I don`t think that this effect can be describe as an "impulse" - In system theory an "impulse" has another definition.

I can assure you it is an impulse because an analysis shows this to be true. Do an analysis of the non inverting circuit as that is the simplest. What you end up with is one or two impulses and a 'regular' term that indicates an integration.

No - I don`t think that this effect will have any influence on stability. It is something like "overshoot".

I think I have to agree with that, given that this circuit is not new by a long shot.

But my main concern is your claim (without evidence) that "the impulse polarities are different for the two configurations".
Why do you think that there is a difference in polarity?
For the non-inverting configuration the amplifiers response will be positive ("impulse" and "normal" signal) - and for the inverting configuration both "parts of the response" will go in the other direction.
Why do you expect a "subtractive" effect?

I am glad you brought this up. I may have described that incorrectly I'll have to find a better way to describe the different effects between non-inverting and inverting versions. Right now I think it is just a matter of one having less of an impulse than the other. For example, one having 5cd and the other having 10cd, where I use "cd" to mean the "curly delta" symbol often used for writing out impulses. I will come up with a more quantitative analysis so we can see the difference. However, since we do not expect this to be too much of a problem, we might not worry about it too much except for academic reasons.

 

Hello again,

I should mention that after all the analysis and talking about the impulses and the like I had forgotten to mention a more important fact regarding this circuit.
That fact is that this kind of circuit has been used for years and years in conjunction with stepper motor drivers.
As we all know, when we want a stepper motor to jump quickly from one step to another, we bang it with as high a voltage as it will take or what we have available as long as it can take that voltage level. This causes the fastest jump from one angle to the next. To do this, a circuit just like this is used where we use a sense resistor to sense the current and apply the drive voltage. As the current rises, the voltage is reduced.
This can be found on the data sheet for a stepper motor driver chip. I'd have to look up the number of the part I am more familiar with though. For example, we might use a 2 Ohm sense resistor to sense the current and the chip does the rest.

This also ties in with a method to control a buck regulator. By sensing the current in the inductor and using state variable feedback, we can provide a fast response as long as we have enough input voltage. Unfortunately this is not always the case, but it's still interesting from a theoretical standpoint. As someone said, if we had an infinite voltage (impulse) we could change the current in the inductor instantaneously and that would provide us with the fastest response time.
For the circuit we are talking about in this thread, the impulse will not be as effective as that because the power supply voltage is very limited just like with the buck.

I am posting the theoretical output of this circuit just to show what the impulse(s) do. Keep in mind this extreme example will never happen as bad as this because the power supply voltages are always limited, very limited.

In this illustration, the output with the non-inverting version is twice as high as with the inverting version. Because this illustration uses op amps that are ideal with a gain of 100000, the output voltage can reach very high values like 50000 volts and 100000 volts. These are both with the same voltage input (except for polarity).
The two resistors used for the inverting version are equal in value, so that could be why the inverting version puts out only 1/2 of what the non-inverting version puts out at t=0.
They both settle to the correct output voltage which is not visible here because of the vertical scale.
I guess it is also a little interesting that the non-inverting version actually settles faster than the inverting version.

 

Another example can be found in late-model Cars that have "Direct-Injection" into the Combustion-Chambers.
Direct-Injection must be timed with great precision, and, within a very small time-window,
so the manufacturers have installed a Boost-Converter inside the Fuel-Injection-Computer
that bumps-up the normal ~12-Volts, to over ~65-Volts,
this is to overcome the inevitable Induction caused by the Solenoid-Coil inside of the Injectors,
then the Current though each Injector is regulated-down to about ~2-Amps with
a PWM-Current-Regulator-Circuit, to "hold them open" for as long as required.

This makes the Injectors open extremely fast, about ~5-times faster than a normal Fuel-Injector.
.
.
.

 

Another example can be found in late-model Cars that have "Direct-Injection" into the Combustion-Chambers.
Direct-Injection must be timed with great precision, and, within a very small time-window,
so the manufacturers have installed a Boost-Converter inside the Fuel-Injection-Computer
that bumps-up the normal ~12-Volts, to over ~65-Volts,
this is to overcome the inevitable Induction caused by the Solenoid-Coil inside of the Injectors,
then the Current though each Injector is regulated-down to about ~2-Amps with
a PWM-Current-Regulator-Circuit, to "hold them open" for as long as required.

