Op amps with feedback on both input terminal

[dann11]

I just saw an op amps configuration with feedback at both input terminals, what configuration it is? in what application does it is use.. thanks..

 

[KlausST]

Hi,

where did you see it?

Isn´t there a discription?

I don´t think it makes sense .. at least on first sight.

Klaus

 

[FvM]

No it doesn't make sense. It could work as a current source - with the output at a different node.

I suspect a trivial drawing fault, R2 could be connected to +5V instead of the OP output.

 

[KlausST]

Hi,

Yes, then it is a differential amplifier (voltmeter) with Vcc/2 offset at opamp output.
Able to measure +/- voltage = AC voltage.
Makes sense.

Klaus

 

[AMS012]

As long as the negative feedback factor is more than the positive feedback factor, you can hook up circuits like that (assuming a stable circuit required and not an oscillator). This might help you in that regard https://www.maximintegrated.com/en/app-notes/index.mvp/id/1868

 

[FvM]

As long as the negative feedback factor is more than the positive feedback factor, you can hook up circuits like that (assuming a stable circuit required and not an oscillator).

True, but the circuit as shown would be purposeless, though. An unloaded Howland current source would have balanced positive and negative feedback, but as said, a different output connection.

 

[Audioguru]

Another problem with the schematic of the differential amplifier is that since it has a single positive supply and has its inputs biased at half the supply voltage then input coupling capacitors to block the DC are missing and an output coupling capacitor might also be needed.

 

[dann11]

but commonly, what configuration that uses a feedback network on both the input terminal of the op amps? and in what application are they used? Thanks.

 

[Audioguru]

We think the schematic you found is drawn wrong.

 

[FvM]

but commonly, what configuration that uses a feedback network on both the input terminal of the op amps?

Not related to your schematic, but a Howland current source does.

 

[Dan Mills]

Also negative impedance converters, gyrators, and simple hysteresis oscillators.

Also some filter topologies, biquads and such.

Regards, Dan.

 

[Eshal]

I want to be the part of this discussion.

As you all said, the schematic is drawn wrong. But I want to ask why? In post#5 of @AMS012, he has given link in which opamp has positive and negative feedback simultaneously. And @dann11's circuit has also both feedbacks simultaneously. The one which is maxim's is correct and @dann11's is wrong. Why why why??

Regards,
Princess

 

[Audioguru]

The first circuit posted here is wrong because it has two inputs like a differential amplifier but wrongly has positive feedback on one of the inputs.
Maxim's circuit has only one input.

 

[LvW]

I want to be the part of this discussion.
As you all said, the schematic is drawn wrong. But I want to ask why? In post#5 of @AMS012, he has given link in which opamp has positive and negative feedback simultaneously. And @dann11's circuit has also both feedbacks simultaneously. The one which is maxim's is correct and @dann11's is wrong. Why why why??
Regards,
Princess

Who said it would be wrong? How a circuit can be "wrong"? This can be decided only if compared with a specific task.
Either a circuit can fulfill its task or it can`t.
Here the main question was: What is the purpose of the circuit?

In general: For linear operation, positive feedback is possible and allowed if - at the same time - we also have negative feedback that dominates over positive feedback (neg. feedback factor>pos. feedback factor).

 

[AMS012]

I think the circuit posted in the first post should work fine. Though it has two inputs, isn't it just the superposition of the two inputs?

 

[Audioguru]

I simulated the circuit in the first post:
1) With both inputs having the same sinewave of 10V peak, the output was 1.8V peak so it is an attenuator with a high input impedance and a very low output impedance.
2) With the 1M inverting input resistor grounded and a 10V peak sinewave at the 1M non-inverting input resistor the output was the same as the input except the impedances are transformed again.
3) With the 1M non-inverting input resistor grounded and a 10V peak sinewave at the 1M inverting input resistor the output was about 1.18V peak.

 

[FvM]

I'm keeping my point the post #1 circuit is useless as shown. If anybody sees a useful application of this particular circuit (with the given resistor values, input- and output connections etc.), please let me know.

The circuit is a stable amplifier because positive feedback is only half of negative (I overlooked this point previously), but you can achieve the same gain with negative feedback only. So why would you design-in positive feedback?

 

[Eshal]

Positive feedback the biggest problem as all discussion says. However, I understand @Audioguru @FvM says.

 

[vGoodtimes]

I think the circuit posted in the first post should work fine. Though it has two inputs, isn't it just the superposition of the two inputs?

If assumptions about the opamp are met. (or if the opamp has no internal delays...) In a practical design you would need to ensure that this can't become a latch. FvM has shown this for your case.

I'm trying to figure out if there is a valid use case based on the specific opamp used. Seeming pointless designs like this sometimes make sense when other factors are included. I don't see one yet though.