Continue to Site

Welcome to EDAboard.com

Welcome to our site! EDAboard.com is an international Electronics Discussion Forum focused on EDA software, circuits, schematics, books, theory, papers, asic, pld, 8051, DSP, Network, RF, Analog Design, PCB, Service Manuals... and a whole lot more! To participate you need to register. Registration is free. Click here to register now.

Register Log in

can any one help me to find gain bandwidth product of the bode plot i attached

Status
Not open for further replies.
Y

yadhu

Junior Member level 1
Joined
Sep 2, 2012
Messages
19
Helped
3
Reputation
6
Reaction score
3
Trophy points
1,283
Location
Bangalore
Activity points
1,409
can any one help me to find gain bandwidth product of the bode plot i attached
 

Attachments

  • bodeplot.JPG
    bodeplot.JPG
    69.6 KB · Views: 111

yadhu said:
can any one help me to find gain bandwidth product of the bode plot i attached
Click to expand...

The gain-bandwidth product is given for an amplifier with a lowpass response only.
In contrast, your BODE plot has a bandpass shape.
However, one can assume that the plot represents an opamp circuit that has differentiating properties up to 10 kHz and that the falling part of the
curve shows the open-loop characteristics of the opamp used.
Therefore: The used opamp has a GBW of approx. 2 MHz (0 dB crossing frequency) .
Does this answer your question?
 
i rigged the circuit by using TL082 opamp. in its datasheet they given GBW as 3Mhz
 

It is probably a Multiple-Feedback-Bandpass-Filter circuit that has perfectly straight sides attenuating at 6dB per octave.
But notice that the right side begins bending down at about 2MHz, probably caused by the 3MHz gain-bandwidth product of the opamp.
 

Audioguru said:
It is probably a Multiple-Feedback-Bandpass-Filter circuit that has perfectly straight sides attenuating at 6dB per octave.
But notice that the right side begins bending down at about 2MHz, probably caused by the 3MHz gain-bandwidth product of the opamp.
Click to expand...

No audioguru, I don't think that the response belongs to a second-order bandpass (why do you think of a MFB filter ?). Again: I think it's a circuit that differentiate (very poor phase margin as a typical property).
However, I think it's funny that we are guessing here - and the questioner (yadhu) does not tell us what kind of circuit he is investigating.
 
LvW said:
No audioguru, I don't think that the response belongs to a second-order bandpass (why do you think of a MFB filter ?). Again: I think it's a circuit that differentiate (very poor phase margin as a typical property).
However, I think it's funny that we are guessing here - and the questioner (yadhu) does not tell us what kind of circuit he is investigating.
Click to expand...
Of course it is a Multiple Feedback Bandpass filter with a high Q giving the sharp narrow peak and gradual single-order lowpass and gradual single-order highpass slopes for most of its response.
 

Attachments

  • Multiple Feedback Bandpass Filter.PNG
    Multiple Feedback Bandpass Filter.PNG
    14.2 KB · Views: 124

Of course it is a Multiple Feedback Bandpass filter with a high Q
Click to expand...
You mean, the transfer function looks similar.

But what does your example tell about the bode plot in the original post and the circuit producing it?

As a trivial point, a TL082 isn't able to produce 80 dB gain at 15 kHz without a transformer or a LC resonator. We can further assume, that the term GBW has been used erroneously in the question.
 
Audioguru said:
Of course it is a Multiple Feedback Bandpass filter with a high Q giving the sharp narrow peak and gradual single-order lowpass and gradual single-order highpass slopes for most of its response.
Click to expand...

Audioguru, perhaps you have overlooked that the "midband" gain is 80 dB.
More than that, are you able to detect a certain type of filter topology based on a BODE diagram only?
 
The Multiple Feedback Bandpass Filter is the only filter I have seen with a sharp and narrow high-Q peak then gradual single-order slopes.
The high-Q of the circuit produces a peak that has an amplitude much higher than the opamp can do by itself.

EDIT: I am just guessing because the filter is so bad that I have never made one.
 

Audioguru said:
The Multiple Feedback Bandpass Filter is the only filter I have seen with a sharp and narrow high-Q peak then gradual single-order slopes.
Click to expand...

Each of the various filter topologies is able - in principle (!) - to produce high Q values.
For ideal conditions (amplifier, parts tolerances) ALL topologies exhibit the same properties. However, the behaviour of real circuits depends on the selected pole frequency and the type of opamp used (GBW).

Audioguru said:
The high-Q of the circuit produces a peak that has an amplitude much higher than the opamp can do by itself.
Click to expand...

Your opamps can produce an output voltage amplitude "much higher than the opamp can do by itself" ? That means: Higher than the supply voltage ?
 
The high-Q of the circuit produces a peak that has an amplitude much higher than the opamp can do by itself.
Click to expand...
I assume you are talking about filter gain. That's true e.g for an oscillator which achieves infinite gain. Similarly, the gain peaking of the present circuit can be explained by a near to zero phase margin.

My previous comment about the impossibility of a filter with 80 dB gain is only valid for an ideal filter. But a it's well possible utilizing the peaking of a real OP circuit.

