Lutz von Wangenheim: "As another consideration, it is necessary to add that the classical active opamp implementation (NIC) allows two alternatives:

NIC that is

* short-circuit stable, or

* open-circuit stable.."

Lutz, would you explain what you mean... maybe INIC and VNIC?

Yet what is NIC?

Lutz, I will repeat here the simple truth about true negative resistors (NICs) - they are just sources... self-variable sources... functional converters: VNIC is a 1-port current-to-voltage functional converter (voltage source - see the attached picture); INIC is a 1-port voltage-to-current functional converter (current source). The function is set by a sample "positive impedance" element (resistor, capacitor, diode...) inside the NIC...

I just can not believe it - three months after I asked this question there is no any (at least one!) reaction to it! And that about the negative impedance converter - the most mysterious, unexplainable, mystic electronic circuit...

Maybe I am wrong? Maybe this circuit is trivial and is explained in detail in many sources? Then be pleased to list them here:

1. .............................

2. .............................

3. .............................

Two days ago, I (Circuit dreamer) began discussing the sophisticated INIC circuit in Wikipedia arguing with my main opponent (Zen-in):

https://en.wikipedia.org/wiki/Talk:Negative_impedance_converter#Questions_on_this_circuit_description

https://en.wikipedia.org/wiki/User_talk:Circuit_dreamer#A_discussion_about_NIC_and_TIA

Perhaps these discussions will awaken your interest in the mystic circuit?

Well, let's consider the discussion... First, it will show what orthodox Wikipedians are and how they think; second, it will reveal more secrets of NICs. It all started after my opponent decided finally to begin thinking about the circuit of NIC:

"I think a circuit analysis (math) is essential, especially with a speculative circuit like this. These NIC circuits all seem to ignore some basic properties of opamps and rules about circuit analysis. For example there is an element that is both a current source and a voltage source but there is nothing that says what the relationship is between the two. There is an independent I and V so I suspect the analysis is incorrect. The reference cited on the negative resistance page has a date 1953. Are there any more recent references or any evidence someone has actually built this circuit and demonstrated the claimed negative resistance effect? ...I'm not convinced the circuit would really work as stated... " And I replied as follows to these confused thoughts:

There is nothing speculative in this circuit (see the attached picture below). It is based exactly on the "basic properties of opamps and rules about circuit analysis". To show it, here is another ("functional") analysis based on the properties of the classic non-inverting amplifier and the two versions of the Ohm's law (I = V/R and V = I.R).

The op-amp keeps up the voltage drop VR1 across R1 equal to the input voltage VS (the op-amp acts as a voltage-to-voltage converter or voltage follower) by passing a current I2 = VR1/R1 = VS/R1 through the resistor R2 (so R1 acts as a voltage-to-current converter). The current I2 creates a voltage drop VR2 = I2.R2 across R2 (so R2 acts as a current-to-voltage converter). The op-amp keeps up the voltage drop across R3 equal to the voltage drop across R2 (the op-amp acts as another voltage-to-voltage converter or voltage follower) by passing a current IS = VR3/R3 = VR2/R3 = (I2.R2)/R3 = ((VS/R1).R2)/R3 through the input source. So, the input resistance is really -R3.R1/R2.

If R2 = R3 (the usual case), the circuit injects the same current IS = -I2 that would be drawn by the resistor R1 if it was connected directly to the input source. So, it behaves as a "negative resistor" R1 having the same voltage as the positive R1 but with an inverted current; thus the name of the circuit - "negative impedance converter with current inversion" (INIC). The circuit "inverts" every positive/negative element (resistor, capacitor or inductor) connected in the place of the resistor R1 to the "opposite" negative/positive element with equivalent impedance; it is just a "current inverter" (actually, the very INIC consists of the two resistors R2 and R3, and the op-amp). According to this explanation, the direction of IS in the figure should be the opposite (the current should enter the input source when it produces a positive voltage).

"The circuit would really work as stated" (would be stable) if the internal resistance Ri of the input source is low enough; in the case of an "ideal" voltage source (Ri = 0), it is unconditionally stable. More precisely, the inequality Ri/(Ri + R3) < R1/(R1 + R2) must be satisfied so that the negative feedback dominates over the positive one.

What do you think about these my insights?

Well, it is seen that I will have to talk with myself even during the Christmas holidays ...

