Filter caps with a battery?

[SonicVI]

Are power supply filter caps (the 47uF cap in the Tone Bender MK2 for example) necessary when you are only using a 9V battery?  Since the power is not converted DC there shouldn't be ripple. Is there some inherent property (noise or something similar to ripple) of a 9V battery that still makes a filter cap beneficial?

[petemoore]

  Batteries and Caps have in common the ability to store DC.
  I've never detected ''ripple'', unless you count the discharge rate .
  I used battery as a DC standard by which I would compare to and judge small DC power supplies.
  Then I started using floating, filtered/regulated DC supplies, and yes, filtering right at the circuit.
 

[OC16]

Avoid motorboating sound if the circuit tries to self oscillate at low frequencies.

[Mark Hammer]

Let us say you have a circuit which consists of 3 dual op-amps,  At the point where the power source reaches the board, you have a smoothing cap of, let's say 220uf.  Why do each of the individual op-amp chips not have their own smoothing cap?  because the power has already been smoothed, prior to distribution across the chips.

A pure DC supply is already smooth.  Consequently the puny little individual transistor circuit fed from the battery is good to go being hooked up directly to the battery.  This is why you see so many "classic" circuits not including a smoothing cap; it was assumed they would be exclusively powered by a battery.  This is also why you see so many redrawn schematics of classic circuits including an additional smoothing cap; because the re-drawer has no idea if you will be powering it with an external supply or not, and whether that supply has any audible ripple.

There are exceptions to the no-cap-if-it's-battery-only principle.  The textbook example would be anything that includes an LFO sub-circuit.  These can often require sudden current draws that introduce spikes on the power lines.  In that instance, providing smoothing for the LFO and audio path on an individual basis can help to reduce audible ticking.  In that case, it is not the stability of the original battery source that needs regulation and smoothing, but rather the "competition" between segments of the circuit for that power.

Which reminds me.  Recently I was tinkering with a phaser and was surprised by the audible ticking.  I was surprised because I had not heard it previously from that very same board.  When I replaced the battery with a fresh one, however, the ticking stopped abruptly.  No circuit changes, only the battery.  So why did this happen?  Because the ticking occurs as a result of one segment of the circuit suddenly taking away precious (and limited) current away from the rest of the circuit.  When the available current was increased by means of a fresh battery, there was enough to go around for everyone and the circuit segments did not have any problems sharing.  So, no more ticking.

That might explain some of the problems with ticking that people have posted over the years, involving circuits that opthers had never experienced ticking with, and that were supposedly "protected" against audible ticking.

[Paul Marossy]

That's a good explanation, Mark. IMO, phasers, flangers and similar pedals with lots of IC chips in them ought to be run on a power supply anyway because they often eat batteries very quickly. I wonder, how long a battery would last with an ADA Flanger? It would not be long, that's for sure!  

[SonicVI]

 This is also why you see so many redrawn schematics of classic circuits including an additional smoothing cap; because the re-drawer has no idea if you will be powering it with an external supply or not, and whether that supply has any audible ripple.

Many classics originally did have a cap though. I gave the Tone Bender MK2 example, because it was obviously intended to only used with a bettery yet still has a 47uF smoothing cap.  The Rangemaster would be another example as well as most if not all of Sola/Colorsound fuzz pedals.  Did the designer probably just not know any better, or were they predicting a future in which AC adapters would be commonplace and that their pedals would still be used in that time? 

[Electric Warrior]

Even more interesting: They started out with a 25µF and then switched to using two 25µF caps in parallel. Later units use up to 64µF. They must have thought it was an improvement for something. Why waste a cap if it isn't doing anything?

[alanlan]

All batteries have internal resistance which is not easy to define or derive from the specs on batteries.  For the Duracell alkaline it's stated as 1.7 ohms @ 1KHz.  But what is it for an old non-alkaline battery?  What is it at higher frequencies?  Without smoothing caps, any change in current draw from a circuit will modulate the supply voltage.  This may or may not cause any problems depending on the circuit and the battery.  If you had a circuit with a high gain amplifier (without much power supply rejection) and something else drawing significant changing current from the supply, then you will no doubt have a problem.

So it's not true to say a 9V battery supply will always be clean when in use, but it's also not true to say that you always need smoothing caps.  Consider the circuit and how each stage works and then decide.  It's usually easier not to think though, for the price of a cap.