Why is there a 50mV difference between the two input terminals of an op amp working in negative feedback?

sfcsdc

Why is there a 50mV difference between the two input terminals of an op amp working in negative feedback? [Copy link]

 

As shown in the figure, the test voltage is marked on the figure

The voltage on the left side of 330K becomes lower when it reaches the op amp's non-inverting input, but the voltage at the feedback-regulated inverting input is correct.

The offset voltage of 358 itself is not that large, and it will be the same if you change to other op amps.

The input impedance of the op amp itself is very large, but different output voltages will affect the voltage at the common-mode input terminal.

What is the problem here?

maychang

"The offset voltage of the 358 itself is not that large. It will be the same if I change to other op amps."

What instrument did you use to measure this? A multimeter?

If it is a multimeter, what is the input resistance of its voltage range?

maychang

Why is there a 50mV difference between the two input terminals?

From the markings in the figure, it can be seen that the difference is 40mV. Moreover, the difference of 40mV is also the result of two measurements, not the result of one measurement (one measurement refers to the measurement using a certain instrument connected between the two input terminals of the op amp).

chunyang

The internal resistance of the meter is not enough, the current drawn is too large, and a voltage drop occurs on R18.

sfcsdc

It is true that after removing the 330K resistor, everything is back to normal.

The equivalent output resistance of the non-inverting input and the inverting input is different.

But I use the oscilloscope to see that the two channels are different

The input impedance of the oscilloscope and multimeter should be relatively large.

But is this a problem with the op amp or a problem with the test?

maychang

sfcsdc posted on 2022-4-7 09:43 It is true that after removing the 330K resistor, it is normal. The equivalent output resistance of the non-inverting input terminal and the inverting input terminal is different. But I use the indicator...

"The input impedance of the oscilloscope and multimeter should be relatively large."

The input impedance of an oscilloscope is generally 1 megohm (at the panel socket), and the voltage range of a digital multimeter is usually 10 megohms. Whether 1 megohm or 10 megohm, compared with 330 kilohms, it cannot be ignored.

maychang

sfcsdc posted on 2022-4-7 09:43 It is true that after removing the 330K resistor, it is normal. The equivalent output resistance of the non-inverting input terminal and the inverting input terminal is different. But I use the indicator...

"But is this an op amp problem or a test problem?"

Clearly a testing issue.

The LM358 input bias current is about 50nA, and can reach up to 100nA. The bias current of the op amp must be taken into account when testing. If the bias current is not taken into account, then there is something wrong with the test method.

maychang

sfcsdc posted on 2022-4-7 09:43 It is true that after removing the 330K resistor, it is normal. The equivalent output resistance of the non-inverting input terminal and the inverting input terminal is different. But I use the indicator...

In the first circuit, the left side of R18 is 1.548V to ground, the right side is 1.508V to ground, the voltage across R18 is 40mV, and the current through R18 is 40mV/330000 ohms = 121nA. It can be seen that the bias current of the op amp cannot be ignored here.

maychang

sfcsdc posted on 2022-4-7 09:43 It is true that after removing the 330K resistor, it is normal. The equivalent output resistance of the non-inverting input terminal and the inverting input terminal is different. But I use the indicator...

In the first circuit, the voltage on the right side of R18 to ground is 1.508V. Assuming that the internal resistance of your multimeter voltage range is 10 megohms, the current through the multimeter is 1.508V/10000000 ohms = 151nA, and the direction is from the op amp non-inverting input to ground.

The difference between 151nA and 121nA passing through R18 is 30nA, which is the bias current of LM358 (the bias current of the LM358 op amp input flows from inside the chip to outside the chip).

maychang

sfcsdc posted on 2022-4-7 09:43 It is true that after removing the 330K resistor, it is normal. The equivalent output resistance of the non-inverting input terminal and the inverting input terminal is different. But I use the indicator...

From the calculation results on the 9th post, there is nothing wrong with your LM358 op amp. So the answer to the question on the 5th post, "But is this a problem with the op amp or a problem with the test?" is a problem with the test method.

maychang

sfcsdc posted on 2022-4-7 09:43 It is true that after removing the 330K resistor, it is normal. The equivalent output resistance of the non-inverting input terminal and the inverting input terminal is different. But I use the indicator...

"The equivalent output resistance of the non-inverting input terminal and the inverting input terminal is different."

I don't know where to start. How can there be an "equivalent output resistance" at the non- inverting input or the inverting input ?

maychang

sfcsdc posted on 2022-4-7 09:43 It is true that after removing the 330K resistor, it is normal. The equivalent output resistance of the non-inverting input terminal and the inverting input terminal is different. But I use the indicator...

In the first circuit, you use a multimeter to measure the voltage between the non-inverting input terminal and the ground. The current direction in the multimeter is from the non-inverting input terminal to the ground, causing the current in R18 to flow from left to right.

If you don't use a multimeter to measure, the direction of the current in R18 is from right to left (for your specific circuit, the current is 30nA), because the LM358 bias current flows from inside the chip to outside the chip .

All the confusion in this picture is caused by the fact that you used a multimeter to measure a high internal resistance circuit. Isn't this a test problem?

sfcsdc

maychang posted on 2022-4-7 11:58 "The equivalent output resistance of the non-inverting input terminal and the inverting input terminal is different." I don't know where to start. The non-inverting input terminal or the inverting input terminal, how...

Here is what I think:

For a multimeter, the equivalent output resistance of pins 5 and 6 of U3B in the measurement diagram is different.

Pin 5 is 330K, and pin 6 is around 10K.

The input resistance of the multimeter is not ideally infinite.

So in fact, the same voltage is measured at pin 5 as 1.508V and at pin 6 as 1.548V

maychang

sfcsdc posted on 2022-4-7 13:57 This is what I think: For a multimeter, the equivalent output resistance of the 5th and 6th pins of U3B in the measurement diagram is different. The 5th pin is 330K, ...

Here is what I think:

For a multimeter, the equivalent output resistance of pins 5 and 6 of U3B in the measurement diagram is different.

Insert the word " power supply " into "equivalent output resistance" and it will become "the output resistance of the equivalent power supply is different".

zlk1999

Which one should be taken as the standard?

maychang

zlk1999 posted on 2023-6-21 13:42 So which one should be taken as the standard'

[Which one should be taken as the standard?]

The result with the smaller system error shall prevail. When measuring the inverting input of U3B, the source impedance is 10 kilo-ohms, and when measuring the non-inverting input, the source impedance is 330 kilo-ohms. Obviously, the measurement result of the inverting input shall prevail.