Power Filtering

JakeAC5253 · Post by **JakeAC5253** » 28 Dec 2010, 17:53

Can someone introduce me to power filtering basics with regard to stompbox design? I've seen designs with one 100Ohm resistor in front of a 100uF capacitor and other designs with 100Ohm resistors surrounding a 220uF capacitor. I've seen designs with one cap, and designs with two caps. Can someone explain the reasoning behind these designs and what are the benefits of designing something a certain way over the other. Or can you refer me to some online literature that I can read to better understand the design process. Thanks!

JakeAC5253 · Post by **JakeAC5253** » 02 Feb 2011, 04:47

Anything?

.Mike · Post by **.Mike** » 02 Feb 2011, 06:41

According to my understanding...

Using a 100R in series with V+ followed by a 100uF in parallel creates a low pass filter. The formula for figuring out the corner frequency is Fc = 1/(2*pi*R*C). R is in Ohms, C is in Farads. So, for 100R/100u, Fc = 1/(2*3.14*100*.0001), or 15.9Hz. If you increased the R to 1K, the corner frequency would shift to 1.59Hz. Reduce the R to 10R, and the corner frequency shifts to 159Hz. Similarly, if you stuck with the 100R and increased the C to 1000u, the corner frequency shifts to 1.59Hz, decrease the C to 10u and the corner frequency shifts to 159Hz. If R or C goes up, the corner frequency goes down, and vice versa.

The resistor causes a voltage drop according to Ohm's law (V=I*R). So, if you were using a 100R and your circuit was pulling 10mA, the voltage drop across the resistor would be .01*100 = 1V. If your supply is 9V, you would measure 8V after the resistor. If you use a 10R, the drop would be 0.1V, and you would measure 8.9V after the resistor. If you were using a 1K, the drop would be 10V (!), and the drop would be greater than the battery can put out. As current increases or the R value increases, the voltage drop increases.

When using a basic LPF like this, you have to balance what works best for you circuit. Too much resistance combined with too much current draw, and you lose too much voltage. Not enough resistance, and the corner frequency increases to the point where the filter is ineffective. Use too large of a capacitor and you physically might not be able to fit it in the box.

For more information, a search on power supply filtering or low pass filters would be a good start.

Mike

JakeAC5253 · Post by **JakeAC5253** » 02 Feb 2011, 15:13

Thanks for writing back! I'm actually familiar with how RC filters work, and have a basic understanding of how supply filtering/smoothing works, but what I searched on google didn't answer my question. I was more asking why would one choose to use one filtering setup over the other in the original examples. You've answered the resistor part of the question, but how about the cap? My overdrives each have a single 100uF cap, and my fuzz has two 220uF caps in them. According to RC filter rules, all you get is more low end going to ground, but what does it mean in power filtering land? Less smoothing? More smoothing? More stable AC sin wave?

Hides-His-Eyes · Post by **Hides-His-Eyes** » 02 Feb 2011, 19:11

More capacitance = cleaner power, more current available for moments when more is needed (think LED based switch pop, etc!)

JakeAC5253 · Post by **JakeAC5253** » 02 Feb 2011, 19:19

Hides-His-Eyes wrote:More capacitance = cleaner power, more current available for moments when more is needed (think LED based switch pop, etc!)

That's great, thank you! So more capacitance would lessen the switch pop?

And how does cleaner power affect the sound? Is it a similar effect to putting a gain smoothing cap across a clipping stage, but for the entire circuit instead of just the one aspect? Like chopping the random peaks and such?

merlinb · Post by **merlinb** » 02 Feb 2011, 21:28

JakeAC5253 wrote: And how does cleaner power affect the sound? Is it a similar effect to putting a gain smoothing cap across a clipping stage, but for the entire circuit instead of just the one aspect? Like chopping the random peaks and such?

If the power isn't clean then you may be able to hear hum mixed in with the audio. The cleaner the power, the less the background PSU hum. (And the less the LED pop, as mentioned already)

Transistor circuits like fuzzes are far more sensitive to PSU hum / pop than opamp circuits, so they often have more brutal power filters than those you find in an opamp design. (e.g., opamp circuit might just have a modest cap, whereas a fuzz might have an RC filter).

JakeAC5253 · Post by **JakeAC5253** » 04 Feb 2011, 13:05

OK, thanks. Now is it like a drag and drop thing? Like if I wanted to upgrade the filter caps on any of my projects all I would have to do is buy a larger cap value and install it right?