Cap switching pops: harmful to other components?

[Bucksears]

Since I can't find a cap-switching toggle-switch method that I can fully understand (feel like a moron on this one), I'm about ready to just bite the bullet and use a 'popping' config to simplify a cap switch, but I'm concerned that with the caps releasing voltage that it would hurt other components or throw off the (biasing of a transistor).

Should there be any concerns with excess voltage being released?

Thanks,
Buck

[R.G.]

Should there be any concerns with excess voltage being released?

Mostly, no.

If you're after the very, very best noise performance, maybe. It takes about 5-7Vdc to reverse-break the base-emitter of a bipolar, and that's what degrades. In general, the caps in effects do not store enough charge to burn things out. Maybe the pop through a Goliath (tm) amp would burn out a tweeter.

cap-switching toggle-switch method that I can fully understand (feel like a moron on this one), I'm about ready to just bite the bullet and use a 'popping' config to simplify a cap switch

Let's un-moron you, since switching configurations are hard to do.

To fully understand the switching method, draw the full circuit one time for each position of the switch. Then erase the switch position that's not active, and any caps that have been open circuited. The open switch poles/throws effectively vanish from the circuit when they're open.

To understand popping, all you need to know is that setups where caps are open circuited will pop. Setups which hold the caps to the same DC voltage with big value resistors instead of open circuits will not pop. Imagine putting a 4.7M resistor between the pole of a switch and every throw. That connection does not pop as a result of capacitors when the switch is thrown.

[remmelt]

What about like this?



Also see this thread:
http://www.diystompboxes.com/smfforum/index.php?topic=54829.0

As noted by Mark Hammer in the other thread, you'd need to figure out cap values that make sense for your application.

[remmelt]

Or, to reply to my own post, like this:



S1 is SPDT, with either ON/ON or ON/OFF/ON. Especially the centre off one is fun: three different modes with in the centre setting all three caps in series. Caps in series are the same as resistors in parallel, click link for more.

Since all caps are always connected to ground, no popping should occur. This is how I understand it, at least.

[Bucksears]

Do both ends of each cap have to be attached to something or can just one end be switched to ground & the other open and still avoid popping?

Thanks for the responses guys!

- Buck

[R.G.]

Any time a cap is open - for instance when a switch disconnects one end of it - it will relax toward 0V. The higher the quality of the cap's insulation, the more slowly this happens. Teflon and polystyrene relax very slowly. Electros just roll over.

If (a) a cap is set up so that when it's connected it has a DC voltage across it and (b) it's ever open circuited then it will leak down and pop when reconnected. The size of the pop depends on how much it changes while open circuited.

You can fix the popping by using high value resistors to hold the "open" terminal so it's never really opened, only in series with a  high value resistor. When you do this, the high value resistor holds the otherwise-open terminal at the correct DC voltage to not pop. Caps in series can still pop because all of the caps will still leak down to different DC values, as determined by their equivalent parallel leakage resistances.

A definite way to prevent popping on two-toggle setups: Input to a resistor to ground, 1M to 4.7M; one end each of two caps ties to this resistor. The free end of each cap ties to its own 4.7M resistor. The free ends of the 4.7M resistors tie together and this goes to the signal in into the circuit. The switch is connected with its pole at the junction of the two 4.7Ms and one throw each to the junction of a cap and its 4.7M resistor. The toggle switch shorts out one of the 4.7M resistors at a time. The remaining cap is held at the proper DC level by its own 4.7M resistor and the pulldown on the input.

If you use a center-off toggle, you can tie a third cap permanently between the input pulldown resistor and the pole of the switch. Now when the switch is center-off, you get this third cap. The other two switch settings parallel one of the two side-caps with 4.7M resistors.

[remmelt]

All right, so it would look like this for the ON/ON switch:

[Bucksears]

Thanks again everyone.
What I'm trying to do is this: on the BSIAB II, I'm trying to toggle between the stock C2 cap (.068uF) and a higher value cap w/pot so I can switch to an 'upper mids' boost setting; I don't think I can do the 'center tap' between two caps on this method.
Seeing as how the cap doesn't connect directly to ground, it's a little complicated (I think) to get switch it. I may just skip it for now and think on it later.

Thanks again,
Buck

[R.G.]

Why wouldn't you just replace C1 in remmelt's drawing with your cap and the pot? Looks ideal.

[Bucksears]

If the top two poles of the switch are connected in remmelt's example 3, what is the path of the signal?
IN --> C1 --> OUT?

Wouldn't R2 & R3 affect the sound given that they are connected directly to OUT?

[R.G.]

If the top two poles of the switch are connected in remmelt's example 3, what is the path of the signal?
IN --> C1 --> OUT?

Yes.

Wouldn't R2 & R3 affect the sound given that they are connected directly to OUT?

Yes. But it's a matter of degree. First of all, R2 and R3 are chosen deliberately to be BIG. The switch is a toggle, and is always shorting out one of them, either R2 or R3, so only one of them matters at any given moment. So the remaining switch that's still in series with its capacitor limits the signal that the "off" capacitor passes. For all but very unusual circuits, you can ignore the cap that's left in series with its resistor. Put another way, I can design circuits where R2 or R3 would make a difference, but I'd have to design the circuit specifically to do that. In ordinary circuitry, the one with a resistor in series with it is effectively removed from the circuit.

Look at it another way:
It's obvious that if R2 and R3 were infinite (i.e. an open circuit) then they would have no effect on the sound and you'd be left with switching between two caps and still having popping.
It's also obvious that if R2 and R3 were, say, 100 ohms, that it would never pop, but you would not get very good switching.
Is there a place between 100 ohms and infinite that both does not pop but has an acceptable degree of audio interference between the two caps?
My experience leads me to think yes, there is, and it's in the 1M to 4.7M region. Shoot, use 22M for them. All that does is get closer to popping, but with poly caps it might be fine.