the effect of Cgd due to intrinsic load.
Click to expand...

Yes.. I see your point. I understand my mistake. Y11 = jw(Cgs+Cgd) (because the output would be shorted and both capacitances would come in parallel). Z11 expression will be longer but it will include the effect of Cgd. I apologize for misleading.
 

Status
Not open for further replies.

Similar threads

Part and Inventory Search

Welcome to EDABoard.com

Sponsor

Back
Top