Power switching.

[kaycee]

So most people use the switched input jacks to 'turn on' the power in their pedals, whether using a battery or a power supply. Putting a jack in connects circuit ground to power ground and your away. I know that there are issues with this methodology, but that aside it means that if you leave a jack in the socket you are powering the circuit ( potentially draining your battery if you use one) .

If you are running a transistor fuzz etc the fact that its running all of the time doesn't really matter much other than the battery issue if you use one. But if the circuit has a part with a finite life, such as a filament bulb, or a valve then on those occasions when you want to leave that pedal on your board but not use that session, having a power on/ off switch would be handy.

The answer seems obvious, break either the power or ground line with a SPST. Does it matter which? Is one better that the other?

[GibsonGM]

As you noted, the plug connects the grounds to complete the circuit. So you could put a SPST in the ground line at that point, just before or after the jack.

Traditionally, however, circuits break the hot line just after it enters the enclosure.  Not important on a 9V pedal, but a 300V tube amp, more so ) You NEVER switch the neutral side with house wiring, for example - always hot, or your chassis can be hot.  Good practice to get into in case of internal faults and whatnot.   

I wouldn't worry for this application, though - wherever it's convenient! 

[R.G.]

Like "true bypass" signal switching, input-jack power switching of the ground-side power of an effect is a solution to a problem we don't have much any more.

It persists because of a number of reasons. Some of these are "That's the way all the good old vintage pedals did it," its simplicity and ease-of-understanding for beginners, which the vast majority of pedal builders are, and it's value in advertising copy.

The 9V battery is rapidly becoming a voltage standard, not a thing most people use to power their pedals. Back when people had one or two pedals, it made sense. When a small pedalboard has half a dozen pedals and 9V batteries cost $2 or more each, that gets expensive FAST, especially when unplugging every pedal in a pedalboard is not easy to do. In most countries, buying a DC power adapter which is designed specifically to power pedals costs about the same as 10-15 batteries, sometimes less. Both ease of use and economics say "use a DC adapter".

But there are technical reasons not to switch the ground of a pedal. GGM notes some of these. One reason I keep coming back to is that switching the negative side of power to a pedal with a stereo jack on the input ->maximizes the potential for grounding problems <- in high gain pedals.

If you use the stereo jack trick, 100% of the "sewer ground" currents from the pedal flows on the input ground wire, and through jack contacts of questionable integrity. This maximizes the voltage across that "wire" and it's directly in series with the input signal. This sets up the biggest potential for current-induced feedback to the input of the pedal. It often works fine - until the jack gets dirty  or old. If you have to do jack switching, do it on the ground wire to the output, not the input.

Even better, switch the power side, not the ground side with a single bipolar transistor.  See
http://www.geofex.com/FX_images/PNP_power_switching.pdf
which is a cleaned-up version of a much earlier article on the same issue.

[kaycee]

Thanks GGM and RG, I knew that there were some worthwhile considerations to be made. In this instance I'll break the hot line in, for future builds that need to incorporate a battery I'll look at RG 's transistor switching idea.  Thanks!

[JRB]

Definitely break the 9V wire and not the GND. I have implemented a power on/off switch in al my pedals and the first batch I wired it into the GND just following the good old use a jack to switch it.
On its own it worked perfectly. However whenever you used multiple effects powered by the same source it only worked if you switched everything off.
The power found a way back to the source through the patch cable via the other effect bypassing the switch. Doing it the other way around fixed it.