power amplifier question [hlp]
suppose that a rf power amplifier is fed with 6 signal with frequencies:
2250MHz , 2350MHz , 2450MHz , 2300MHz , 2400MHz , 2500MHz
what can you say about frequency of output signals from the power amp? What will be the spurious signals? What parameter determines the spurious signals?
suppose that the power amp frequency range is 2-4Ghz and that it is operating in full power.
the equation that gives you the two tone shoul give you all of the cross products provided that you derivate all of them, but there is got to be a better solution. can sombody point to this for this guy and me as well.
I don't think there is a simpler way to do it than all the cross products of the tones. This just tells you where they are but not the amplitudes. Just pug all the tones into a simulator through a nonlinear element and see where they fall; set the harmonic order high enough.
As far as where they come from it depends on the nonlinearities in the device. For instance the calssic JFET transconductance curve is IDS=IDSS/Vp^2 * (Vg-Vp)^2 and this will only generate a second harmonic with no intermods since the third derivative is zero, hence the third order intermods are zero. Now multiply this by other functions such as (1+lambda*VDS) for output conductance and tanh(alpha*VDS) for knee region and you get 5th, 7th, etc order of intermods. Hit breakdown of junctions and you get very sharp nonlinearities which cause more.
Since the frequencies are evenly spaced by 50 MHz, the IMD tones will also be spaced by multiples of 50 MHz. Some will fall on top of the desired signals and some out of band.
The PA will output all the frequencies in general as specified. The signals are 50MHz apart so it will definitely have all the derivatives of these signals.
This PA will work as multicarrier system and in that case if its not backoff with the proper system calculated margin it will generate all the intermods and that will interfere with the desired frequencies.
I suppose to test PA linearity you don't have to input so many signals only two tone tests are sufficient.