Audio High Pass Filter (4th order)

[Jester]

I'm working on an audio circuit that will run from a 9V battery and requires a high pass filter, to reduce 50-60Hz noise/hum. The pass band is specified as 180Hz to 2400Hz. The present design has a 2nd order filter and the hum is unacceptably high.

I hope to use one capacitor value in this filter to reduce cost, does this look like a reasonable capacitor for this application? https://www.digikey.ca/product-detail/en/GRM3195C1H273JA01D/490-1762-1-ND/587544

Below is the proposed filter circuit, does this look reasonable? or are there obvious improvements that can be made?

Present design uses a LM833. Is this a good choice and if not suggestions for a more suitable op amp?

I will use a switched cap low pass filter following this stage to reduce noise beyond the 2400Hz cut-off.

 

[FvM]

No idea which filter type you wanted to implement and how you calculated it but it's far away from a reasonable 180 Hz high pass implenetation. The circuit is 5th order rather than 4th order, by the way.

Equal cap design shouldn't be a problem.

 

[Jester]

No idea which filter type you wanted to implement and how you calculated it but it's far away from a reasonable 180 Hz high pass implementation. The circuit is 5th order rather than 4th order, by the way.

Equal cap design shouldn't be a problem.

I used this calculator: https://www.daycounter.com/Filters/Sallen-Key-HP-Calculator.phtml

The box around the active portion is the 4th order.

3dB @ 150Hz, then chose closest 1% resistor values

I think I just clued in, R8 should be 102k, not 10k2

 

[FvM]

I think I just clued in, R8 should be 102k, not 10k2

That looks much better.

A 5th order Butterworth implementation gives slightly different values:

R25 31k6
R23 48k7
R9 12k1
R8 127k
R22 39k2

 

[Audioguru]

The extremely small capacitor is ceramic. A ceramic capacitor should not be used in audio coupling circuits because it is microphonic and its capacitance changes with voltage so it causes low frequency distortion. Use a film capacitor instead.
Audio? Most audio frequencies will not be passed. Audio goes from 20Hz to 20kHz. Your audio will sound much worse than a telephone, more like a throat microphone.

 

[FvM]

The capacitor is low permitivity NP0 type, it has no significant voltage dependency and no or very little piezolelectric effect. I think it's o.k. for general purpose audio.

 

[D.A.(Tony)Stewart]

THe 4th filter you describe is suitable for Audio. If it to be digitized, there are better Nyquist filters for telephony SLIC chips that are 7th order to maximize bandwidth with maximum 2f harmonic bandstop to block alias distortion.

But your initial problem is hum is unacceptably high

Thus you may need a twin T notch filter or more likely a balun for the audio source to prevent the common mode hum from becoming differential.

More info on your entire circuit, layout and connections for power source , type etc need to be defined to isolate the best solution.

 

[FvM]

But your initial problem is hum is unacceptably high

That's true of course. And 50 Hz hum is often accompanied by harmonics and other interferences. So the low-pass might lack the intended effect.

I would check with an audio processing software if the filter achieves what you expect before making actual hardware.

 

[Jester]

That's true of course. And 50 Hz hum is often accompanied by harmonics and other interferences. So the low-pass might lack the intended effect.

I would check with an audio processing software if the filter achieves what you expect before making actual hardware.

Audio processing software, can you elaborate?

I'm working under a NDA for this project, so I can't disclose as much as I would like.

All I can say is:
- that the audio source is non-conventional, it's not a microphone or guitar pickup or anything similar.
- the source is ac coupled
- the post filter signal will not be digitized, it will be amplified

I think I will have to build it and then determine the source of the hum, hopefully the 5th order filter (initial prototype that I don't have has a 2nd order HPF) will improve the hum situation.

Comment regarding the choice of capacitor, in the past I experienced some manufacturing production problems when using small film capacitors. The capacitors were being damaged during the soldering process, it seemed the heat profile used for the solder process was not compatible with these small film capacitors. In the end we replaced with a NPO type and this resolved the issue.

Any suggestions for a suitable op amp (9V battery source)

Thanks to everyone for the advice so far.

 

[Audioguru]

Any suggestions for a suitable op amp (9V battery source)?

You selected a very tiny surface mount capacitor so I guess the opamps also must be tiny surface mount?
Since the bandwidth is so narrow then It will not produce hiss so any low voltage opamp will work.

 

[D.A.(Tony)Stewart]

Floating battery powered audio amps suffer from imbalance in high CM E fields from imbalanced mics.

The noise must be filtered from the source by balancing the audio input channels with large ferrite baluns.

 

[chemelec]

Why not make a NOTCH FILTER for either 50 or 60 Hz.
Probably Better than a High Pass filter.

 

[Jester]

Why not make a NOTCH FILTER for either 50 or 60 Hz.
Probably Better than a High Pass filter.

Perhaps:
- it would need to be configurable 50Hz or 60Hz
- it's possible there could be harmonics of 60Hz, so the notch would let these pass through

Until I actually build one of these, I will not have a sense for the hum; is it just 60Hz or also harmonics of 60Hz?

- - - Updated - - -

Floating battery powered audio amps suffer from imbalance in high CM E fields from imbalanced mics.

The noise must be filtered from the source by balancing the audio input channels with large ferrite baluns.

The input source is not a microphone, its a photo input.

- - - Updated - - -

You selected a very tiny surface mount capacitor so I guess the opamps also must be tiny surface mount?
Since the bandwidth is so narrow then It will not produce hiss so any low voltage opamp will work.

OPA2170?

http://www.digikey.ca/product-detail/en/OPA2170AIDGKR/296-36694-1-ND/4341325

 

[chemelec]

I'm not even sure Why your worried about 50 and 60Hz Hum.
What is your Line Power Frequency?
At Most, you would only need a Filter for that Frequency.

But If you have Proper Shielded Wire going to the Cartridge or whatever your using on the phono input, than Hum shouldn't be a Problem.

I Still have and use my Record Player.
No Filters and Hum is not an issue.

 

[D.A.(Tony)Stewart]

The input source is not a microphone, its a phono input.

Same thing. Use a balanced shielded cable.
Minimum get a balun or Shielded twisted pair.

Otherwise if only fundamental, notch filter if band-aid fix. If a buzz, no filter will work.

I cannot make you understand.

 

[chemelec]

If you have HUM, It could be GROUND LOOPS.

This just requires good wiring techniques to avoid them.

 

[vGoodtimes]

It depends on the source of hum. Industrial 60 Hz can have 30%+ THD with a lot of high harmonics (big triac loads for example).

Some photoelectronics are significantly nonlinear, this will add harmonics, probably more even-ordered ones.

The only other idea would be if you could get a noise reference for cancelation.

Or find a way to use DSP to solve the problem.

 

[Audioguru]

The OPA2170 dual opamp has a slew rate that is too low for high fidelity audio but will be fine for your very narrow bandwidth.

 

[Jester]

I have not built or tested this device yet, however someone that has, stated that external light striking the photo detector was introducing significant 60Hz hum. If the photo detector was covered up the hum all but disappeared. This was with a simple 2nd order HPF, hopefully the 5th order filter will help.

 

[chemelec]

What Photo Detector?