Inverting Amplifier Design
- Thread starter JasonMcG
- Start date
Think of the feedback resistor that will be required.Gain=Rfeedback/Rinput
Gain is known.
Rinput is known.
Solve for Rfeedback.
Then wire it up like this:
Think about the effect of the input bias and offset currents.
Think of the implications for the bandwidth.
We need more information about the requirements and constraints.
You are going to have a lot of problems if you use a 741.
Look up the datasheet for it and you will see that the input resistance can be as low as 300 kΩ, which is just a few percent of the value of the feedback resistor you need using that topology. Then the input bias current can be as high as 1.5μA while the current in your feedback resistor is never going to exceed about 1.2μA. And with a gain-bandwidth product of 1 MHz, you are looking at a closed loop bandwidth of only 1 kHz.
Look up the datasheet for it and you will see that the input resistance can be as low as 300 kΩ, which is just a few percent of the value of the feedback resistor you need using that topology. Then the input bias current can be as high as 1.5μA while the current in your feedback resistor is never going to exceed about 1.2μA. And with a gain-bandwidth product of 1 MHz, you are looking at a closed loop bandwidth of only 1 kHz.
The input resistance of op amp itself is usually in MegaOhms. As in Millions of Ohms. Get it?
The main point of WBahn's is that if you have input signal higher than 1 kHz, you will have problem. The op amp in this situation acts as a simple filter. It simply filters OUT any signal that is higher than 1 kHz.
So.
Keep input signal below 1 kHz. See? No problem.
If you have access to some simulation software, build the circuit, it is only 6 parts:
signal source
10 kOhm resistor
feedback resistor
741 op amp
positive power supply
negative power supply
Then check it out, see what it does when your signal source is 1 Hz, 100 Hz, etc. And see what it does when you put in 2 kHz, 10 kHz, etc.
The input resistance of an amplifier is the resistance as seen by whatever is driving the amplifier's input. For some opamp-based topologies it is the input resistance of the opamp, but in most it is not. That includes the topology that shteii01 recommended.
You want an opamp that has a max input bias current that is at most about 10% of the current in your signal path and an input resistance that is at least about 10x the value of any of the resistors in the path. You also need it to have a gain-bandwidth product of at least 1000 times the highest signal frequency you are interested in.
No, that was not the main point. Since it hasn't been said what the max signal frequency is, that GBW product may or may not be an issue.
The other items WILL be issues.
The 741 has a hard time making it to within 2 or 3 volts of the rail, so the maximum output signal will be ~±12V. That means that the maximum input signal will be about ±12mV. With a 10kΩ input resistance, that means that, at the max signal, the current into the input resistor will be 1.2μA. The gain calculation is based on essentially all of that current then flowing through the feedback resistor, but in reality it may not even be enough to furnish the input bias current needs of the opamp! At lower signals, say something that will result in a 3V output signal, the input signal current almost certainly won't be enough to supply the needs and so the output voltage will be determined by the input bias current and not the input signal. Even the polarity of the output signal will likely be the opposite of what is expected.
