High Pass Active Filter Distortion

[TXRX]

Hi,

I built the attached "HPF Active Filter Design" File.

You can see at the 3 Scope pictures Attached - the distortion that i receive at the output in high level of input signal.

I built in the past active filters and i did not see any distortion at high level input signal.

Can anyone explain why there is a distortion at the output, you can see that the duty cycle change dramatically ?

Thanks,

 

[FvM]

I don't think that you should expect meaningful output signals for a totally overloaded filter circuit.

 

[crutschow]

What is the in-band gain of the filter?

At what input level does it start to distort?

 

[Audioguru]

A highpass filter passes harmonics (distortion).

 

[goldsmith]

A highpass filter passes harmonics (distortion).

Hi Audioguru
Of course yes but it should not add anything to the signal ! it should cover the input signal with adding some restrictions not adding something into that .

Best Regards
Goldsmith

 

[godfreyl]

What is the in-band gain of the filter?

22dB, in theory. With the given opamp and supply voltage, 100mV rms input will cause clipping.

 

[LvW]

TXRX,

did you consider during the design process also the alternative with split (double) supply?
Do you know about the restrictions connected with single supply?

Later: A second look on your circuit revealed a positive bias of app. 2 volts. Initially, I have read 216mV Thus, my comment above is not valid anymore.

 

[Audioguru]

Hi Audioguru
Of course yes but it should not add anything to the signal ! it should cover the input signal with adding some restrictions not adding something into that .

Best Regards
Goldsmith

If the input signal has distortion then the highpass filter makes the distortion appear much worse if the fundamental frequency is attenuated by the filter.

 

[goldsmith]

If the input signal has distortion then the highpass filter makes the distortion appear much worse if the fundamental frequency is attenuated by the filter.

Hi Audioguru
If the input signal has distortion then distortion should be appeared in out put , yes that is absolutely right . but as we can see the the input is not a distorted signal , right ? so it is NOT a usual behavior . ok ? Because the THD created by the circuit !
Respectfully
Goldsmith

 

[FvM]

Because the THD created by the circuit

Yes, quite obviously. The circuit is simply overloaded.

 

[Audioguru]

The "sinewave" at the input shows plenty of distortion. The highpass filter cuts the fundamental level then the percentage of harmonics (distortion) appears to be increased.

The overload (because the input signal level is too high for the voltage gain of the circuit and the power supply voltage) causes the opamp to severely clip the signal at its output. A clipped signal has plenty of distortion.

 

[TXRX]

I don't think that you should expect meaningful output signals for a totally overloaded filter circuit.

Hi FvM,

I excpect to see a Square Wave at the output or a HPF behaviour with duty cycle the same as the Input - But not changing of duty cycle.

- - - Updated - - -

What is the in-band gain of the filter?

At what input level does it start to distort?

Hi,

The in-band Gain is 22dB.

- - - Updated - - -

Yes, quite obviously. The circuit is simply overloaded.

Hi FvM,

How the circuit overloaded if the Load impedance is the Scope with 1Mohm ?

- - - Updated - - -

The "sinewave" at the input shows plenty of distortion. The highpass filter cuts the fundamental level then the percentage of harmonics (distortion) appears to be increased.

The overload (because the input signal level is too high for the voltage gain of the circuit and the power supply voltage) causes the opamp to severely clip the signal at its output. A clipped signal has plenty of distortion.

Hi Audioguru,

How can you explain that i do not see this distortion in Active Filters in the past with high input voltage? Why i do not see any problem in the simulation ?

 

[LvW]

TXRX, just one comment regarding the duty cycle:

A high pass is a "bad" (lossy) differentiating circuit.
Now - imagine what happens if a squarewave with 50% duty cycle is differentiated.
I think, you know the answer.

 

[FvM]

How the circuit overloaded if the Load impedance is the Scope with 1Mohm ?

The term overload is also commonly used for exceeding the input signal range suitable for linear operation. This is surely the case in your measurement.

I excpect to see a Square Wave at the output or a HPF behaviour with duty cycle the same as the Input - But not changing of duty cycle.

I believe that you don't expect it at first sight, but it can easily happen. Exceeding the linear range can also cause a shift of operation point. If you probe the OP input voltages as well and review the datasheet, you'll most likely understand what's going on.

 

[TXRX]

Hi,

Is there any idea how to improve the circuit in order to keep the Duty Cycle of 50 % ?

 

[FvM]

Is there any idea how to improve the circuit in order to keep the Duty Cycle of 50 % ?

Why do you want this?

It's surely possible somehow, it might be necessary to select a different OP (e.g. rail-to-rail type) with more symmetrical voltage limits. In addition possibly signal limiting at the input side. It's generally difficult if you can't specify a reasonable signal range.

 

[TXRX]

Hi,

I put 2 resistors R97 R98 of 100 ohm and there is improvement in the distortion but you can see that at high signal voltage the duty cycle increase from 52% to 57%, you can see also that in high input voltage the output to input phase moved by 90 degree - it is look like there is phase reversal.

 

[Audioguru]

The first and third inputs are severely distorted.
The input levels are much too high causing the outputs to be clipping like mad. Why??

It is normal for a filter to cause phase shift.