It is obvious that both intuitive and mathematical approach are useful and necessary when analyzing circuits, and they should be applied in this sequence. I have a very good intuitive understanding of these kinds of circuits and I want to share it with you; that is why I continue elaborating this discussion. I have also considered step-by-step the INIC's operation in a Wikibooks story. I have implemented this story more like a "movie" with separate "frames" but the text explanations are few:

https://en.wikibooks.org/wiki/Circuit_Idea/Linear_Mode_of_Current_Inversion_NIC

To understand this circuit, it is crucial to see where and in what direction currents flow. Note that in this simple case (an ideal input voltage source with Ri = 0) the circuit is nothing else than a non-inverting amplifier. Of course, there is a resistor connected between the op-amp output and its non-inverting input but it does not affect the operation of the non-inverting amplifier since there is no positive feedback and only the ideal voltage source sets the voltage of the non-inverting input. In practice, the input source has some internal resistance; then there is a positive feedback and the circuit becomes even more interesting...

Generally speaking, the op-amp keeps up the voltage drop VR2 across R2 equal to the input voltage VIN (the op-amp acts as a voltage-to-voltage converter or voltage follower) by passing a current IF = VR2/R2 = VIN/R2 through the resistor R1 (so R1 acts as a voltage-to-current converter). The current IF creates a voltage drop VR1 = IF.R1 across R1 (so R1 acts as a current-to-voltage converter). The op-amp keeps up the voltage drop across R equal to the voltage drop across R1 (the op-amp acts as another voltage-to-voltage converter or voltage follower) by passing a current IOUT = -VR/R = -VR1/R = -(IF.R1)/R = -((VIN/R2).R1)/R through the input source. So, the input resistance is really RIN = -R.R1/R2.

If R1 = R2 (the usual case), the circuit injects the same current IOUT = -IF that would be drawn by the resistor R2 if it was connected directly to the input source. So, it behaves as a "negative resistor" R2 having the same voltage as the positive R2 but with an inverted current; thus the name of the circuit - "negative impedance converter with current inversion" (INIC). The circuit "inverts" every positive/negative element (resistor, capacitor or inductor) connected in the place of the resistor R2 to the "opposite" negative/positive element with equivalent impedance; it is just a "current inverter" (actually, the very INIC consists of the two resistors R and R1, and the op-amp). The current enters the input source when it produces a positive voltage instead to exit it as in the case of a positive resistance...

Let’s now see in detail how it all happens...

In the beginning, the input voltage is zero... the op-amp output voltage is zero... all the circuit voltages are zero... so there are no currents flowing... Now we begin increasing the input voltage VIN. The output voltage VIN.(R1 + R2)/R2 increases as well. A current begins flowing starting from the positive supply rail, then enters the positive supply pin of the op-amp and exits the op-amp output. Then this current splits into two currents flowing through the left and the right parts (legs) of this (bridge) circuit. The op-amp will stop changing its output voltage when reaching the equilibrium - V(-) = V(+). In this simple case (ideal voltage source), the op-amp achieves its goal only changing the voltage at the inverting input; but when the source has some internal resistance, the op-amp changes both the voltages (of the non-inverting and non-inverting inputs). The current produced by the circuit enters the input source when its voltage is positive... and this is the INIC idea - to obtain negative resistance by inverting the current. Thus both the "half currents" exit the op-amp output and flow down through the two legs to the ground...

The circuit would be stable if the internal resistance Ri of the input source is low enough; in the case of an "ideal" voltage source (Ri = 0), it is unconditionally stable. More precisely, the inequality Ri/(Ri + R) < R2/(R1 + R2) must be satisfied so that the negative feedback dominates over the positive one.

In the beginning of this year, I asked a question closely related to negative impedance converters:

https://www.researchgate.net/post/What_is_Wikibooks_compared_to_Wikipedia_Can_we_believe_cite_it_and_to_what_extent_I_suggest_you_find_out_that_by_joining_live_in_Wikibooks

There I suggested to develop live together the page about NIC both here, in ResearchGate, and in Wikibooks:

https://en.wikibooks.org/wiki/Circuit_Idea/Negative_Impedance_Converter

For this purpose, today I prepared a number of conceptual illustrations to this page to use them as a "backbone" to reveal the ideas behind NICs. I will redraw them on the whiteboard; then shoot them and place in the Wikibooks and ResearchGate stories about NICs...