This makes the Injectors open extremely fast, about ~5-times faster than a normal Fuel-Injector.
.
.
.

Hi,

That's very interesting, thanks for sharing that.

 

Hello again,

I should mention that after all the analysis and talking about the impulses and the like I had forgotten to mention a more important fact regarding this circuit.
That fact is that this kind of circuit has been used for years and years in conjunction with stepper motor drivers.
As we all know, when we want a stepper motor to jump quickly from one step to another, we bang it with as high a voltage as it will take or what we have available as long as it can take that voltage level. This causes the fastest jump from one angle to the next. To do this, a circuit just like this is used where we use a sense resistor to sense the current and apply the drive voltage. As the current rises, the voltage is reduced.
This can be found on the data sheet for a stepper motor driver chip. I'd have to look up the number of the part I am more familiar with though. For example, we might use a 2 Ohm sense resistor to sense the current and the chip does the rest.

This also ties in with a method to control a buck regulator. By sensing the current in the inductor and using state variable feedback, we can provide a fast response as long as we have enough input voltage. Unfortunately this is not always the case, but it's still interesting from a theoretical standpoint. As someone said, if we had an infinite voltage (impulse) we could change the current in the inductor instantaneously and that would provide us with the fastest response time.
For the circuit we are talking about in this thread, the impulse will not be as effective as that because the power supply voltage is very limited just like with the buck.

I am posting the theoretical output of this circuit just to show what the impulse(s) do. Keep in mind this extreme example will never happen as bad as this because the power supply voltages are always limited, very limited.

In this illustration, the output with the non-inverting version is twice as high as with the inverting version. Because this illustration uses op amps that are ideal with a gain of 100000, the output voltage can reach very high values like 50000 volts and 100000 volts. These are both with the same voltage input (except for polarity).
The two resistors used for the inverting version are equal in value, so that could be why the inverting version puts out only 1/2 of what the non-inverting version puts out at t=0.
They both settle to the correct output voltage which is not visible here because of the vertical scale.
I guess it is also a little interesting that the non-inverting version actually settles faster than the inverting version.

Hello again,

I went over the analysis a little more carefully and looking at the responses for both non-inverting and inverting, I can explain the reason for the theoretical impulse in a much simpler way.

For the non-inverting version, when we apply a test voltage of 1 volt, the differential input to the op amp is that full voltage of 1 volt. That's because we assume we start with little voltage across the sense resistor, and if the sense resistor is small then the voltage there will always be low. Thus, we get an impulse due to that 1 volt differential input.

Looking at the inverting version, when we apply a test voltage of -1 volt, with that sense resistor voltage being zero in this case also, we do not get a differential voltage at the input of the op amp of 1 volt, we get 1/2 volt. That is because for an inverting op amp amplifier if we make the input resistor equal to the feedback resistor, the differential voltage at the input to the op amp is one-half of the total input voltage at t=0. Thus, we see nearly one-half of the impulse at the output.
We could actually lower it if we make Rin different than Rfb, but then we change the gain also. That could be useful or not useful.

That's about the size of it, except we have to look at the practical side of this. If the speed of the op amp is limited in such a way that it acts like a low pass filter to the differential input, then we could see much less of an output impulse, and it could be very small. Thus, there is probably no reason to worry about this unless we were using a larger coil and faster op amp, and possibly we have to keep EMI down low.

The other remark I made about the subtraction of part of the impulse was not exactly correct. It's really that one part is lower than the other and that other dominates the response. Much thanks to member @LvW for catching this misrepresentation.

 

Hello again,
I went over the analysis a little more carefully and looking at the responses for both non-inverting and inverting, I can explain the reason for the theoretical impulse in a much simpler way.

In your considerations (generation of an impulse) you have probably assumed an ideal amplifier.
However, it turns out that in a real OPV the limited "slew rate" covers such a possible impulse (delay of the negative feedback effect due to the inductance).
This results in a rising edge with a clearly recognizable finite slope at the output with an ideal input step.

 

In your considerations (generation of an impulse) you have probably assumed an ideal amplifier.
However, it turns out that in a real OPV the limited "slew rate" covers such a possible impulse (delay of the negative feedback effect due to the inductance).
This results in a rising edge with a clearly recognizable finite slope at the output with an ideal input step.

Hi,

Yes, I had mentioned that in one of my posts, so this is more or less an academic issue.
This is where theory shines. Once you have the ultimate theory, you can always regress.