Bypassing HPF

[warpie]

The following circuit is a simple active 2nd order hpf that I found online.

If I want to bypass the whole circuit, can I simply use a SPDT switch as shown? Using a DPDT for switching both input and output of the filter is not an option.

FWIW, I breadboarded it and I didn't notice any changes in noise when I bypass the hpf (and its input is floating) but maybe I'm missing something?

Any help is much appreciated as always! 

 

[gyraf]

nope, you'll need to use the switch to select to the output of your circuit, not what goes to the input. In your schematic, you're inputting into the output of the opamp in bypass, iznogood

In stead, use the switch to select either the signal from the input of the filter - or the output of the filter, and let this selection be your "output"

/Jakob E.

 

[warpie]

nope, you'll need to use the switch to select to the output of your circuit, not what goes to the input. In your schematic, you're inputting into the output of the opamp in bypass, iznogood

In stead, use the switch to select either the signal from the input of the filter - or the output of the filter, and let this selection be your "output"

/Jakob E.

Thanks Jakob. Do you mean having the output of the filter floating instead of its input?

 

[L´Andratté]

bypass only the filter capacitors:
switch opamp pin 3 between the 2nd 150n cap or the circuit input (via a BIG cap and resistor to ground)

check out steve dove´s mic pre below...(it says 150uF, that´s a typo, 2M2 resistor to de-click the switch, ignore the push-pull transistors)

also the filter has gain to steepen q, bypasing it as a whole would change circuit gain!

 

[warpie]

Thank you, I will investigate further.

Can someone please tell me what's the benefit of switching both the input and the output rather than just the output?

I remember seeing some schematics (SSL maybe?) where they're using a DPDT instead of a SPDT.

 

[abbey road d enfer]

Can someone please tell me what's the benefit of switching both the input and the output rather than just the output?

It has the advantage of completely disconnecting the filter. In some circumstances, it is possible that the filter clips (notice it has about 3dB gain), which often results in producing spikes that get into the power rails and ground and pollute other circuts.

I remember seeing some schematics (SSL maybe?) where they're using a DPDT instead of a SPDT.

Actually, most of the switches in mixers are Dual-Pole (or more) so it doesn't impact cost.

 

[warpie]

It has the advantage of completely disconnecting the filter. In some circumstances, it is possible that the filter clips (notice it has about 3dB gain), which often results in producing spikes that get into the power rails and ground and pollute other circuts.

In case where the (2nd order) filter is just two buffers with caps in-between (i.e zero gain), is there any benefit of using a dual pole switch instead of a single pole at the output (or maybe on the input)?

The reason I'm asking is because now I've been looking into using buffers with switchable capacitors (in order to avoid the gain changes when bypassed). So I was thinking of using the first position of a rotary switch for bypassing the filter as in the attached image.

Obviously in this configuration I can't use the rotary switch for switching the output of the filter.

Actually, most of the switches in mixers are Dual-Pole (or more) so it doesn't impact cost.

Good point.

 

[abbey road d enfer]

The reason I'm asking is because now I've been looking into using buffers with switchable capacitors (in order to avoid the gain changes when bypassed).

Are you aware that the response of two cascaded 1st-order filters is not the same as a one-opamp 2nd-order filter?

 

[warpie]

Are you aware that the response of two cascaded 1st-order filters is not the same as a one-opamp 2nd-order filter?

This is the schematic in question.

 

[warpie]

And this is the response with C3=C5=150nF

 

[abbey road d enfer]

This is the schematic in question.

OK, the response is good in positions 2 and 3, but in position 1, the input signal is shorted by the second stage.

 

[warpie]

OK, the response is good in positions 2 and 3, but in position 1, the input signal is shorted by the second stage.

Yep. I was just wondering whether there's any way of using the first position of the switch for bypassing the filter without the need for additional switches.

 

[abbey road d enfer]

Yep. I was just wondering whether there's any way of using the first position of the switch for bypassing the filter without the need for additional switches.

Since you're not adverse to using more than one opamp, why don't you consider having two filters and switching the outputs.

 

[warpie]

Since you're not adverse to using more than one opamp, why don't you consider having two filters and switching the outputs.

Not sure I follow. Do you mean to "duplicate" the circuit I posted and switch the outputs? If so, it'll become way too complicated since I was planning to use a (dual-gang) rotary switch with 5-6 different Fc.

 

[abbey road d enfer]

Not sure I follow. Do you mean to "duplicate" the circuit I posted and switch the outputs?

Yes.

If so, it'll become way too complicated since I was planning to use a (dual-gang) rotary switch with 5-6 different Fc.

It's not complicated at all; it may just take more real estate than you thought.
Anyway you can't get round it, or you need a 3Pole switch.

 

[warpie]

I guess the simplest solution would be to leave the circuit always in and use a couple of lytics (maybe 22uF?) on the first position but I don't know how much this would affect the snr and thd.

 

[abbey road d enfer]

I guess the simplest solution would be to leave the circuit always in and use a couple of lytics (maybe 22uF?) on the first position but I don't know how much this would affect the snr and thd.

Don't you understand that a 22uF capacitor is like a short-circuit at audio frequencies?