Experiment No. 4 The LM 741 Operational Amplifier By: Prof. Gabriel M. Rebeiz The University of Michigan EECS Dept. Ann Arbor, Michigan The LM * 741 is the most widely used op-amp in the world due to its very low cost (<10 cents in bulk) and good, but not excellent, electrical characteristics. It is composed of 18 transistors integrated together on a tiny silicon chip. This silicon chip is inserted into an 8-pin or a 14-pin and/or signs. They package with the connections shown below. Always refer to the indicate the location of pin #1. The essential electrical characteristics of the LM 741 op-amp are presented below. * 1 LM is a trademark of National Semiconductor, Inc. The µA 741 was actually developed by Fairchild, Inc., one of the early and very successful electronic companies. However, they did not survive in the Silicon Valley man-eat-man atmosphere, and National Semiconductor bought Fairchild. Supply Voltage: up to +12 V typ., +16V max. Gain Bandwidth: 0.4–1.5 MHz (Gain = 1, BW = 1 MHz) Gain = 100, BW = 10 KHz) Input Bias Current: 0.1–0.5 µA typical Input Resistance: 2 Mž typ. Output Current: 25 mA max. Output Resistance: 20 ž typ. Total Harmonic Distortion: <0.2% at 1 KHz (in the linear region!) The 741 op-amp is NOT suitable for driving speakers since the output current is limited to 25 mA. However, it is an excellent amplifier resulting in medium gain (20-100) for audio signals in the mV range. Of course, one can design 741 amplifier with a gain of 500, but then, the bandwidth will be limited to 2 kHz! In the lab, you will be using the LM 747 chip, which contains two LM 741 op-amps in the same package. The pin configuration is: 2 Taken from National Semiconductor–Operational Amplifiers Databook, 1995 Edition. 3 Taken from National Semiconductor–Operational Amplifiers Databook, 1995 Edition. 4 Taken from National Semiconductor–Operational Amplifiers Databook, 1995 Edition. 5 Experiment No. 4. Variable Gain Amplifiers; Summers; Intermodulation Products Goal: To build and test a variable gain audio amplifier (100 Hz – 20 kHz), and then upgrade it to a two-channel summer. Also, to see the intermodulation products when two signals are fed into a non-linear amplifier. R Read this experiment and answer the pre-lab questions before you come to the lab. 4.1 Variable Gain Audio Amplifier: Equipment: • Agilent E3631A Triple output DC power supply • Agilent 33120A Function Generator (Replacement model: Agilent 33220A Function / Arbitrary Waveform Generator) • Agilent 34401A Multimeter • Agilent 54645A Oscilloscope (Replacement model: Agilent DSO5012A 5000 Series Oscilloscope) An inverting variable-gain amplifier suitable for audio frequencies is shown below: R2 = 100 kΩ 1 µF R = 200 Ω s1 + ~ R1 = 2.4 kΩ + Ra Rb leave open Vi1 100 kΩ R4 = Pot. 100 kΩ – Vs1 – Signal Generator DC block (1 µF Cap.) and increase source resistance to 200 Ω (Rs1) + LM 741 + –+ –+ RL = 1 kΩ Vo 220 µF 220 µF NEG (12) GND POS (12) – Ra + Rb = 100 kΩ + Vcc = + 12 V - Vcc = - 12 V Ra Rb Potentiometer 1. R Draw the circuit in your notebook. 2. Assemble the circuit on the breadboard. When the circuit is ready, connect the cables to the power supply, but do not apply the voltage.Show your circuit to your lab instructor and he/she will check it and help you test it. 6 VARIABLE GAIN AMPLIFIER: 3. R Measure the DC voltages V–, V+ and Vo. They should all be in the mV range. 4. Set the source at 1 kHz and Vs = 200 mV ppk and connect it to the amplifier. Connect Vs1 to Channel 1 of the scope. 5. R Connect Vo to Channel 2. Active Channel 2, measure Vppk (1) and Vppk (2). Vary the potentiometer from 0 Ω to 100 kΩ. Determine the minimum and maximum Vo. Determine the minimum and maximum gain (Vo/Vs1). Write down your results and calculations. Check to see if the experimental data agree with your pre-lab calculations (Gain from ~1to ~40). Check the frequency domain representation at both gain positions. Basically, if your amplifier is linear, and you are not clipping the output (amplifier saturation) then you should not observe any harmonics, (2fo, 3fo, ...). Sketch the FFT spoectra, write a brief conclusion. 