Ross Phaser, Univibed - Stop popping when switching modes?

[Freddy205]

Hi there, I recently built a ross phaser which can be switched into univibe mode via a tpdt switch. However when I do switch modes I get a loud POP which can get very irritating. Is there any solution to this? For anyone who needs info on univibing a Ross just search the forums Thanks for any help.

[Mark Hammer]

Ever go for a long time without taking or being able to take a leak?  When you finally get to a "proper receptacle", does it feel like your nuts are going to explode?  Are you prompted to let out a little grunt/scream/moan of relief?

If yes, then you know how a capacitor "feels" when it has nowhere to drain the charge it has stored up.  The popping you hear is a semi-regular product of placing a cap in a path where it can accumulate charge, and then lifting one or both ends of the cap so that charge can't drain.  Typically, the larger the value of the cap, the more charge it can store, and the louder the pop when the free end is reconnected again such that current can flow.

There are two solutions to your diliema.  The simplest is just some advice: don't do the switching unless the pedal is bypassed.

The second is probably something you're not going to like because it involves some math and likely will bugger up your board or require a new one.  I credit John Hollis with opening my eyes to this one.

If you are like many, you have probably modded your pedal for Univibing by having an alternate set of capacitors and using a toggle switch to select between the original values and the Univibe-friendly values.  Chances are, the "other" set is sitting with one end free after you make the switch.  The pop you hear is the sound of the unused set getting some breathing space once reconnected.

Is it possible to do the needed switching without ever lifting an end of any cap?  Yes.  The answer lies in use of series caps.  Remember that the effective capacitance of any caps in series is given by the formula 1/C = 1/Ca + 1/Cb +...1/Cn.  Hollis used this in the Zombie Chorus to shift the delay range by x2.  A pair of 1000pf caps are placed in series, straddling the pins that set clock range.  In series, they produce 500pf of capacitance.  When ONE of them is shorted out (by a straight wire connection bridging it), the effective capacitance becomes 1000pf.

The analogy to the "popless" Phaser/Univibe conversion is that each phase shift stage will need to have two caps placed in series.  The switch will "select" by shorting out the one you don't want to use at the moment.  Since the shorted out one always has a path to drain any charge (can't get much more efficient than a piece of wire), no pops will occur.

A 4PDT would allow you to avoid all the math, by having the switch shunt the Univibe caps in one direction, and shunt the phaser caps in the other direction (the switch common goes at their junction).  That should do it.  The "easier" method might be to use something like a 4016 or 4066 CMOS and a SPST switch to do all four from one switch.  It would need a very different board though, obviously.

[R.G.]

Mark, I have to admit that your explanation of capacitor popping puts a whole new spin on the "water analogy".  8-)

There are actually two ways to do popless capacitor switching, and I only add this one for completeness. If you have two capacitors in parallel and open/close one of them, you get popping because the open cap leaks down a bit when it's open. The solution in this case is our old friend, the pulldown resistor. Only, used this way, it's more of a pull sideways resistor.

The way you do this is to call one of the caps the primary cap; this one is always in the circuit. In parallel with the primary cap is the secondary cap in series with a 1M or so resistor. The switch "shorts out" the 1M resistor, and when it does, the total capacitance becomes the parallel capacitance of the primary and secondary caps.

Actually, you can put 1Ms in series with *both* caps and alternately short one or the other 1M resistor; this is the same as a selector switch to select one or the other cap, but without the popping. Generally, one cap is left as a "primary" cap, because this saves one throw on the switch. This also works with rotary switches. But then the series string of caps does too if you just hook a rotary up to each node of a string of caps and get fancy about setting up the cap values in series.