Hi team,
I want to use OPA4348 like below, I know that this device voltage output swing from rail is about 18mV typical value. But I find that the output voltage vary with different VCM, could you please explain what happend here? Thanks.
BR,
| VCM | Viff | VO |
| 0 | 0 | 2.2mV |
| 1 | 0 | 10mV |
|
2 |
0 | 18mV |
|
3 |
0 | 26mV |
Hi CL,
I think there are at least two things worth pointing out here.
First, the non-ideal op amp has quite a bit of Common-mode rejection. However, this is not infinite. In other words, when the common mode changes there will be some change in the output. You can figure out how much by looking at the CMRR specification in the datasheet. For this part, it is typically 70 to 80 dB, depending on the common mode voltage range. You can learn about CMRR here.
Another thing to note is that the offset of the amplifier is going to cause a non-ideal output. Take a look at my simulation shown below. The U1 and U2 op amps are in the exact same configuration. So, their difference in output voltage is negligible. This is true in the simulator because both op amps are assumed to have the same offset voltage. In the real world, this would not be the case.
However, the offset voltage of U3 will be seen at the output. If we imagine the offset voltage as a voltage source going into the non-inverting input of U3 with a magnitude of ~+1mV, then we would expect this offset to be gained by the U3 stage.
Please let me know if you have any further questions.
Regards,
Daniel
Hi Lin,
no OPAmp can work properly with the input and output voltages close to the rails. Many parameters like open-loop gain, etc., go bad when approaching the rails. For the OPA4348 the no man's land is 0V...25mV and 4.975V...5.0V for a 100k load and 0V...125mV and 4.875V...5.0V for a 5k load:
Stay away from the no man's land, if you expect proper performance of OPAmp.
Kai