Yesterday I began creating the Wikibooks story (a module of the Circuit Idea wikibook) about negative impedance converters...

https://en.wikibooks.org/wiki/Circuit_Idea/Negative_Impedance_Converter

... and started also the first "lesson" for the beginning Wikibookeans from RG ("ResearchGaters" contributing the discussions here and (possibly) editing the wikibook story:)

https://www.researchgate.net/post/What_is_Wikibooks_compared_to_Wikipedia_Can_we_believe_cite_it_and_to_what_extent_I_suggest_you_find_out_that_by_joining_live_in_Wikibooks

My intention there was to discuss the general questions about Wikibooks, Wikipedia and the relation between them... also, to consider the general questions related to the creating, editing and maintaining Wikibooks articles, particularly the story about negative impedance.

Then there we (I) chose an intriguing title to provoke the interest of web visitors, formulated the basic NIC idea, wrote the lede (intro) and uploaded the first image at the top of the page (the Wikibooks Lesson 2:)...

Now we can begin developing the very story here by opening the wiki editor of the first section:

https://en.wikibooks.org/w/index.php?title=Circuit_Idea/Negative_Impedance_Converter&action=edit§ion=1

First, I think, to make a connection with the more general Wikibooks story about negative impedance...

https://en.wikibooks.org/wiki/Circuit_Idea/Revealing_the_Mystery_of_Negative_Impedance

... we should ask the general question, "What is negative impedance converter?"... and it will be the title of our first section (for this purpose, you must enclose it with two equal signs). The answer is simple; we take it from the article above - It is just an op-amp implementation of a true negative resistor. But now the next question arises, "What is true negative resistor? We can take and adаpt again the explanations from there:

"It is the opposite "element" of the ordinary "positive" resistor; it is a circuit adding (injecting) the same energy that the equivalent "positive" resistor would dissipate. So, it is nothing else than a source... but this is not the ordinary constant source; it is a "self-varying" (dynamic) source. And as there are two kinds of sources (in contrast with the only one kind of "positive" resistor), there are two kinds of true negative resistors (S-shaped and N-shaped), accordingly, there are two kinds of their op-implementations (VNIC and INIC) as well. First, a NIC can behave as a dynamic voltage source producing voltage that is proportional to the current passing through it (named voltage inversion NIC or VNIC) or as a dynamic current source producing current that is proportional to the voltage across it (named current inversion NIC or INIC). You can consider the VNIC as a 1-port current-to-voltage (to voltage, not to voltage drop!) converter and INIC - as a 1-port voltage-to-current converter."

Then we can illustrate the text above with two conceptual pictures about the two kinds of (true) negative "resistors" (they are actually sources). In their description, we place links to them, set the positions (left for the one and right for the other), sizes (280px) and captions:

[[Image:Current-driven_neg_resistor_idea_1000.jpg|left|thumb|280px|Fig. 1a. S-shaped (current-driven) negative "resistor".]]

[[Image:Voltage-driven_neg_resistor_idea_1000.jpg|right|thumb|280px|Fig. 1b. N-shaped (voltage-driven) negative "resistor".]]

I continue developing the page with the hope that someone will support me... So, the next reasonable question is, "How do we create negative impedance converters?"

https://en.wikibooks.org/wiki/Circuit_Idea/Negative_Impedance_Converter#How_to_create_negative_impedance_converters

We can derive the general idea from our rutine where we frequently create some quantity by inverting its opposite quantity. In the general negative impedance WB story, I have shown a few funny (and a bit fabricated:) examples from the life, e.g.

https://en.wikibooks.org/wiki/Circuit_Idea/Revealing_the_Mystery_of_Negative_Impedance#Looking_for_the_basic_S-shaped_NR_idea

So, we can obtain negative resistance (-R) by inverting the equivalent "positive" resistance (R). But how? The answer is simple if only we know the Ohm's law:) It presents the resistance as a ratio between the voltage and the current (R = V/I); so when the two variables are positive, the resistance is positive as well. So, to make negative resistance, we have to invert one of them - the voltage or the current...

