In part 1 of this series we discussed how effective input capacitance is formed. In this section we're going to try different opamps in SPICE and make real measurement. Lastly, compare the results.
Test Setup:
We are going perform the same experiment that we did in SPICE in part 1 .Test setup is very simple oscilloscope and function generator are used to measure the bandwidth, the same method that was shown in figure.1 of part 1 was used. Test PCB shown below in figure.1.
The only thing that changes is the opamp in this setup. R23 in the test PCB is the series resistance, which is 10kΩ.
Here are the results in a spreadsheet. Calculated, simulated and measured results are shown.
As I pointed in part 1, it doesn't make a lot of sense to accurately predict what you are going to get in real life because of the huge variation in open loop gain and differential resistance. For ADA4522 and ADA4898 measured value is much larger than expected. You might be think it caused by layout, but you have to realize I managed to measure 4.7pF for OPA192 with the same PCB, so not all of 58pF coming from layout. I am going to claim that it because of process variation, but I am open to discussion and suggestion.
Conclusion
Althought this experiment seems like a failure (well it kind of is), we can still learn something valuable from the results. For opamps that have huge differential resistance common mode input capacitance is the only thing to worry about, if it's not large expect to have some or a lot more input capacitance.
Comments