Irfd Update
Irfd Update
Frequency Design
October 31, 2001, returned to web on Nov 9, 2005.
I got an e-mail the other day (that is, in 2001) that pointed out an error in
Chapter 7 dealing with oscillators. This had to do with Figure 7.4 which
shows the effect of emitter bias current in an NPN Colpitts oscillator on starting
gain and phase. The circuit used for analysis is presented here:
In an ideal NPN, the signal current presented to the emitter emerges from the
collector. A collector voltage is then generated, defined by the net impedance
presented to the collector. Here is where I made an error. In my analysis back
in the pre PC era, I assumed that the impedance was resonant and constant.
This is not a good assumption as you can show through a more complete
analysis. You really have to do the analysis with more care. The resonance is
altered by the loading from the emitter current, as transformed through the
Colpitts network of C1 and C2.
The next figures show a more up-to-date analysis in MathCad.
The general character of the curve is unchanged, but the values are different.
Specifically, the phase goes through zero degrees at a bias current of about 0.3
mA.
Many thanks to John Broderick from RF Integration Inc. in Lowell, MA, the
sharp eyed reader who caught the error.