Selecting Between Capacitors

[darth_b]

I would like to add some variability to an effect I'm building by being able to choose between a set of capacitors.  Instead of manually swapping out caps by using a socket, I would like to use say a SP5T switch to choose which capacitor is active at any time.  I would wire one lead of each cap to the same spot on the board, and the other cap leads to their respective lug on the switch.  The output of the switch would then go back to the board.
Does anyone see any issues doing this?  Or am I breaking some electronic rule?

[Gus]

Make sure you get the right kind of switch.  This is nothing new.

make before break or break before make switches is what you should read about

[Mark Hammer]

I've been a vocal advocate for using series switching of caps wherever possible.  The problem with selecting between caps is that charge can get stored up in a cap with one leg "hanging" free.  The moment you provide a path for that stored charge to drain off, it drains off super-quick and you hear an audible "pop" as a result.  If what you switch is somethng that is never or rarely done while on a stage with the effect engaged, the pop will not be objectionable, but if there is the remotest chance that you might have the pedal on, and the amp up while you do your switching, the pop can be very annoying.

Fortunately, it is possible to use a series approach to switching, though there are some practical constraints that come with it.  When caps are  placed in sries, their effective capacitance is calculated the same way parallel resistors get calculated.  So, a .22uf cap in series with a .1uf cap is equal to 1/.22+1/.1 = 1/C  (4.545 + 10 = 1/.0688) or .069uf.  If I shunt (i.e., place a straight wire connection across) the .22uf cap, then I'm left with .1uf.  Shun the smaller cap alone, and the effective capacitance is now .22uf.  If you had a 3-position SPDT (center-off) toggle, and wiored it so that one side position shunted the .22uf cap and the other side position shunted the .1uf cap, you'd have three different capacitance (.22, .1, .069) for the price of two.  More importantly, NONE of the caps would ever have a lead "hanging free" so they would always have a chance to drain and never pop.

Not a perfect solution to every circumstance, but very useful for a great many.

Depending on the application, you can also reduce the impact of two or more caps by placing each in series with a large-value resistor and using a switch to simply shunt/defeat the resistor.

[96ecss]

Here is an example from my gallery of how to make a cap switchable http://www.aronnelson.com/gallery/v/Daves-Layouts/Hot+Silicon.gif.html

It works great and does not pop at all.

Dave

[96ecss]

Oops, I just realized you were referring to switching between 2 different caps. The example I gave just switches 1 cap in and out. Sorry.

Dave

[darth_b]

I like the series cap switch idea.  It's still hard for me to wrap my head around that having caps in series actually results in a capacitance lower than either of the individuals.
So, in order to get a 5 position cap switch for the following values
0.0047 uf
0.0033 uf
0.0022 uf
0.001 uf
0.00047 uf

Then I could use
0.0047 uf
0.0110785714 uf
0.0066 uf
0.0018333333 uf
0.0008867925 uf

in series, adding each one in turn using the switched shunt.  Of course these are calculated values and I would have to find a close actual cap.

[Mark Hammer]

For that sort of intended range of values, the math may not exactly work out in your favour, and maybe not the switch architecture either.

It might help if you tell us what you want to witch the cap values for.  It is quite possible that 6 values is way overkill, implying that a 3-position toggle might do what you want with less hassle and cap-choice difficulty, or that you can even accomplish what you want with a variable-big-cap-bypass like Joe Gagan pioneered.

[Gus]

Here is another hint
neve 1081 EQ switch wiring

[MikeH]

The matchbox project at runoffgroove.com does exactly what you're talking about.  You can check out a schematic there and see how they do it on a pcb or perf.

[axg20202]

Also check out the Big Muff Pi project document at Tonepad - it shows a cap switching mod shown with a description of how the cap values add up to the desired capacitance value per switch position.

http://www.tonepad.com/getFileInfo.asp?id=94