frequency response curve?

Mr. G. · Post by **Mr. G.** » 14 Oct 2008, 15:50

Is there any software out there, similar to the Duncan TSC, that will let you put in the tone section of a circuit (or the whole circuit for that matter), to see what the frequency response will be?

I'm trying to learn about tone stacks, and have learned a lot from the TSC, but it's kinda limited when working with tone stacks for pedals.

Got any suggestions?

madbean · Post by **madbean** » 14 Oct 2008, 17:37

You can use a circuit simulator like SPICE that kind of stuff, but really some simple arithmetic is all that's needed for all passive and most active tone circuits.

Your corner freq for a passive RC filter is determined by the formula:

f=1/(2*pi*R*C)=1/(6.28*R*C)

where f is in Hertz, R is in k-ohms, and C is in microfarads. The behavior (low pass versus high pass) only depends on the orientation of the RC filter (which component goes to ground).

So, for example, in the TS tone stack you have a passive low-pass filter (attenuates the highs) just before the second stage of the opamp where there is a series 1k resistor followed by a 220nF to ground. The formula then follows:

f=1/(6.28*.001*.22)=1/.0013816 which is approximately 724Hz.

You can use this formula to analyze almost any passive tone control. You simply need to account for all the RC filters to get the most accurate result.

For active tone circuits, the same formula applies, but you also have to deal with gain. There are some basic rules to follow for first, second order active filters. To find out more about these, try finding a decent Active Filter Cookbook. There are also some free pdf's on the Texas Instrument website.

I suggest you really try to master the concepts because you will need to know how to analyze this stuff by rote if you ever want to design your own pedals.

There's lots of good info about tone circuits on http://www.muzique.com

rackham · Post by **rackham** » 14 Oct 2008, 17:43

There are a few RC filter calculators available on t'internet too that can speed things up for you.

I think there's one on muzique.com

Duckman · Post by **Duckman** » 19 Aug 2010, 05:33

Hey guys!

This old thread seems to be the right place to ask, since my search wasn't successful:

Once you know what frequency corner you need to tame, you can reach the same fc value with differents combinations of R and C... example

2k2/.039uF - 1855.9 Hz
22k/.0039 - 1855.9 Hz

How this affect the RC filter performance?
I suppose that higher values of R, less voltage through the filter, so less signal... is that right?

Thanks for your help!

marshmellow · Post by **marshmellow** » 19 Aug 2010, 09:41

Duckman wrote:I suppose that higher values of R, less voltage through the filter, so less signal... is that right?

No, it's not. In an ideal environment, they are perfectly the same. It all depends on your surrounding circuitry, what comes before the filter and what comes after it. The filter has a frequency dependent impedance that has to be driven by the part that comes before it. Say you have 1V sine signal, if you want to put that into a 1k load, you can determine what current the driver has to be able to deliver.

Take the (almost) ideal circuit device, an opamp, it has near zero output impedance. A 12AX7 in a typical guitar amp implementation might have somethinkg like 50k. The output impedance of your previous stage, and the impedance of your filter, now form a voltage divider. That means, if you don't want to lose a lot of your signal, and to not overly stress your preceding stage, the impedance of your filter has to be higher than the output impedance of the driver. General rule of thumb is a factor of 10.

Practically speaking, for an opamp, the combination 2k2/.039µ should be fine. For a transistor stage 22k/.0039µ would be better suited in most cases. For the tube guitar amp even another order of magnitude higher.

After the filter comes the next stage. The filter obviously also has an output impedance, so this next input stage again has to have an appropriate, high enough, input impedance.

Duckman · Post by **Duckman** » 19 Aug 2010, 21:22

marshmellow wrote:...That means, if you don't want to lose a lot of your signal, and to not overly stress your preceding stage, the impedance of your filter has to be higher than the output impedance of the driver. General rule of thumb is a factor of 10....

Thanks, marshmellow!

Clear enough for me (finally!!)

RiZsho · Post by **RiZsho** » 29 Aug 2010, 08:02

A good (and free, at least for circuits under 20 parts) is ICAP 4.
http://www.intusoft.com/demos.htm
I worked with it a lot in my 4 years at the university.