Amplifier singularity
时间：04-04
整理：3721RD
点击：
Hi!
I am designing an oscillator and its amplifier has some interesting characteristics (picture enclosed). There are some frequencies where the Rollett factor has a local maximum, and at these frequencies the transistor does not amplify at all. (It is not a matched amplifier as it is to be used in an oscillator, it only contains a BJT, bias circuitry (radial stubs and resistors) and coupling capacitors at the input and output.)
I am familiar with the definition of the Rollett factor and it is clear why it would have a maxmimum where S21 has a minimum, but I would like to know why it happens. I have tried the simulation with a different transistor (but same manufacturer and package) and the results were similar (even the frequencies), so my guess would be that it is caused by package parasitics, however, I would like to see a second, third, fourth opinion. :)
The reason I'm curious is because I have built a VCO using this amplifier for 3GHz with a 10% tuning bandwidth and at some tuning voltages a spurious oscillation appears around the marked 880MHz frequency (and of course mixing products appear because of this spurious signal and the nominal frequency).
My wild guess would be that at these frequencies the parasitic inductances of the BJT and its package do not allow amplifying, but these inductances resonate with the feedback filter and that causes the spurious oscillations, but again, I'm not sure and the opinion of more experienced people cannot hurt. :)
Thanks in advance for any insights!
I am designing an oscillator and its amplifier has some interesting characteristics (picture enclosed). There are some frequencies where the Rollett factor has a local maximum, and at these frequencies the transistor does not amplify at all. (It is not a matched amplifier as it is to be used in an oscillator, it only contains a BJT, bias circuitry (radial stubs and resistors) and coupling capacitors at the input and output.)
I am familiar with the definition of the Rollett factor and it is clear why it would have a maxmimum where S21 has a minimum, but I would like to know why it happens. I have tried the simulation with a different transistor (but same manufacturer and package) and the results were similar (even the frequencies), so my guess would be that it is caused by package parasitics, however, I would like to see a second, third, fourth opinion. :)
The reason I'm curious is because I have built a VCO using this amplifier for 3GHz with a 10% tuning bandwidth and at some tuning voltages a spurious oscillation appears around the marked 880MHz frequency (and of course mixing products appear because of this spurious signal and the nominal frequency).
My wild guess would be that at these frequencies the parasitic inductances of the BJT and its package do not allow amplifying, but these inductances resonate with the feedback filter and that causes the spurious oscillations, but again, I'm not sure and the opinion of more experienced people cannot hurt. :)
Thanks in advance for any insights!
If you post your circuit schematic and-if it's possible-layout, we can comment something.