Lab 4 - ECE 420L 

Authored by Stephanie Silic

silics@unlv.nevada.edu

February 22nd, 2017

  

Lab Description

  

This lab is a discussion of the bandwidth and slew rate of op-amps, using the LM324 op-amp.

   

This report includes the following sections:

  

Lab Report

 

Part 1 - Gain Bandwidth Product    

 

    The bandwidth of an op-amp is the range of frequencies where it's closed loop gain is  constant.  At a gain of 1, for example,  say in an inverting topology, we say its bandwidth is the frequency where its gain starts to drop. The point where the gain has gone down by 3 dB is where this frequency is defined. The Gain Bandwidth Product (GBP) is the closed-loop gain multiplied by the bandwidth at that gain, and it is a constant. The GBP is equal to the frequency of the op-amp at unity gain ( 1*BW = GBP), also called the unity gain frequency, fun.

  

The open loop frequency response of the LM342 is given in its datasheet. We note that the GBP, or unity gain frequency should be around 1 MHz:

  

http://cmosedu.com/jbaker/courses/ee420L/s17/students/silics/Lab4/Datasheet_open_loop.PNG

 
The following circuit and simulation illustrates the frequency response of an inverting topology at gains of 1, 5, and 10.  
 
http://cmosedu.com/jbaker/courses/ee420L/s17/students/silics/Lab4/GBP_ckt_and_bode.PNG

 

We note from the frequency response shown above that the bandwidth gets smaller as gain increases.
 
http://cmosedu.com/jbaker/courses/ee420L/s17/students/silics/Lab4/GBP_datasheet.PNG
 
The Gain Bandwidth Product (GBP) of the LM324 op-amp is 1.3MHz at Vcc=1.3MHz. This means that its bandwidth at a certain gain can be calculated as 1.3MHz/gain. The bandwidths at gains of 1, 5, and 10 can be estimated as follows:
    -  Gain of 1:  bandwidth = 1.3MHz,
    -  Gain of 5:   bandwidth  = 260kHz
    -  Gain of 10:  bandwidth  = 130kHz
.            - Provide schematics of the topologies you are using for experimental verification along with scope pictures/results.
            - Associated comments should include reasons for any differences between your estimates and experimental results.
 
Circuit used in experiment non-inverting topology, gains of 1, 5, and 10 (left to right):
 
 
http://cmosedu.com/jbaker/courses/ee420L/s17/students/silics/Lab4/Spice_noninv_1_sch.PNG http://cmosedu.com/jbaker/courses/ee420L/s17/students/silics/Lab4/Spice_noninv_5_sch.PNG http://cmosedu.com/jbaker/courses/ee420L/s17/students/silics/Lab4/Spice_noninv_10_sch.PNG
 
The experimental procedure for determining the bandwidth of the op amp a gains of 1, 5, and 10 is to first set the input signal to a fairly low frequency so that the gain is accurate, then increase the frequency to the point where the output has dropped by 3dB, or is 70.7% of its value. The frequency where this happens is the bandwidth of the op amp at that gain.
 
The oscilloscope screenshots below show the gain at a low frequency (left), and the gain where it has dropped by 3dB (right).  
 
If I found that the gain was not exactly as I expected, I still noted the bandwidth at a 3dB drop from the actual gain seen on the oscilloscope.
 
In all oscilloscope screenshots, the input is the blue waveform (channel 2) and the purple waveform (channel 3) is the output.
 
  Gain of 1:
http://cmosedu.com/jbaker/courses/ee420L/s17/students/silics/Lab4/gain_of_1.PNG  http://cmosedu.com/jbaker/courses/ee420L/s17/students/silics/Lab4/non_inv_BW_1.PNG
 
Gain of 5:
http://cmosedu.com/jbaker/courses/ee420L/s17/students/silics/Lab4/gain_of_5.PNG  http://cmosedu.com/jbaker/courses/ee420L/s17/students/silics/Lab4/non-inv_BW_5.PNG
 
Gain of 10:
http://cmosedu.com/jbaker/courses/ee420L/s17/students/silics/Lab4/gain_10.PNG  http://cmosedu.com/jbaker/courses/ee420L/s17/students/silics/Lab4/non-inv_BW_10.PNG
 
The experimental results versus the estimated bandwidths are summarized in the table below:
 
http://cmosedu.com/jbaker/courses/ee420L/s17/students/silics/Lab4/non-inv_GBPtable.PNG
 
Inverting topology circuit used:
 
http://cmosedu.com/jbaker/courses/ee420L/s17/students/silics/Lab4/Spice_INV_1_sch.PNG
   
Again, the screenshots on the left show the gain at a frequency within the bandwidth, and the screenshots on the right show the gain decreased by 3dB, or when it has dropped to 70.7% of its value.

Gain of -1:
 
 http://cmosedu.com/jbaker/courses/ee420L/s17/students/silics/Lab4/gain_of_-1.PNG  http://cmosedu.com/jbaker/courses/ee420L/s17/students/silics/Lab4/inverting_BW_gain_-1.PNG
   
Gain of -5:
 
http://cmosedu.com/jbaker/courses/ee420L/s17/students/silics/Lab4/gain_of_-5.PNG  http://cmosedu.com/jbaker/courses/ee420L/s17/students/silics/Lab4/inverting_BW_-5.PNG
 
Gain of -10:
 
http://cmosedu.com/jbaker/courses/ee420L/s17/students/silics/Lab4/gain_of_-10.PNG  http://cmosedu.com/jbaker/courses/ee420L/s17/students/silics/Lab4/inverting_BW_gain_-10.PNG
   
The experimental results versus the estimated bandwidths are summarized in the table below:
   
  http://cmosedu.com/jbaker/courses/ee420L/s17/students/silics/Lab4/inverting_GBPtable.PNG
 
Part 2 - Slew Rate:
 
    The output of an op amp can only change at a certain rate.  This rate of change is called the slew rate. On the datasheet of the LM324, we find that the slew rate is about 0.4 V/us.
 
http://cmosedu.com/jbaker/courses/ee420L/s17/students/silics/Lab4/Slew_rate_datasheet.PNG\
http://cmosedu.com/jbaker/courses/ee420L/s17/students/silics/Lab4/Datasheet_voltage_follower.PNG  http://cmosedu.com/jbaker/courses/ee420L/s17/students/silics/Lab4/Datasheet_voltage_follower1.PNG
 
To experimentally determine the slew rate of an op amp, we put the op amp in the voltage follower configuration as shown in the datasheet:
 
http://cmosedu.com/jbaker/courses/ee420L/s17/students/silics/Lab4/Spice_noninv_1_sch.PNG
 
    If we apply a square wave to the input of the voltage follower, the output will not be able to change as fast as the edge of the square wave so we will see the slew rate directly as the output changes as fast as it can:

     

http://cmosedu.com/jbaker/courses/ee420L/s17/students/silics/Lab4/Slew_square.PNG

  

We can calculate the slew rate experimentally from the experiment shown above by finding the slope of the output, or change in voltage divided by change in time. This turns out to be about 500mV/1.32 us = .37V/us which is very close to the datasheet's value of .4V/us!

  

    If we use a sinusoid as the input, we notice that as the amplitude of the sine wave increases, the slope between the peaks of the sinusoid increases as well, even at relatively low frequencies (say, 100kHz).  if this slope is steeper than the slew rate, the output of the op amp will be distorted, since it cannot change fast enough to keep up with the sine wave. The slew rate can be seen again in the following screenshot:

    

   http://cmosedu.com/jbaker/courses/ee420L/s17/students/silics/Lab4/Slew_sine_cursors.PNG\

  

  

 

Return to my labs