1. INVERTING THE VOLTAGE POLARITY. Well, let's first invert the voltage; thus we will obtain the so called "S-shaped" (current-controlled) negative resistance RS = -V/I = -R. This means that if we pass a current through the S-shaped negative resistor, the input terminal becomes negative (instead positive as in the case of the ordinary "positive" resistor). That is why, circuits implementing this technique are named "voltage-inversion negative impedance converters" (VNIC). Note the power is also inverted (PS = -V.I = -P). Let's insert this text in the story...

https://en.wikibooks.org/wiki/Circuit_Idea/Negative_Impedance_Converter#Inverting_the_voltage_polarity

... and illustrate this arrangement with an extremely simple but useful for understanding picture...

2. INVERTING THE CURRENT DIRECTION. Then, we can invert the current; thus we will obtain the so called "N-shaped" (voltage-controlled) negative resistance RN = V/-I = -R). This means that if we apply positive voltage across the N-shaped negative resistor, the current goes out of the negative resistor and enters the positive terminal of the voltage source (instead to leave the positive terminal of the voltage source and to enter the negative resistor as in the case of the ordinary "positive" resistor). That is why, circuits implementing this technique are named "current-inversion negative impedance converters" (INIC). Note the power is also inverted (PN = V.-I = -P). Let's insert this text in the story...

https://en.wikibooks.org/wiki/Circuit_Idea/Negative_Impedance_Converter#Inverting_the_current_direction

... and illustrate this arrangement with a dual "simple but useful" picture...

But it is interesting to see how the "competitive" Wikipedia page "explains" the fundamental NIC ideas. We skip the lede where I (Circuit dreamer) have written a few introductory lines about NICs and go to the first section (Basic circuit and analysis):

https://en.wikipedia.org/wiki/Negative_impedance_converter#Basic_circuit_and_analysis

What we can see, is only the final op-amp circuit and a fiormal analysis; there are no human-friendly explanations... there is no answer to the most elementary questions...

As I can see, there are no any objections to the written above; so we can continue... We have to invert the voltage or current... but how? Let's first see how to invert the voltage...

When the input current Iin flows through the positive resistor R, it creates a voltage drop VR = IIN.R (see the attached conceptual picture). We can use this voltage to drive an additional "helping" voltage source (VCVS, on the right in the picture) so that to produce the inverted voltage VOUT = -IIN.R. But as the voltage drop VR will subtract from this voltage, it has to be two times higher (2VOUT) so that the resulting voltage VOUT = -IIN.R across the whole "resistor" will be the same but inverted as the initial voltage across the resistor.

So the trick is to add two times higher negative voltage to the initial positive voltage drop with the purpose to create a negative voltage. BTW we can see this solution around us when we invert some (usually "bad") quantity by adding a bigger opposite ("good") quantity. Agreed? If so, let's write it in the subsection below...

https://en.wikibooks.org/wiki/Circuit_Idea/Negative_Impedance_Converter#Inverting_the_voltage_polarity

... and illustrate this arrangement with the attached picture...

Let's how do the same with the current...

Now we can invert (reverse) the current direction by connecting an additional "opposing" voltage source in series, which voltage is two times higher than the input voltage. As a result, the same but opposite current enters back in the input source.

So the trick is to add two times bigger reverse current to the initial direct current with the purpose to create a "negative" current. Agreed? If so, let's write it in the subsection below...

https://en.wikibooks.org/wiki/Circuit_Idea/Negative_Impedance_Converter#Inverting_the_current_direction

... and illustrate this arrangement with another picture...

But an interesting question arises now... In this arrangements, actually we have connected two elements in series - a "positive" resistor and a negative "resistor" (the varying voltage source on the right). The negative element dominates and the resulting resistance is negative. The question is, "Why have we connected two kinds of resistance - positive and negative, after we ultimately want only negative? Why do not we just use the negative resistance? For what we need the positive resistance?

No answer, no vote (even negative)... no any interest in the most interesting electronic circuits... Obviously, ResearchGate is not the place where to discuss great circuit ideas... Okay, I will still finish this "discussion" and then I will continue to develop it in Wikibooks...

Hi Cyril - I only can speak for myself. However, there are as much as 20 (twenty!) succeeding contributions from your side - in some cases filled with drawings that are hard to read (and to comment). And - in addition - a lot of links to wikipedia or to wikibooks. That´s too much!

As a consequence, it was not easy to jump into a discussion and to respond with questions or with answers or even with own contributions. I simply did not know where to start. Can you understand?