In fact, the shown transfer characteristic can be achieved by simple inverting differentiator, as already assumed by LvW in post #2.
 
  • Like
Reactions: LvW

    LvW

    LvW

    Points: 2
    Helpful Answer Positive Rating
FvM said:
In fact, the shown transfer characteristic can be achieved by simple inverting differentiator, as already assumed by LvW in post #2.
Click to expand...
A simple single-order differentiator is not a active bandpass filter. If it is followed by a simple single-order (second-order?) integrator then the "peak" is not sharp and narrow.
 

Attachments

  • bandpass response.PNG
    bandpass response.PNG
    118.3 KB · Views: 104

Please take a look at this simulation result. It's using an "ideal" OP with first order transfer function, open loop gain of 1e5 and 2 MHz GBW.



The peaking is created by nearly zero phase margin of the differentiator circuit.
 

Attachments

  • diff.zip
    1 KB · Views: 55
Last edited:
  • Like
Reactions: LvW

    LvW

    LvW

    Points: 2
    Helpful Answer Positive Rating
Audioguru said:
A simple single-order differentiator is not a active bandpass filter. If it is followed by a simple single-order (second-order?) integrator then the "peak" is not sharp and narrow.
Click to expand...

Who has claimed that? Who spoke about an integrator device?
Don't you remember that the problem of realizing an opamp based (inverting) differentiator always consists of a phase margin approaching zero - causing a peak as shown at the beginning of the thread (and repeated by FvM, showing the corresponding circuit) ?
 
Last edited:
  • Like
Reactions: FvM

    FvM

    FvM

    Points: 2
    Helpful Answer Positive Rating
FvM said:
The peaking is created by nearly zero phase margin of the differentiator circuit.
Click to expand...
Thank you.
I have never made a differentiator like that. I see that as the output phase of the opamp increases at a high frequency then its output peaks.
so it is a Multiple Feedback Bandpass filter circuit using only one capacitor and one resistor.
 

I have never made a differentiator like that.
Click to expand...
Me neither, you see why. Although the circuit is apparently acting as a high Q second order filter, I won't rely on the exact Q value, or at least circuit stability. You only need to add some capacitive load and make the circuit oscillating.

I think, the original post gives a nice interview question.
 

Audioguru said:
so it is a Multiple Feedback Bandpass filter circuit using only one capacitor and one resistor.
Click to expand...

A "multiple feedback bandpass filter" with just one feedback path?
Audioguru, I suppose you know that there are some other bandpass topologies available than MFB?

- - - Updated - - -

FvM said:
Me neither, you see why. Although the circuit is apparently acting as a high Q second order filter, I won't rely on the exact Q value, or at least circuit stability. You only need to add some capacitive load and make the circuit oscillating.
Click to expand...

Yes, a filter circuit relying on the open-loop opamp response is certainly not an option.
In this context, I remember the so called "R-filter" exploiting the frequency characteristics of (at least) two opamps only (without any external capacitors).
There are lot of papers on this subject - however, I never have heard about practical use of this concept. It is more or less an academic exercise.
 
  • Like
Reactions: FvM

    FvM

    FvM

    Points: 2
    Helpful Answer Positive Rating
There is an almost useless 3-band audio spectrum analyser project that used hopeless MFB filters.
I would never make something as bad as that.

The bass filter peaks at about 170Hz and cuts off any real bass, the high frequency filter peaks at about 3.3kHz and cuts off the top two highest octaves of sound and the mid filter peaks at 700Hz and does not show much else.
 

Attachments

  • 3 channel spectrum analyzer.png
    3 channel spectrum analyzer.png
    122.1 KB · Views: 107

The Bode plot shows clearly a transfer function with a zero at s=0 and a high-Q pair of poles. An additional pole at high frequency produces an increased slope at the right of the plot.
As far as I can see, the exact topology of the circuit can not be completely inferred.

FvM's circuit in post #13 produces practically the same response.
The "open loop" configuration of this circuit has two poles: one of them is the real dominant compensation pole of the OA at 20 Hz, and the other is at -1/(R1*C2). Closing the loop with R1 (think in the root locus) they move forming a high-Q pair. The high frequency pole of the OA must be responsible of the higher frequency pole .
Assuming that the dominant pole in the OA is obtained with a feedback capacitor, then this is a sort of multiple feedback circuit (there are two nested loops), but not like the ones swown in posts #6 and #18.
Regards

Z
 

zorro said:
Assuming that the dominant pole in the OA is obtained with a feedback capacitor, then this is a sort of multiple feedback circuit (there are two nested loops), but not like the ones swown in posts #6 and #18.
Regards
Z
Click to expand...

I think, following this philosophy, even the simplest opamp amplifier (inverting, non-inverting) would be classified as multiple feedback circuits.
I doubt, if this is a definition that makes sense. What about feedback caused by the collector-base capacitance of a BJT?
 

Status
Not open for further replies.

Similar threads

Part and Inventory Search

Welcome to EDABoard.com

Sponsor

Back
Top