Hello, and welcome to the TI Precision Lab supplement for Intrinsic Op Amp Noise. This lab will walk through detailed calculations, SPICE simulations, and real-world measurements that greatly help to reinforce the concepts established in the noise video series.

The detailed calculation portion of this lab can be done by hand, but calculation tools such as MathCAD or Excel, can help greatly since noise calculations can involve many steps.

The simulation exercises can be performed in any SPICE simulator, since TI provides generic SPICE models of the op amps used in this lab. However, the simulations are most conveniently done in TINA-TI, which is a free SPICE simulator available from TI website. TINA simulation schematics are embedded in the presentation.

Finally, the real-world noise measurements are made using a printed circuit board, or a PCB, provided by TI. If you have access to standard lab equipment, you can make the necessary measurements with any plus or minus 12 volts power supply and oscilloscope. However, we highly recommend the VirtualBench from National Instruments. The VirtualBench is an all-in-one test equipment solution which connects to a computer over USB or Wi-Fi and provides power supply rails, analog signal generator, and oscilloscope channels, and a 5 and 1/2 digit multimeter for convenient and accurate measurements. This lab is optimized for use with the VirtualBench.

In experiment 1, we'll determine the total output voltage noise in a circuit with no filtering. First, calculate the expected total RMS and peak-to-peak output noise voltage for the circuit shown here using the techniques and equations given in the noise lecture. Note that the switch JMP9 is open, so filter capacitor C5 is not connected to the circuit. Make this calculation twice. First, with the OPA188 selected for U3 and U4, then with OPA211. The different parameters of these op amps will give you different noise results.

In order to perform the noise calculations, you need to know certain parameters of each op amp. The key parameters are the op amp gain bandwidth product, or GBW; input voltage noise density, or e sub n; and input current noise density, or i sub n. This slide gives a table of these key specs for both the OPA188 and OPA211. Enter your answers in the bottom of the slide. The solutions are already provided to allow you to check your work.

I won't go through the entire calculation, but rather will give an overview of the key steps. First, the resistor thermal noise spectral density is calculated using the circuit resistor values. Next, the total input noise spectral density is computed based on the resistor thermal noise, op amp input voltage noise, and op amp input current noise once it's been converted to voltage. The op amps gain bandwidth and the circuit's closed loop bandwidth are used to calculate the noise bandwidth, which enables you to determine the total input referred RMS noise.

Finally, the input referred noise is multiplied by the closed-loop gain in order to compute the total RMS and peak-to-peak output noise. This slide shows the complete calculation for the OPA188.

And this slide shows the calculation for the OPA211. The op amp's input voltage and current spectral density as well as the op amp gain bandwidth product are different, but the steps of the calculation are exactly the same. Simply substitute in the new values and take note of the result.

The next step is to run the SPICE simulation analysis for the total output noise. Simply open the TINA-TI simulation schematics embedded in the presentation. Ensure that the filter capacitor jumper is open. Then, select Analysis followed by Noise Analysis. Make sure that Total Noise is selected, then run the analysis from 1 Hertz to 100 megahertz. Run the simulation for both the OPA188 and OPA211.

Over the 100 megahertz frequency range, the OPA188 circuit has a total noise of 3.59 millivolts RMS, or 21.5 millivolts peak to peak. The OPA211 circuit has a total noise of 10.75 millivolts RMS or 64.5 millivolts peak to peak. Make sure to disable the DC power supply before setting up the test PCB.

In the VirtualBench software, click the Power button in the DC power supply area to turn off the power. Next, check the front panel of the VirtualBench to make sure the LEDs are off. Also, ensure that the function generator is Off.

To prepare the test board for the measurement, install the jumpers and devices on circuit 3 and circuit 4 as shown here. On circuit 3, install JMP11 and JMP12, as well as the OPA188 in sockets U3 and U3.

On circuit 4, install JMP15 and JMP16 as well as the OPA211 in sockets U5 and U6. This gives the full schematic for circuit 3 on the TI Precision Labs Test Board. You will use this circuit to measure the total output noise of the OPA188.

This slide gives the full schematic for circuit 4 on the TI Precision Labs Test Board. You will use the circuit to measure the total output noise of the OPA211.

For the test board to function properly, it is important that you only install jumpers and devices in circuits 3 and 4. Do not install any jumpers or devices in any other circuits on the PCB. Remove any jumpers or devices from the unused circuits and store them in the storage area at the bottom of the test board.

This gives the connection diagram between the TI Precision Labs Test Board and the National Instruments VirtualBench. Connect the provided power cable to the DC power supply of the VirtualBench and power connector J4 on the test board. Connect V out 1 on the test board to VirtualBench oscilloscope channel 1 and V out 2 on the test board to VirtualBench oscilloscope channel 2 using BNC cables.

