ECE 304: Feedback Examples Read S&S § 8.3-8.6, pp. 798-831 1 Given the voltage amplifier shown in Fig. 1 with input resistance Ri and output resistance Ro, and a choice of drivers from Fig. 2, we are going to connect the appropriate driver from Fig. 2 to a resistor load RL using two different feedback amplifiers. + Vi – AυVi + – Figure 1: Voltage amplifier RTHIN RN Thevenin voltage driver Norton current driver + VTH – Figure 2: Two possible sources to drive the feedback amplifier. Use an idealized feedback network to produce each of the following 1.1 A VOLTAGE AMPLIFIER 1.1.a Select the appropriate driver from Fig. 2. Explain your choice. Solution: The gain of the voltage amplifier has dimensions V/V. That is, input variable is voltage: therefore we want to use a Thevenin voltage source from Fig. 2. 1.1.b If the input resistance seen by VTH or IN with feedback is called Ri,f, is Ri,f increased or decreased by feedback? Explain why the change is desirable for this type of amplifier. Solution: Ri,f should be increased by feedback so that more voltage will be coupled into the amplifier from the input voltage divider. 1.1.c If the output resistance seen by VTH or IN with feedback is called Ro,f, is Ro,f increased or decreased by feedback? Explain why the change is desirable for this type of amplifier. Would you model the entire amplifier including feedback as a Norton source or as a Thevenin source? Explain. Solution: For a voltage amplifier the output variable is voltage. Therefore the entire amplifier should be modeled as a Thevenin source. For a Thevenin source we want the Thevenin Unpublished work © 3/29/05 JR Brews Page 1 3/29/2005resistance to be as low as possible, so the maximum voltage will be delivered to the load by the output voltage divider. Therefore, RO,f should be reduced by feedback. 1.1.d Show the feedback network, including the type of dependent source, i.e. VCVS, VCCS, CCVS, CCCS. Don’t forget RL. Solution: To increase Ri,f we want a series input connection. To decrease Ro,f we want a shunt output connection. The dimensions of βFB are reciprocal to the amplifier gain, so V/V. The input variable to the feedback source is voltage and the output variable is voltage so Vf = βFB Vo, and we have a VCVS. Voltage Amplifier VCVS + Vi – AυVi + –+ Vo– + –Vf = βFBVo Unpublished work © 3/29/05 JR Brews Page 2 3/29/20051.2 A TRANSCONDUCTANCE AMPLIFIER 1.2.a Select the appropriate driver from Fig. 2. Explain your choice. Solution: The gain of the transconductance amplifier is A/V. That is, voltage is the input variable. Therefore the appropriate driver is the Thevenin voltage source. 1.2.b If the input resistance seen by VTH or IN with feedback is called Ri,f, is Ri,f increased or decreased by feedback? Explain why the change is desirable for this type of amplifier. Solution: Feedback should increase Ri,f to insure more voltage is input to Ri of the amplifier by the voltage divider at the input. 1.2.c If the output resistance seen by VTH or IN with feedback is called Ro,f, is Ro,f increased or decreased by feedback? Explain why the change is desirable for this type of amplifier. From the standpoint of the output, would you model the entire amplifier including feedback as a Norton source or as a Thevenin source? Explain. Solution: For the transconductance amplifier gain is A/V → output is current. Therefore we model the entire amplifier as a Norton source, and we want Ro,f as large as possible so that the Norton resistance of the entire amplifier will be large, coupling more current from the output current divider into the load resistance. Therefore, RO,f should be increased by feedback. 1.2.d Show the feedback network, including the type of dependent source, i.e. VCVS, VCCS, CCVS, CCCS. Don’t forget RL. Solution: To increase Ri,f we want a series input connection. To increase Ro,f we want a series output connection. The dimensions of βFB are reciprocal to the amplifier gain, so V/A = Ω. The input variable to the feedback source is current and the output variable is voltage so Vf = βFB Io, and we have a CCVS. Transconductance Amplifier + – CCVS + Vi – AυVi Vf = βFBIo Io + – Unpublished work © 3/29/05 JR Brews Page 3 3/29/20051.3 A TRANSRESISTANCE AMPLIFIER 1.3.a Select the appropriate driver from Fig. 2. Explain your choice. Solution: The gain of the transresistance amplifier is V/A. That is, current is the input variable. Therefore the appropriate driver is the Norton current source. 1.3.b If the input resistance seen by VTH or IN with feedback is called Ri,f, is Ri,f increased or decreased by feedback? Explain why the change is desirable for this type of amplifier. Solution: Feedback should decrease Ri,f to insure more current is input to Ri of the amplifier by the current divider at the input. 1.3.c If the output resistance seen by VTH or IN with feedback is called Ro,f, is Ro,f increased or decreased by feedback? Explain why the change is desirable for this type of amplifier. From the standpoint of the output, would you model the entire amplifier including feedback as a Norton source or as a Thevenin source? Explain. Solution: For the transresistance amplifier gain is V/A → output is voltage. Therefore we model the entire amplifier as a Thevenin source, and we want Ro,f as small as possible so that the Thevenin resistance of the entire amplifier will be small, coupling more voltage from the output voltage divider into the load resistance. Therefore, RO,f should be reduced by feedback. 1.3.d Show the feedback network, including the type of dependent source, i.e. VCVS, VCCS, CCVS, CCCS. Don’t forget RL. Solution: To decrease Ri,f we want a shunt input connection. To decrease Ro,f we want a shunt output connection. The dimensions of βFB are reciprocal to the amplifier gain, so A/V = S. The input variable to the feedback source is voltage and the output variable is current so If = βFB Vo, and we have a VCCS. If = βFBVo + Vo – + –+ Vi – AυVi VCCS Transresistance Amplifier Unpublished work © 3/29/05 JR Brews Page 4 3/29/20051.4 A CURRENT AMPLIFIER 1.4.a Select the appropriate driver from Fig. 2. Explain your choice. Solution: The gain of the transresistance amplifier is A/A. That is, current is the input variable. Therefore the appropriate driver is the Norton current source. 1.4.b If the input resistance seen by VTH or IN with feedback is called Ri,f, is