Opa2350 negative half of voltage

[Midgard]

I am using opa2350 as a current sensor in inverter as that combination:

and when output voltage is >20 VAC i have problem with sensing current in negative sinusoidal:

Positive half looks quite okay. I try it with diffrent shunt resistor value, but the problem stay always for negative voltage.
Why my opamp have trouble with negative half?

[StillTrying]

You'd see more with the scope on DC coupled.

I think the input phases of the op amps are the wrong way around, they'd be better with the +ve input going to the top of the 0.04Rs to give a +ve going output with increasing 0.04R current.

[SiliconWizard]

I think the input phases of the op amps are the wrong way around, they'd be better with the +ve input going to the top of the 0.04Rs to give a +ve going output with increasing 0.04R current.

Agreed.

[Midgard]

Thank you for answers. I switched 20k resistors with 1k and the output is a lot better.

You'd see more with the scope on DC coupled.

When I turn on DC coupling my signal is moving out of range of oscilloscope.

[StillTrying]

"I switched 20k resistors with 1k and the output is a lot better."

You reduced the gain from -20 to -1 ?
We were suggesting changing the gain to +21 by swapping the 1k inputs on the 0.04Rs so that the op amp outputs rise towards 3V with more current rather than 0V.

You need to know the maximum current to be measured through the 0.04Rs to set the op amp gain to give about 0-3V output, or whichever voltage range you need from the op amp output.

"When I turn on DC coupling my signal is moving out of range of oscilloscope."

You can move the channel's 0V position to the bottom of the screen as the outputs from the op amps are all only positive(0-3V).

[Midgard]

"I switched 20k resistors with 1k and the output is a lot better."

You reduced the gain from -20 to -1 ?
We were suggesting changing the gain to +21 by swapping the 1k inputs on the 0.04Rs so that the op amp outputs rise towards 3V with more current rather than 0V.

I mean i swap the resistor 20k with 1k and 1k with 20k. Sorry for not accurate explanation.

"
You need to know the maximum current to be measured through the 0.04Rs to set the op amp gain to give about 0-3V output, or whichever voltage range you need from the op amp output.

How to define maximal current on shunt resistor?

[MarkF]

Is there a reason for using differential input configuration?
I agree with others.  It looks like you are inverting the signal and trying to correct for a negative output be raising the non-inverting reference.

Why not just use a non-inverting configuration with a gain of 20?

[Midgard]

Is there a reason for using differential input configuration?
I agree with others.  It looks like you are inverting the signal and trying to correct for a negative output be raising the non-inverting reference.

Why not just use a non-inverting configuration with a gain of 20?

(Attachment Link)

Wouldn't be affecting a Bias current error?

"I switched 20k resistors with 1k and the output is a lot better."

You reduced the gain from -20 to -1 ?
We were suggesting changing the gain to +21 by swapping the 1k inputs on the 0.04Rs so that the op amp outputs rise towards 3V with more current rather than 0V.

You need to know the maximum current to be measured through the 0.04Rs to set the op amp gain to give about 0-3V output, or whichever voltage range you need from the op amp output.

"When I turn on DC coupling my signal is moving out of range of oscilloscope."

You can move the channel's 0V position to the bottom of the screen as the outputs from the op amps are all only positive(0-3V).

Now I have another trouble. I can't probe the output of opamp correctly. It looks like this (Pink channel):

I try to solder different capacitors at top of feedback resistor but it won't help. Any ideas?

[MarkF]

Show us your new circuit and where you are probing.

Use DC coupling on the scope.  Can't see the actual DC level of the signal.

[Midgard]

Show us your new circuit and where you are probing.

My new circuit (I probe on blue circles):

Use DC coupling on the scope.  Can't see the actual DC level of the signal.

[MarkF]

You have an inverting configuration with a fixed voltage of 1.57V at the +In of each op-amp
With a gain of 1K / 20K = 0.05



What are you expecting?  Surely, you want to have some gain?
I still think you should use my previously posted circuit.
What are the voltages at RP1 and RP2?  What is the max voltage you expect across these resistors?

[MarkF]

The circuit you have is probably NOT what you think it is?
Look at it redrawn.

[Zero999]

Wouldn't be affecting a Bias current error?

The OPA2350 has a bias current of just 0.5pA typical, 10pA maximum and will cause an additional offset voltage of under 10nV, with an input resistance of 1k in parallel with 20k, which is tiny compared to the input offset voltage of 500µV.