6. R Measure the frequency response from 10 Hz to 1 MHz. Take data at 10, 20, 50, 100, 200, 500 Hz, 1, 2, 5, 10, 20, 50, 100, 200, 500 kHz, and at 1 MHz. At each frequency, measure Vi and Vo. Do these measurements for three gain settings: for the minimum gain setting (|G| ~ 1), maximum gain setting (|G| ~ 40) and for the midband gain setting of |G| ~ 10. Organize your data in a table form. 7. Something interesting happens for G ≈1 (unity gain). You will have a +4-6 dB peak in the frequency response at ~200 KHz. This is due to the internal “compensation” capacitor (C1 = 30 pF) which ensures stable operation of the opamp under all negative feedback conditions. At unity gain, the interaction of C1 with the transistors around it create this peak. This is advanced analog circuit design and you will see it in EECS 413. However, you can always say that you measured it first in EECS 210! Congratulations; you have built a variable gain audio amplifier with a gain of ~1-40 (0-32 dB). 4.2 Summer and Intermodulation Products: 1. Push the Output/Off button of the power supply to Off position (no voltage applied to your circuit). In a different (but close) part of the proto-board, connect the following circuit to the amplifier. The complete circuit is shown on the Experiment No.4 Worksheet. 1 µF Rs2 = 200 Ω + ~ Vs2 – Wavetek Signal Generator 7 2. R + Vi2 R3 = 2.4 kΩ Ra Rb 100 kΩ Pot. leave open Negative terminal of the op-amp – DC block and increase source resistance to 200 Ω Redraw the entire circuit (with the op-amp) in your notebook. 3. Connect the output of the WAVETEK function generator to Vs2 and set it at 800 Hz and Vppk = 200 mV. Choose the gain of channel 2 to be around 10. 4. Connect the Agilent source to Vs1 and set it at 1 kHz and Vppk = 200 mV. Choose the gain of Channel 1 to be around 10. 2–CHANNEL SUMMER: 5. R Now, turn on both sources and look at the output waveform. The output waveform closely resembles the telephone dial tone! Look at the frequency domain and see the output spectrum. Plot the spectrum of Vo (frequency domain). Again, if your amplifier is linear, then you should only measure f1 (1 KHz) and f2 (800 Hz), and no intermodulation products between the two signals (2f1-f2 , 2f2 -f1, etc. ...). 6. Spend 5 minutes with the potentiometer (to vary the gain of each channel) and signal sources (to vary the frequency of each channel) to get any waveform you wish. You need not be adding only sinusoidal waves. You can try sine waves with triangular waves! However, make sure that you never drive the amplifier into clipping. Congratulations, you have built a two-channel audio summer, called a “mixer”, with a gain of 1-40 for each channel (0-32 dB). INTERMODULATION PRODUCTS: 7. R Disconnect the cables with T-connector from the Wavetek source. Increase the Agilent source voltage (Vs1) at f1 = 1 kHz until you drive the amplifier well into clipping. A Vs around 0.65 Vppk and maximum gain will be good. The output signal should have a fundamental frequency component of 17–18 dBV (at 1 kHz) and a third harmonic frequency component of around –10 dBV (at 3 kHz). Verify this using the scope in the FFT (frequency) domain. Choose a frequency span of 9.6 kHz. 8. R Now, turn on the Wavetek source with Vs2 = 200 mVppk and f2 = 600 Hz and set the gain to 10. Notice the JUNGLE of frequencies which turn up. These are called intermodulation products. Using the cursor, record the frequencies of components above the noise level. Verify if the following components are present in the spectrum: f1, f2, 2f1, 2f2, f1 + f2, f1 – f2, 3f1, 3f2, 2f1 – f2, 2f2 – f1, 2f1 + f2, 2f2 + f1. Make a clear record of which components are above the noise level – and which ones are buried in the noise. If in doubt, as your lab instructor for help 8 Explanation: The summer is the perfect circuit to see the intermodulation product between two tones in a non-linear circuit. Basically, if you have a circuit driven into non-linearity by a large input signal, then it will generate large amplitude harmonics. If a new but much smaller signal is fed into the amplifier, it will mix with all of the harmonics and will create a jungle of frequencies (see figure below). This example is only with two signals. Can you imagine what will happen if you have 3, 4, 5, ... signals? The simple answer is: HI-FI HELL. SO, THE GOLDEN RULE IS: NEVER DRIVE AN AMPLIFIER INTO THE NON-LINEAR REGION! 