Filters in opamp feedback (configuration and calculating)

[mark2]

Referencing this GeneralGuitarGadgets article and specifically the following schematic...



It says
> R5 and C4 form another low pass filter. R4 and C3 form the high pass filter, and since R4 is a pot, the roll-off frequency will vary as you turn the pot

The electrosmash TS article calculates the filter similarly, so I'm assuming these weren't typos.

I would have thought R4 and C5 form a LPF, and C3 and R5 form a HPF since these configurations more closely match the textbook examples of filters. i.e. the signal into the opamp being taken from the junction of the resistor and cap.

Can anyone shed any light or give any pointers to help me understand these filters?

[antonis]

C4 & R5 form a HPF..!!
C3 & R4 form a LPF..!!

(nothing more, nothing less..)

[mark2]

I think my confusion is that I wanted to treat them as regular passive RC filters, but clearly that's not how it works with opamps.  And I lacked the vocabulary or background knowledge to get many helpful google results. But I've discovered some resourced on active filters which is proving much more helpful, so I think I may be on my way now.

Thanks.

[PRR]

Remember these are in a feedback loop and work differently than if they were in the direct path. {EDIT- you discovered...}

I do think they could have been worded better. R4 against R5 sets the midband gain. C4 against R5 sets a bass rolloff. C3 R4 sets a treble rolloff.

C6 seems to do nothing but destabilize the opamp (since the opamp can, or will try to, drive C6 full level).

[mark2]

C6 seems to do nothing but destabilize the opamp (since the opamp can, or will try to, drive C6 full level).

They note R6+C6 being used for tone... should this have a buffer in front of it to avoid loading the op amp?

[antonis]

They note R6+C6 being used for tone... should this have a buffer in front of it to avoid loading the op amp?

Nothing so fancy..
(just a 1k or so "stopper" series resistor..)

P.S.1
An op-amp buffer should also be loaded by C6 when R6 almost zero..

P.S.2
IMHO, both HP & LP filter corner frequencies are by far high..
[C4/R5 at 3.2kHz and C3/R4 at 32kHz (min)]

[cctsim]

The op-amp (only) transfer function and cut-offs are as below:




And the graph of the transfer function:




And the Matlab/Octave code if you want to experiment:

clear, clc, close all

R5=1e3;
R4=100e3;
C4=50e-9;
C3=47e-12;

f=20:40e3;
jwC3=2*pi*f*1j*C3;
jwC4=2*pi*f*1j*C4;

Hjw=1+(R4*jwC4)./((1+R5*jwC4).*(1+R4*jwC3));

semilogx(f,20*log10(abs(Hjw)),'r')
grid on
xlabel('f (Hz)')
ylabel('|H(f)| in dB')