Regards

Lutz

Exactly Lutz, I understand... Of course, everyone has the right to respond or not to respond to my questions, I do not force anyone to do so. I guessed, just looking at all of these sections in the field of electronics including hundreds of thousands members (and that scientists and specialists), and given the unexplained problems of these circuit decisions, that my questions will interest them. And I started outlining the problem gradually expecting some feedback (at least by the people who know me and working together for a year). It may be positive, negative, whatever is... just I can not accept to have no any feedback...

Cyril - your first three posts on this subject were written between end of August and beginning of September. I was off (vacation) until end of September.

Thus, I missed the start of this thread - perhaps you can accept this as something like an "excuse". I fully understand that you are somewhat disappointed - however, if you reread your first post you will notice that it is rather long and it is not very easy to "extract" the "honey" (that means: the real question) from it.

Good night (for today)

Lutz

But let's get back to the essential things in circuitry... I will continue the above "scenario" in pictures (sort of like comics) that can help us to build in a logical way the two versions of NIC (VNIC and INIC).

Lutz, much of the ideas below were inspired by your question about the Antoniou`s GIC circuit. In fact I got inspiration from there, but I struggled with the full awareness of the CIC; so I decided to go back and fully realize NICs... and then try to do this with GICs...

We can realize the ideas presented in the above pictures first by using fixed gain amplifiers. Thus we can build a VNIC by a differential "inverting" amplifier (why?) having a gain of 2...

... and an INIC - by a non-inverting amplifier having a gain of 2.

But nowadays we use op-amps to realize precise fixed-gain amplifiers. So, we can build a voltage-inversion op-amp NIC (VNIC) by making the little exotic "inverting" amplifier below - K = (R1+ R2)/R2...

... and a current-inversion op-amp NIC (INIC) by making the classic non-inverting amplifier below - K = (R1+ R2)/R1...

Yes Lutz, it was not very easy to "extract" the "honey" about this 2-resistor op-amp circuit... and your question about GIC helped me. On the other hand, I to a large extent complicate the situation deliberately to stimulate thinking and reasoning that are so necessary for such an interesting discussion...

I continue exposing the "scenario"... I will briefly outline it to to leave room for discussion...

The next powerful idea is to think of the op-amp NIC as of a balance bridge. Let's see the two conceptual (equivalent) circuits - first, of a VNIC...

... and second, of an INIC:

According to them, the two op-amp implementations can be drawn - first, of a VNIC...

... and second, of an INIC:

Thus we can discern well-known functional building blocks (converters) and to draw two functional block diagrams - of a VNIC...

... and of an INIC:

Finally, it is interesting to see what element (of the total three elements, e.g. resistors) is actually "inverted" - in the case of a VNIC...

... and in the case of an INIC:

Cyril - sorry, this is not a comment to the NIC subject.

It is - more or less - a kind of "consolation" for you.

On the 5th of August I have given a link to an article I have written which gives an improvement/extension of the well-known Barkhausen oscillation criterion:

(See article: "On a rigorous oscillation condition")

https://www.researchgate.net/publication/255171263_On_a_rigorous_oscillation_condition

Of course, I did ask for (and I have expected some) comments because I have proposed a more rigorous criterion with the aim to come closer to a crierion that is sufficient (instead of being "only" necessary like Barkhausen).

The result was: ZERO response.

This was really surprising for me.

Data On a rigorous oscillation condition

Lutz, your article is too theoretical for me and so I could not react to it, if I could, I would do it with pleasure. Then I was looking for intuitive explanations of Wien bridge oscillator but reached an impasse, and the same happened later with the GIC circuits. So I went back to explain the most basic circuits with negative resistance - NICs...

Dear Prof. Cyril Mechkov,

I got interested in this topic since it related to my small current research. I hope to make a study under your guidance and supervision. (to make a review paper which really need your supports). (because in my view, it is just only an interesting topic if it related to some on-doing works)

If you concern, please let me know. My email address is: [email protected]

Thank you so much

Sincerely,

Nguyen Ha

Dear Van,

I am interested in circuit phenomena and in particular in NICs, so far as to be able to reveal the basic idea behind them... and stop here...

Thank you very much for your reply. Your works here help me a lots. 

On Page 3, I began telling a funny story about this "devilish circuit":) My guess was that we can think of it as of a balanced bridge. If it is possible, we can say that it is invented in the middle of 19th century by Sir Charles Wheatstone:)

Here I will elaborate in details this story. I will do it in a form of a 4-step scenario where I will consider in parallel bridge and op-amp versions. Thus the electrical implementation of the 19th century (Wheatstone's bridge) and the electronic implementation of the 21st century (NIC) as though will be connected with some kind of a "bridge" through time.