9 Experiment No. 4. Variable Gain Amplifiers; Summers; Intermodulation Products Pre-Lab Assignment 1. This question assumes an ideal op-am (Golden Rules apply): Calculate Vo/Vi1, Vi1/Vs1, and Vo/Vs1 for the potentiometer set at Ra = 0 Ω and Ra = 100 kΩ. Assume the 1 µF capacitor (DC block) to be a pure short circuit for your analaysis. YOU WILL NEED THIS FOR YOUR LAB. These questions deal with the non-idealities of the LM 741 op-amp. 2. Why is the load resistor (RL) of the LM 741 variable gain amplifier set at 1 kΩ? Calculate the max. output voltage swing if RL = 200 Ω. 3. The LM 741 is connected to a DC source of +12 V with RL = 1 kΩ. What is the maximum swing (approximately) of Vo before clipping occurs? If RL = 1 kΩ, what is the maximum power that can be delivered to the load (PL = Vpk2/2RL = Vrms2/RL). 4. You will notice on in Experiment #4 that there is a 100 kΩ resistor connected between the ground and the positive (non-inverting) input of the LM 741 amplifier. Can you explain why this is done? (See Lab #5 Introduction). Also, why 100 kΩ and not 5 kΩ or 500 kΩ? 10 Experiment No. 4. Variable Gain Amplifiers; Summers; Intermodulation Products Lab-Report Assignment 1. 2. Plot the frequency response of the inverting amplifier for the low, medium, and high gain (Vo/Vs1) settings on the same graph using MATLAB. Use a dB scale for the gain and a logarithmic-scale for frequency. Explain your results. Calculate the Gain•Bandwidth product at each gain setting. Do your calculations agree with what you expect? Explain. Design: 3. Design a variable gain amplifier with a maximum gain of ~100 and a minimum gain of ~0.5. You can use resistances up to 100 kΩ and a 0-200 kΩ potentiometer. An input resistance of 1 kΩ or above is required. 4. For the amplifier/summer on page 65, assume Rs1 = 10 kΩ and Rs2 = 1 kΩ. What is the max./min. gain for Vo/Vs1? What is the max./min. gain for Vo/Vs2? As you can see, there is a resistive divider occurring between Rs1 and (R1 + Ra), and between Rs2 and (R3 + Ra). Draw a circuit which ensures that both channels have exactly the same gain even if they have different source resistances (you can use other op-amps if you wish). You cannot use potentiometers in your design. Intermodulation Products: 5. In the lab report of Experiment #3, you calculated the harmonics generated by a non-linear amplifier. Now, you are going to calculate how intermodulation products occur in nonlinear amplifiers. A non-linear amplifier transfer function is given by: Vo = A Vi + ß Vi2 + γ Vi3 where A ≡ gain of amplifier ß, γ ≡ non-linear components and Vi = V1 cos (ω1t) + V2 cos (ω2t) ß, γ << A with V2 << V1. a. Calculate Vo(t) (neglect all components of V22 and V23 since they are very small). Put Vo in the form: Vo = A { ... } + ß { ... } + γ { ... }. As a result of your calculation, you should get rid of all powers, such as cos2 (ωt), replacing them with various harmonics such as cos (2ωt); get rid of the products, such as cos (ω1t) . cos (ω2t) replacing them with intermodulation products such as cos ((ω1 + 2ω) t), cos (( ω1 – ω2) t). Use trigometric formulas such as mentioned on p. 51 of this Manual. 11 b. Draw the output spectrum in dB for: V1 = 0.2 Vrms, V2 = 0.01 Vrms, f1 = 1 kHz, f2 = 1.1 kHz, A = 40, ß = 4, γ = 1. You can simply sketch this spectrum – or you can plot it with MATLAB, using stem command. 12 Experiment No. 4. Variable Gain Amplifiers; Summers; Intermodulation Products Worksheet/Notes R2 = 100 kΩ 1 µF ~ Rs1 = 200 Ω + + Vs2 – R1 = 2.4 kΩ Vi2 Ra Rb 100 kΩ Pot. + Vs2 – Wavetek Signal Generator + Vi2 – DC block and increase source resistance to 200 Ω + Vo – R3 = 2.4 kΩ 1 µF Rs2 = 200 Ω ~ + R4 = 100 kΩ Ra Rb 100 kΩ Pot. RL = 1 kΩ leave open Negative terminal of the op-amp – These experiments have been submitted by third parties and Agilent has not tested any of the experiments. You will undertake any of the experiments solely at your own risk. Agilent is providing these experiments solely as an informational facility and without review. 13 AGILENT MAKES NO WARRANTY OF ANY KIND WITH REGARD TO ANY EXPERIMENT. AGILENT SHALL NOT BE LIABLE FOR ANY DIRECT, INDIRECT, GENERAL, INCIDENTAL, SPECIAL OR CONSEQUENTIAL DAMAGES IN CONNECTION WITH THE USE OF ANY OF THE EXPERIMENTS.