Next, apply power to the National Instruments VirtualBench and connect it to your computer with the USB cable. The hardware should be detected as a virtual CD drive. And you can run the VirtualBench software directly from the drive.

Once the software opens, configure the software as follows. Set the time scale to 100 milliseconds per division with acquisition mode set to Auto. Enable channels 1 and 2 on the oscilloscope and set them to 1x AC coupled mode. Set the vertical scale on channel 1 to 10 millivolts per division. And on channel 2, to 20 millivolts per division. Set the plus 25 volt power supply to plus 15 volts, 0.5 amps. Set the minus 25 volt power supply to minus 15 volts 0.5 amps.

Press the Power button to turn on the power supply rails. Enable peak to peak and RMS measurements on both channels in order to read the output voltage of each circuit.

You must also set the mode of the VirtualBench oscilloscope. Click the button shown on the front panel, then set acquisition to Sample and persistence to Disabled.

The expected output voltage noise results from the measurement are shown in the screenshot at the top left. The OPA188 has a measured noise of 19.8 millivolts peak to peak, or 3.3 millivolts RMS. And the OPA211 has a measured noise of 77.4 millivolts peak to peak or 12.9 millivolts RMS. As you can see from the table, this agrees extremely well with the results from calculation and simulation.

In experiment 2, we'll determine the total output voltage noise in the circuit with a filter capacitor in the feedback network. Perform another total noise calculation on the circuit shown. However, this time a filter capacitor is added to the feedback network of the first amplifier stage. This capacitor will reduce the circuit's noise bandwidth so the overall noise performance will be significantly improved. As before, do the calculations with both the OPA188 and OPA211.

The key parameters of op amp gain bandwidth product input voltage noise density and input current noise density are given again for your reference. Enter your answers in the table at the bottom of the slide. The solutions are already provided to allow you to check your work.

The steps in the calculation are nearly the same as before. However, this time the filter capacitor must be taken into account when calculating the circuit's closed-loop bandwidth. With a filter capacitor connected, the closed-loop bandwidth was computed to be 1.59 kilohertz. Without the capacitor, the bandwidth was 19.8 kilohertz. As expected, this helps to reduce the total noise.

Step through the calculation again for the OPA211, substituting in for the noise and bandwidth parameters of that device. Run another SPICE simulation analysis for the total output noise. This time, ensure that the filter capacitor jumper is closed so that the capacitor is connected. As before, select Analysis followed by Noise Analysis. Make sure the Total Noise is selected, then run the analysis from 1 Hertz to 100 megahertz. Remember, run the simulation for both the OPA188 and OPA211.

With the filter capacitor connected, the OPA188 circuit has a total noise of 1.09 millivolts RMS, or 6.54 millivolts peak to peak. The OPA211 circuit has a total noise of 0.585 millivolts RMS, or 3.51 millivolts peak to peak.

The jumper settings on the test board must be modified before rerunning the bench measurement. Simply install jumpers JMP9 and JMP13. All other jumpers and devices remain the same from the previous experiment.

This slide shows the full schematic for circuit 3 on the TI Precision Laps Test Board. You will use this circuit to measure the output noise of the OPA188, this time with the filter capacitor connected.

You will use this circuit to measure the total output noise of the OPA211. Again, with the filter capacitor. The general test board setup remains the same for experiment 2. Do not install any devices or jumpers in unused circuits. The cable connections between the test board and the VirtualBench also remain the same for experiment 2.

In the VirtualBench front panel, change the vertical scale of channel 2 to 10 millivolts per division. All other settings remain the same.

The expected output voltage noise results from the measurement with the filter are shown here. The OPA188 has a measured noise of 6.58 millivolts peak to peak, or 1.09 millivolts RMS. And the OPA211 has a measured noise of 4.12 millivolts peak to peak, or 0.69 millivolts RMS. Like before, the calculated, simulated, and measured results all match quite closely.

The real question is, how much benefit did we get by adding the filter capacitor? This slide compares the noise performance of both the OPA188 and OPA211 with and without the filter. It should be clear just from the oscilloscope capture that the filter gave us a massive improvement in noise.

If you compare the peak to peak noise values in the table, you can see that with the filter we achieved a 3x improvement with the OPA188 and a 19x improvement with the OPA211.

You may wonder why we didn't see more of a reduction in noise with the OPA188. The key reason is that the gain bandwidth product of the OPA188 at 2 megahertz is much lower than that of the OPA211 at 80 megahertz. So the closed-loop bandwidth of the OPA188 circuit is lower as well.

When we added the filter and reduced the closed-loop bandwidth of both circuits to only 1.6 kilohertz, this was much more of a reduction for the OPA211 than the OPA188. Therefore, the filter had a bigger effect for the OPA211 circuit. The OPA188 also has a higher input voltage noise spectral density than the OPA211.

That concludes this lab. Thank you for your time.