With this I want to demonstrate the power of "pure ideas" when they are released from the bondage of the specific circuit implementations (tube, transistor, op-amp, etc...), technology constraints and all kinds of imperfections. The pure idea does not need modern technologies... or sophisticated math analysis and simulations... It is clear and simple... eternal and immortal... To understand how a circuit operate, you have to see on which pure idea this circuit is based... otherwise you have not understood the circuit... you only know the concrete circuit implementation...

The pure idea can be materialized in many possible ways... and this electronic implementation is only one of them. It could be realized, for example in a mechanical form, many years ago (e.g. by Archimedes, centuries BC:) I would gladly consider such a "lever implementation"... and if you want, I will actually do it...

Step 1. We assemble a bridge circuit by four resistors (R1-R4). It is convenient to think of it as of a combination of two voltage dividers - R1-R2 (the left leg) and R3-R4 (the right leg). Let R1 = R3 and R2 < R4.

In the 2nd century BC lever implementation........................................

• In the 19th century electrical implementation (the first picture below), we supply the bridge by a varying voltage source OA, and observe its output voltage by a zero indicator (galvanometer). Our job is to balance the bridge by varying the supply voltage.
• In the 21st century electronic implementation (the second picture below), an op-amp OA supplies the bridge by its output voltage, and "observes" the bridge output voltage by its differential input. Its job here is the same - to balance the bridge by varying the supply voltage.

Step 2. We begin to fulfill your obligations...

• In the 2nd century BC lever implementation........................................
• In the 19th century, when we try to increase the supply voltage VOA, both output voltages of the voltage dividers begin increasing. But since we have chosen R2 < R4, VR4 increases more vigorously than VR2... this makes us decrease the supply voltage... and, as a (quite surprising) result, we do not change the supply voltage... it remains zero. VR2, VR4 and their difference dV also remain zero. Figuratively speaking, the whole circuit is "dead"... there are no voltages across and currents through resistors... but still it is balanced.
• In the 21st century, the op-amp does exactly the same... When the power supply is turned on, its output voltage VOA tries to change (e.g., increase)... and the voltages of its inputs begin increasing. But since R2 < R4, V- increases more vigorously than V+ (as they say, the negative feedback dominates over the positive one). This makes it decrease its output voltage... and again the result is thaty the output voltage VOA remains zero... V+, V- and their difference dV also remain zero. The whole op-amp circuit is "dead"... there are no voltages across and currents through resistors.

Step 3. It is time to break the "bliss" of these well balanced bridge circuits, e.g. by inserting a positive voltage source VIN in the lower part of the left leg.

• In the 2nd century BC lever implementation........................................
• In the 19th century, the output voltage VR2 of the left voltage divider will increase (since VIN adds to VOA by means of the resistive summer R1-R2). We react to this intervention by increasing VOA. Our will is to increase only VR4... but unfortunately, VR2 also changes... but luckily for us, more slowly. Thus finally, we manage to balance the bridge circuit.
• In the 21st century, the op-amp "sees" that V+ increases, and reacts to this intervention by increasing its output voltage VOA (with the purpose to increase V-). Unfortunately for it, V+ also changes... but luckily for it, more slowly. Thus finally, due to the dominating negative feedback, the op-amp manages to balance the bridge circuit.

Step 4. With the same success, we can disturb the balanced bridge circuit by inserting (e.g., a negative) voltage source in the lower part of the right leg.

• In the 2nd century BC lever implementation........................................
• In the 19th century, the output voltage VR4 of the right voltage divider will decrease (since VIN subtracts from VOA by means of the resistive summer R1-R2). We react to this intervention (again?) by increasing VOA. Our will is, as above, to increase only VR4... but unfortunately, VR2 also increases... luckily for us, more slowly. Thus we manage to balance the bridge circuit.
• In the 21st century, the op-amp "sees" that V- decreases, and reacts to this intervention by increasing(?) its output voltage VOA (with the purpose to increase V-). Unfortunately for it, V+ also changes... luckily for it, more slowly. Thus, due to the dominating negative feedback, the op-amp manages to balance the bridge circuit.