Non-Linear HEMT model

[abhijitrc]

What is the role of Dgd ( diode at gate and drain ) in large signal model ?

Since in HEMT/MESFET the equivalent lumped circuit model shows two diodes , one at gate-source side and another at drain-source side!,since both are diodes how the device amplification takes place....

1. S-parameter extraction of the device is showing positive value of S21 at some bias points, but as we know diodes are passive, how we are getting appreciable gain,ie, 20logS21 shows voltage gain ????????, plz help me to sort out the problem.

 

[abhijitrc]

a.What is the difference in the topology of minimum noise matching and power gain matching/conjugate matching ?
b. For an amplifier to drive the succeeding stages (driver amplifier), we require at least 5.0dBm output power from the amplifier,( currently we are using CFY25,as a device for design ), vendor/manufacturer supplied the s-parameter extracted at the power level -20.0dBm, can we use this s-parameter?, or we have to extract large-signal s-parameter? or Nonlinear model is a better option?
c. Can we go for the conjugate matching or power gain matching for this particular design?
d. We want to operate the circuit at P1dB region, can we match at that power level only??

 

[tony_lth]

a) Min noise matching is matched with gamma-opt, but conjugate matching is for max power, due to gamma-opt is different from conjugate point.
b) -20dBm is input power or output power? If input, in my view, you can use the S-params.
c) for driver amlifier, conjugate match is ok
d) It depends on your system, except power, such as harmonics, signal distortion, etc. What is your modulation plan?

 

[abhijitrc]

a) Min noise matching is matched with gamma-opt, but conjugate matching is for max power, due to gamma-opt is different from conjugate point.
b) -20dBm is input power or output power? If input, in my view, you can use the S-params.
c) for driver amlifier, conjugate match is ok
d) It depends on your system, except power, such as harmonics, signal distortion, etc. What is your modulation plan?

i think concept of gamma-opt in noise matching is that generator impedance should match with the device impedance so that noise power should not reflected back and it should not be correlated. sir what i think if the device impedance is matched with the load and source impedance say ~50 ohm then only we can achieved maximum power, then how gamma-opt is different from conjugate matching? plz guide

Apart from that large signal s-parameters are power dependent since we require 5.0dBm output power from the amplifier,s-parameter extracted at -20dBm can be used ,if we increase the input power level there is a change in bias point wrt i/p signal then what will be the topology ??

Can we use conjugate matching at the input for mimimum noise matching ????, Can anybody explain the logic behind not using conjugate matching for minimum noise matching!!!!!

 

[afz23]

Can we use conjugate matching at the input for mimimum noise matching ????, Can anybody explain the logic behind not using conjugate matching for minimum noise matching!!!!!

I think you are getting it all wrong here.
Low noise devices are characterized by Γopt provided
by manufacturer.

If you match such that Γs=Γopt ,please note matching should be
equal to Γopt not Γ*opt,you will get min.NF.

But if you make Γs=s11*,you get maximum gain,but
problem is they don't coincide with min. NF .

So for LNA DESIGN designers prefer gamma opt
matching over max gain.

 

[abhijitrc]

I think you are getting it all wrong here.
Low noise devices are characterized by Γopt provided
by manufacturer.

If you match such that Γs=Γopt ,please note matching should be
equal to Γopt not Γ*opt,you will get min.NF.

But if you make Γs=s11*,you get maximum gain,but
problem is they don't coincide with min. NF .

So for LNA DESIGN designers prefer gamma opt
matching over max gain.

Dear afzal sir,

That i know that noise matching is gamma s=gamma opt and it is provide by manufacturer, i just want to know how the noise behavior will change if it is conjugately matched??? and why noise figure will be poor?