|
bartholomeus87 Post subject: Blf177 amplifier 100mhz Stripline Posted:
Sat Jan 17, 2009 12:05 pm Captain Joined: Sat Jan 17,
2009 11:47 am Posts: 6 Hello people, I have question
about this amplifier i would like to build. I found it on the web in
the elenos sf150 handbook. I am really stuck here because i can't figure
out how to recalculate the value of the stripline in the input matching
circuit. There is a partlist wich contains all values of the components
on the board exept this stripline. I looked at the datasheet of the
blf177 and the s-parameters are calculated with a bias current of 100ma,
wich is far more than on this board. So i don't exactly know what te
input impedance is in complex form. Re+-Im I can arrange a pc
with agilent Genesys software, but i dont know where to start. Who
can help me out? Thanks, Bart Top IR
Post subject: Posted: Mon Jan 19, 2009 3:18 pm Site Admin
Joined: Mon Jun 27, 2005 2:02 pm Posts: 373 Location:
Germany Hello Bart, I would do the following process:
1. I would use the S-paramaters of the transistor to see the
input impedance. 2. Add the impedance of the voltage divider
connected to the transistor's gate. 3. Simulate the impedance
of the matching network made of: L4, C10, C11 and C16. 3.
The stripline should basically transform the impedance seen by the transistor+voltage
divider to the impedance seen by the matching network made of the componenets
in step 3. Both impedance should be very close to each other as the
stripline will have a uniform impedance. 4. Once the impedance
required for the stripline is known, then you are able to calculate
its dimensions based on the substrate properties that you use for the
board. Top bartholomeus87 Post subject: voltage
dividerPosted: Tue Jan 20, 2009 5:08 pm Captain Joined:
Sat Jan 17, 2009 11:47 am Posts: 6 Thanks for the reply, but
i don't understand the voltage divider part. What is the function
of this circuit, and how do i add/compensate the impedance for it?
Bart Top IR Post subject: Posted: Wed
Jan 21, 2009 2:06 am Site Admin Joined: Mon Jun 27,
2005 2:02 pm Posts: 373 Location: Germany Since in this circuit
the Vgs is provided by Vcc and there is no separate voltage supply for
Vgs, a voltage divider is required to set the right value of Vgs for
the required DC operating point of the transistor. The resistor
connected between the gate to GND add their part to the total impedance
that would be seen from the right side of the stripline. Then you should
add them to the simulation to see what is the total impedance seen to
the right of the stripline. Top bartholomeus87
Post subject: Posted: Mon Jan 26, 2009 3:29 pm Captain
Joined: Sat Jan 17, 2009 11:47 am Posts: 6 Thanks for the
help, great forum with a lot of rf theory. Just a question, how
do you calculate te optimum bias voltage for a circuit like this? Or
is just a given value in the datasheet? Top IR
Post subject: Posted: Mon Jan 26, 2009 3:40 pm Site Admin
Joined: Mon Jun 27, 2005 2:02 pm Posts: 373 Location:
Germany Many thanks! I am happy that I can help! There
is no optimum bias for a circuit like this. You calculate the Vgs that
will give you the Ids, based on the gm (Transconductance) curves in
the data sheet. Per given (Vgs, Ids), there is a set of measured S-parameters
provided by the manufacturer. It is easy to calculate the bias
for a JFET transistor because of the high gate impedance, which actually
forms an ideal voltage divider in the gate, compared to a Bi-Polar transistor,
for which you would need a Thevenin transformation to be able to set
the base equivalent voltage and resistance. Top bartholomeus87
Post subject: Posted: Tue Jan 27, 2009 7:00 am Captain
Joined: Sat Jan 17, 2009 11:47 am Posts: 6 Ok, i think i'll
understand that. I've been playing around with Genesys the past
few days, and i figured out a few things: -I can simulate the network
of l4, c10, c11 and c16 so i know what impedance is fed into the stripline.
And graphs of the reflected power and power gain. -I know how to
calculate the dimensions of the stripline. -I know where the s-parameters
stand for. (s11,s21 enz.) But i can't figure out how to extract
the input impedance for the transistor from the s-parameters. Can you
explain how to do this step-by step? Thanks Top
IR Post subject: Posted: Tue Jan 27, 2009 3:32 pm
Site Admin Joined: Mon Jun 27, 2005 2:02 pm Posts:
373 Location: Germany Extracting the internal impedance of the
transistor would require a non-linear model of the transistor (Spice-based
model). For the purpose of your simulation such model is not
required and also the knowledge of the transistor's impedance is not
required, since I guess that in this transistor there is an internal
matching between the gate pin and the die, therefore there is a transformation
of impedances as part of the internal structure. Top
bartholomeus87 Post subject: Posted: Wed Jan 28, 2009 4:13 am
Captain Joined: Sat Jan 17, 2009 11:47 am Posts: 6
I have downloaded a spice model of the blf177 from the nxp website.
It is an s2p file. I saw that the Idq was 100ma, is that a problem?
I don't understand the last part, there's got to be some kind
of match for the transistor? Top IR Post subject:
Re: Blf177 amplifier 100mhz StriplinePosted: Thu Jan 29, 2009 2:37 pm
Site Admin Joined: Mon Jun 27, 2005 2:02 pm Posts:
373 Location: Germany Quote: I have downloaded a spice model
of the blf177 from the nxp website. It is an s2p file. S2p
file is not a Spice model but a linerar model. Spice model would include
the equivalent circuit of the transistor, i.e. internal capacitances,
inductances and other semiconductor physical related parameters.
NXP does not provide a Spice model for this device. If you are interested
in it, you have to contact them. Quote: I don't understand
the last part, there's got to be some kind of match for the transistor?
What I meant is that the impedance which can be derived from
the S-parameters is measured at the device pin. the S-parameters are
de-embedded ususally at the device pin, and this is not the impedance
of the transistor's die. Usually transistors which deliver higher power
levels are internally pre-matched, and therefore include an internal
matching network, which then has to be included as part of the Spice
model. _________________ Best regards, - IR
Top bartholomeus87 Post subject: Re: Blf177 amplifier
100mhz StriplinePosted: Thu Jan 29, 2009 5:17 pm Captain
Joined: Sat Jan 17, 2009 11:47 am Posts: 6 Right, thats clear
to me now. And what to do with the bias? Top IR
Post subject: Re: Blf177 amplifier 100mhz StriplinePosted: Fri Jan
30, 2009 5:05 pm Site Admin Joined: Mon Jun 27, 2005
2:02 pm Posts: 373 Location: Germany If you work with other
bias conditions, you can ask the manufacturer to provide you different
S-paramters for your bias point. I would refrain from working with different
bias than the one recommended by the device manufacturer. _________________
Best regards, - IR
Posted 11/12/2012
|