bfg425w.s2p

Hi all, I need some help from Aplac-lovers to finishing up an old lab I've left behind me years ago.
The goal is to design a narrowband 10Ghz amp using
Philips BFG425W Npn transistor.
It should be designed for maximum gain, and less than -10dB input and output reflection coefficients.

I know how to do it in theory, because I've read the Gonzales book so many times, but I've forgot my Aplac knowledge. And this is the tool one is expected to use...

I need to extract from the example-files the S-parameters at 10Ghz, given by the teacher and which I attached to this message.

I want to edit the file so that I get the S-parameters for the bias-condition
that gives the highest transconductance gm.
That is I also need to find out which bias-currents Ib and Ic gives me that...

Could some kind soul help me with that ? I remember that one could add a text-object and write a PRINT -statement, but one has be careful not adding too many spaces or something in the syntax. I remember it was horrible !


I also have questions about the coupling capacitors in the bias circuit, these are by the teacher set to 1 Farad ! Is it becuase the scope is to get
the "intrinsic" S-parameters of the transistor, that is that we want no interference at this stage from the world outside the transistor itself.

I am fully aware of that the coupling capacitors in the final circuit will be a part of the input and output matching circuits, and there values will be determined by the input and output matching networks. And since the frequency is so high I can use small valued capacitances and still have consider it as short circuit for ac.

I know how to do the rest, stability-,operating power gain-,availible power gain -circles.

Hoping for some help...

Kindest regards,
StoppTidigare

Are you sure that a "certain" gain is available at 10 GHz with this transistor ?
I don't think so..


Essentially it has 25GHz fT ? Additional to, the package is not very suitable for that purpose..
It's very difficult bud !

I downloaded the s parameter file from Philips and put it in a program I have. You can get about 5 dB of gain at 10 GHz and the transistor is unconditionally stable at this frequency.

Hi all what should I do ?
APLAC gives me the same s-parameters though I've tested to change the bias to the craziest values..

I tried (Vce=0.5V Ib=100uA),(Vce=0.5, Ib=50uA) and(Vce=10V,Ib=900uA)

I still get the same S-parameters:
# MHz S MA R 50
10000 0.8379711591 73.34472725 0.2462133656 -66.68669628 0.09987239032 -11.80558728 0.928956038 94.12799952
That is not physical isn't it !

And by the way flatulent, APLAC shows a decreasing K-value from 8-12Ghz starting at around 0.45.....So it shows that it is definitely not unconditional stable. But again its from APLAC....

What should I do friends ?

Below I've pasted in the *.i -file

Kindest regards,
StoppTidigare


AplacVar Ib=50u
AplacVar Vce=0.5
AplacVar dS
AplacVar k
AplacVar MAvG
AplacVar gm
AplacVar fT


#include "C:\\WINDOWS\\SKRIVBORD\\NEW FOLDER\\BFG425W_CHARACTERISTICS\\BFG425W.I"

BFG425W BFG425W1 Collector Base GND

Curr Ib GND Base
+ DC=Ib
Volt Vce node1 GND
+ DC=Vce
Short AmpereMeter node2 Collector
+ I Amperes
Cap Cin In Base
+ 1
Cap Cout node2 Out
+ 1
Ind RFC node2 node1
+ 1
DefNPort nport 2
+ In GND 50
+ Out GND 50
+ Store "BFG425W.s2p" MHz MA
Sweep "BFG425W.1"
+ DC
+ LOOP 7 AplacVar Ib LIN 0 700u
+ LOOP 501 AplacVar Vce LIN 0 10
+ Window 0
+ X "Vce" "V" 0 10
+ Y "Ic" "A" 0 50m
Show W=0 XY Vdc(Collector) Idc(Amperes)
EndSweep
Sweep "BFG425W.2"
+ DC
+ LOOP 100 AplacVar Ib LOG 10u 1m
+ Window 1
+ LOGX
+ X "Ib" "A" 10u 1m
+ Y "Ic" "A" 0 100m
Show W=1 XY Ib Idc(Amperes)
EndSweep
Sweep "Sweeping gm"
+ DC
+ LOOP 101 AplacVar Ib LOG 5u 1m
+ LOOP 1 FREQ LIN 10G 10G
+ Window 1
+ LOGX
+ X "Ig" "A" 1m 100m
+ Y "gm" "S" 0 0.26
Calc
gm=Mag(Iac(Amperes)/Vac(In))
EndCalc
Display
+W=1 XY "gm" Idc(Amperes) gm Marker=1
EndSweep
Sweep "Philips BFG425W"
+ LOOP 201 FREQ LOG 8G 12G
+ Window 5 FineSmith
Calc
dS=mag(S(1,1)*S(2,2)+S(1,2)*S(2,1))
k=((1+dS^2-mag(S(1,1))^2-mag(S(2,2))^2)/
+ (2*mag(S(1,2)*S(2,1))))
MAvG=(Mag(S(2,1)/S(1,2))*sqrt(abs(k^2-1)))
EndCalc
Display
+ Window 0
+ dB "S11" S(1,1)
+ dB "S12" S(1,2)
+ dB "S22" S(2,2)
+ Window 1
+ dB "S21" S(2,1)
+ Window 2
+ Y "H21" MagdB(h(2,1))
+ Window 3
+ Y "MSG" (MagdB(S(2,1)/S(1,2)))/2
+ Y "MAG" 10*(log10(MAvG))
+ Window 4
+ Y "K" k

EndSweep