HomeBrew RF Siganl Generator FET Follower

Transcription

1 V1 is 9.0 Volts dc V2 is 0.5Vp-p at 30 Mhz Purpose and Function The Follower is used to isolate the oscillator from the loading effects of the amplifier stage. It isolates the oscillator from changes downstream providing a more stable design The FET follower provides a high impedance load to the oscillator. Direct loading of the oscillator caused frequency shifts and loss of amplitude or even stopped oscillation. Theory and Design Classic FET source follower. R4 determines the bias and current flow see calculated below. William R. Robinson Jr. p1of 8

2 Calculated o The data sheet does not give information o Simulation exceeds 10 Ghz but is the wrong FET Bias Vd = Vcc o Vd = 9V Vg = J1Vs (see Hartley_FET_Oscillator) o Vg = 1.5V Vg I choose Vgs at 2.0 as for J1 therfore Id and Rs are the same as J1 o Vsg 2 Id Idss(1 ) Vp Idss and Vp are dependent on the FET and can be found in the data sheets Vp is between 8.0 and 0 Volts 2 I will use the average of 4.0 volts Idss is between 2.0 and 20.0 ma 2, I will use the average of 11.0 ma Id 11ma(1 ) 4.0 Id = 2.75 ma o Rs = Vs * Id Rs = 2.0V * 2.75ma Rs = 5.5K so used readily available 4.7K o Vs = Vg +Vgs Vs = Vs = 3.5 As it turns out a 4.7K resistor created Vs = 2.44Vdc in simulation and 4.3 Vdc in the real circuit o So these calculations are off significantly o The simulation model FET is not an exact match o Looking at the Data sheet 2 we see a large variance for Vgs and Idss. This and Reference 3 show that it is very difficult to calculate the biases for and FET with any degree of certainty William R. Robinson Jr. p2of 8

3 Simulation Csim Rsim1 and Rsim2 were added to provide the same Vg as the output of the Hartley_FET_Oscillator. Rload has little or no affect on Vout o For Rload = 1K to 10K in 1K steps Bottom is 1K VDB(vout)[0] VDB(vout)[1] VDB(vout)[2] VDB(vout)[3] VDB(vout)[4] VDB(vout)[5] VDB(vout)[6] VDB(vout)[7] VDB(vout)[8] VDB(vout)[9] FET_Follower-Small Signal AC-8-Sweep-Graph m m m m m m m m m m k k k 1.000M M M 1.000G o Rload was picked as the input load of J2 o Rload = Xc5 + R5_Gain + R6//R7 4 The lowest Rload is the worst case so we solve for High frequency and R5_gain set to 0 ohms Rload = 68K//18K 1 Xc 2FC Xc = 1/(2 * pi * 27 Mhz * 0.47uf) Xc = 12.5K 68K//18K = 14.5K Rload = 12.5K K Rload = 27K R4 not critical 1K to 10K as only moves the dc bias point which could affect the swing picked same as R2 = 4.7K o For R4 = 1K to 10K in 2K steps Bottom trace is 1K William R. Robinson Jr. p3of 8

4 v(vout)[0] v(vout)[1] v(vout)[2] v(vout)[3] v(vout)[4] william-transient-10-sweep-graph u u u u u u Time u u u u The chat below shows the frequency plot VDB(vosc) VDB(vout) FET_Follower-Small Signal AC-7-Graph m m m m m m m m m m 1.000k k k 1.000M M M 1.000G Fc the cutoff frequency is way above the designed 27 Mhz (> 10 Ghz) It took nearly 200 pf of stray capacitance on the output to bring the simulation cutoff frequency into alignment with the measured circuit. o This seems excessive even with ugly construction Bias Vd = 9.00 Vg = 1.27 Vs = 2.44 William R. Robinson Jr. p4of 8

5 Real Circuit Below is the input and output of the at 10 Mhz o Channel 1 is the from J1s Vs = 1.4 Vp-p + 1.5Vdc at 9.6 Mhz o Channel 2 is the from J2s Vs = 0.92 Vp-p + 4.1Vdc at 9.6 Mhz o Note scope probe detuned the oscillator frequency from the 10 Mhz which was measured at the RF Generator output before the scope probe was attached. FET_Follower Gain db -3-4 Gain db -5-6 Mhz William R. Robinson Jr. p5of 8

6 Fc upper the upper cutoff frequency is just under 13Mhz which is way below the desired range of 27 Mhz The attenuation is 1 db in the pass band Bias measured Vd = 9.0 Vg = 1.5 Vs = 4.1 William R. Robinson Jr. p6of 8

7 Comparison Cutoff frequency varies greatly Biases vary significantly as discussed above. Real-Measured Simulation Calculated Fc Mhz 12.9 > 10 G hz N/a Bias Volts Vd Vg Vs William R. Robinson Jr. p7of 8

8 References 1. Hayward, Wes, W7ZOI, et al, Experimental Methods in RF Design (First Edition), JFET Bias circuit and Equations, Fig 2.19, p UNKNOWN, MPF102 N-Channel RF Amplifier, (FAIRCHILD 2004), online, accessed Horowitz, Paul and Hill, Winfield, The Art Of Electronics Second Edition, (Cambridge University Press 1989) Section 3.05 Manufacturing Spread of FET Characteristics, p Horowitz, Paul and Hill, Winfield, The Art Of Electronics Second Edition, (Cambridge University Press 1989) Section 2.07 Common-emitter Amplifier (Input Impedance), p77. William R. Robinson Jr. p8of 8