High & Low Pass Filters - help a newb....

[Greg_G]

Hi guys
Possibly a dumb question, but I don't know enough to know better... OK....  so be nice...

A simple RC filter is either Low or High pass depending on the order of components right ?
Just looking at the schematics of a Tubescreamer, and Jack's Son of Screamer - The filter which comes off the feedback loop is a high pass on a TS, and a low pass on the SOS from what I can see. Then there's a low pass on both circuits after the first stage.

I'm just wondering why the difference, and what the effect of switching the parts off the loop in the SOS would be ?

[zerohero]

the way i think of it is in what the capacitor is connected to,

a capacitor varies in impedance depending on frequency, the lower the frequency the HIGHER the impedance, therefore DC is blocked.
so if a capacitor is in the signal path, it is a high-pass filter because as the frequency of the signal become higher, the capacitor impedes it less.

if the capacitor is connected to ground, it is a lowpass filter, because as the frequencies get higher the capacitor channels it to ground, and the lower frequencies are not routed to ground.

if it doesnt make sense im sorry its late for me.

hope i helped a little

[Greg_G]

That helps a lot.
So the filter off the feedback loop is a High pass on both circuits, even though the components are reversed.
Thank You

[Mark Hammer]

A volume pot or "voltage divider" consists of two resistances, end-to-end.  When you rotate the wiper of the pot, you change the resistance of one "leg", relative to the other.  When the resistance between the "input" and the wiper is big, relative to the other side of the wiper, the amplitude of the signal available at the wiper is drastically reduced or "divided down" (in terms of voltage, not frequency).

Okay, so the basis principle is that what you get out depends on the relationship between the two sides of the wiper.  If it's a whole lot easier to leak to grund than it is to get from the input to the wiper, then you end up losing a lot of signal.

So far so good.

A capacitor is many things, but one of the things it is might be described as a frequency-dependent resistor.  That is, it is a resistor whose resistance depends on the frequency of the signal passing through it.  Regardless of the absolute capacitance value (10uf, .1uf, 10pf, etc.), the higher the frequency, the lower the resistance that capacitor poses.  Of course, if the cap is a larger value, it poses an even lower resistance, such that a 100uf cap obstructs a 1khz signal much less than a 100pf cap would.

Now, lets look at a capacitor/resistor pair as if it were a pot.  That, we are interested in what sort of signal we would see at the wiper or junction of these two components.  Keep in mind that with a "real" pot, the resistance on each side of the wiper will be identical for all frequencies.  In this, case, however, the resistance will depend on what frequency we're looking at.

So, let's consider the case of a lowpass filter section first.  Here the "input side" of our pot consists of a fixed resistor.  On the other side of the wiper is a cap.  For low frequencies, that cap presents a very high impedance.  So it is as if we have a small resistance on one side of the wiper, and a large resistance on the other.  Result = very little signal loss.  For higher frequencies, though, that effective resistance is smaller/lower.  So, it is as if we have a medium-high resistance on one side of the wiper, and a much lower resistance on the other.  Result = much greater signal loss.

IN other words, it is as if we have a different volume pot setting for every single frequency, and the volume pot is turned down more the higher up you go.  Voila!  Lowpass filtering and treble cut.

Okay, flip it around.  Now we have the input leg of our imaginary pot being a capacitor, and the ground leg being a fixed resistor.  For high frequencies, that cap is functionally equivalent to a very small resistance.  That gives us a small resistance on the input leg, and a higher resistance on the ground leg, similar to a volume pot turned up.  Result = very little signal loss as far as that high frequency is concerned.  For lower frequencies, that cap starts to act like a larger resistance, such that the lower you go in frequency the more that input leg starts to behave like a larger resistance than what is on the other side of the wiper.  Result = lots of signal loss for lower frequencies.  Voila!  Highpass filtering and bass cut.

And that's why the order matters even though the two components might be exactly the same values (e.g., .01uf, 10k).  It's a question of which side of the imaginary pot the specific component is standing in for.  That's also why you use the exact same formula to calculate the rolloff/corner frequency.  Our example of .01uf/10k results in a corner frequency of 1/[2*pi*.01uf*.01M] = 1591hz.  If it's C->R then the rolloff starts occurring below that requency.  If it's R->C, the rolloff starts occurring above that frequency.  Either way, what's driving it is the relative resistance for any given frequency on one side of the imaginary pot vs the other.

I hope that clears it up nicely.

[Auke Haarsma]

very very good explanation!

[Ben N]

BUT, does the order matter within each? In other words, in the case of the simple 1-pole LP filter, does it matter if the series resistor is before the cap to ground (as we are accustomed to see it) or after, and likewise in the case of the HP filter, does the R to ground before or after the seres cap matter? And while we are at it, how about for the voltage divider/pot case--as between tabs 1 & 2, does it matter which is connected to signal source and which to output, so long as tab 3 is connected to ground?

Ben

[gaussmarkov]

BUT, does the order matter within each? In other words, in the case of the simple 1-pole LP filter, does it matter if the series resistor is before the cap to ground (as we are accustomed to see it) or after, and likewise in the case of the HP filter, does the R to ground before or after the seres cap matter? And while we are at it, how about for the voltage divider/pot case--as between tabs 1 & 2, does it matter which is connected to signal source and which to output, so long as tab 3 is connected to ground?

Ben

yes, order matters.  both filters should be viewed as a pair of components in series to ground.  we are just tapping into to the junction between them.  some people like to point out that these RC filters are like voltage dividers made from a pair of resistors.  in that case, you also think about the pair of resistors as in series between the source and ground.