Liam Brocvielle

Well, in our courses , we slightly used (like 1 hour or 2 hours in april) the National Instrument AWR software. It seems that it does the Harmonic Balance simulation quite well, the load pull simulation and for the transistor model, I have not idea yet if it is a solid model or not. I guess it is. I have a student license for that software, so I think I'll dive into the AWR world and discover it. It seems that there is also some example projects concerning power amplifiers. Thanks !

Hi Andres, Thanks a lot for your useful tips. I was so far to properly understandthe implications of desiging a power amplifier. I get more in depth into their world. Until now, I never learned properly the methodology to design such an amplifier. As I can see, the gain optimization becomes meaningless when designing PA. Instead, the loadpull (as you mentionned in your first message) is very frequently used. The goal, as I understand it, is to find what is the best load impedance to put in order...

Hi, thank you for your answer. Computing the u parameter (unilaterality criterion), it is equal to 0.150, and so an error of around 50%, I think. So not really unilateral. Well, to calculate the conjugate, we used this formula in the course : rhoIN = S11 + (S12S21rhoL/(1-S22rhoL) ) = rhoS leading to ZS = ((1+rhoIN) / (1-rhoIN) ) rhoOUT = S22 + (S12S21rhoS/(1-S11rhoS) ) = rhoL leading to ZL = ((1+rhoOUT) / (1-rhoOUT) ) And so as I can say, is that if we assume the device to be unilateral, S12 = 0,...

Hi all, I'm trying to design a power amplifier at 1.9GHz. I made the 1st stage. DC bias has been set to have a collector current of about 10 mA. The simulation get me some S-Parameters, and as S11 was bad and S21 as well, adaptation was required. (as well, I have put a stabilization resistor, as K factor was <1). I choose a quarter wavelength adaptation. The substrate is a classical FR4, and the guided wavelength I have calulated is equal to 74.4 mm. I have attached the circuit's schematic and its...

Great ! I quite understood that if we want to achieve a MAG adaptation, a noise adapatation is not possible. and MAG adaptation is not possible if we do a noise adaptation. (optimum impedance has to be at the input and be the conjugate of the source impedance in this latter case.) Noise adaptation is an important thing when designing LNA, for example. Let's say I have 2 bipolar transistors. 1st one is common emitter (called A) and 2nd one a common collector (called B ) .Let's say as well that after...

Thanks for this really useful tip and the pdf ressource ! So it is stable, but it may have a great return loss, and no transmission ? Key is good power transfer. Yesterday I went more in depht about these impedances adaptations, and as I understood, the goal is to find, for example for a given source impedance, which value of load impedance give the maximum avaiable gain (also called MAG). For example, I took an example from the courses I had. We saw that for a 50 Ohms source, the load's impedance...

Thanks a lot Andres ! It was a scale problem,as you said it. I have attached a picture to show you the rescaled Smith chart. I remember this shape of drawing when talking about stability of devices in courses. As you said as well, we can clearly see that all the circles are out of the Smith chart (for the source here ! ) meaning that any impedance value can be choosen for the source. As I had only a few courses at school about devices in high frequencies, I have a few questions : -In case we take...

Hello, Thanks for your fast answer ! Here is the circuit's schematic. It is a order 3 Chebyshev bandpass filter created with coupled microstrip lines, it aims to be centered around 1.9GHz. Its cutoff frequencies are located at 1.85GHz and 1.95GHz, with bandpass ripple of 0.5 dB.