Level loss on Baxandall circuit

[dogears]

I made a Baxandall circuit. In the original document he mentions using a switch to alter the behavior of the high shelf by including or omitting the center tap of the high shelf potentiometer (section just before figure 8).
https://learnabout-electronics.org/Downloads/NegativeFeedbackTone.pdf

I have this behavior in the attached circuit, and it works as advertised. This circuit is DC coupled to the feedback point, so the 'center tap' of the switch is AC coupled to 0V using a capacitor. All good.

However, when the center tap is grounded, and the attached inductor low pass filter is engaged, I lose 12 dB of level. 

If I lift the center tap, the level comes back.

If I switch off from the center position with the HF switch, everything behaves as normal. Back to the center, lose 12 dB. I haven't been able to figure out why.

Any help?

 

[Bo Deadly]

So it's a 12dB drop across all frequencies and only when the LPF is engaged? That is a clue.

Your LPF has very little series resistance (just the inductor DCR) which means the Q could be pretty high. Meaning it will have a resonance. Not sure if that's what you want.

Also, it's not terribly common to have filters on an output because it will screw with the output impedance. Generally you want an output that will have a consistent output impedance. You might be able to move those filters to before the bax. If there are bad interactions, you might need to just follow them with another buffer to drive the bax.

 

[dogears]

So it's a 12dB drop across all frequencies and only when the LPF is engaged? That is a clue.

Your LPF has very little series resistance (just the inductor DCR) which means the Q could be pretty high. Meaning it will have a resonance. Not sure if that's what you want.

Also, it's not terribly common to have filters on an output because it will screw with the output impedance. Generally you want an output that will have a consistent output impedance. You might be able to move those filters to before the bax. If there are bad interactions, you might need to just follow them with another buffer to drive the bax.

Yes, just straight level loss. Move the high shelf off of the center / flat position, and the level comes back.

The filters are not at the final output “to the world”, there’s another block that comes after this one. 

The LPF works fine when the center tap cap is not in place. The .01uF cap after the inductor flattens the resonance. The LPF engaged is removing the short around the inductor - all of the resistors that form the load are always in place.

Put another way - neither the LPF or the center tap cause the level loss independently. It has to be both.

I am really scratching my head here.

 

[Bo Deadly]

I think you must have a wiring error. Is this on a breadboard? Maybe you're missing a ground somewhere.

The difference in the circuit resulting from switching the treble control one step from the center is only the 330 ohm resistor. That should not result in a straight across level loss.

Use a DMM to check ground continuity at various points. Look for shifts in DC level when switching things.

 

[dogears]

And yet it does. No wiring error.

But I think it’s fixed. Putting literally any series resistance with the cap corrects the behavior. 1 ohm, or even the wire long enough to get to the resistor box.

 

[mjrippe]

Try a different cap.

 

[dogears]

Try a different cap.

Yes, I tried several. It seems to need some series resistance. A smaller electrolytic / higher ESR works properly. Film cap or large / low ESR electrolytic doesn’t.

 

[JohnRoberts]

Is it possible instability? Oscillation due to load  C to ground?

JR

 

[dogears]

Is it possible instability? Oscillation due to load  C to ground?

JR

Could be, but it only happens with the inductor also. It seems like the inductor would stabilize rather than exacerbate, no?