Bypassing an ARP Quadra Phaser-Suggestions?

[Gripp]

Hi Forum!

My interest in synthier sounding guitar textures has led me on a building spree. It started out with the Rocktave Divider via Tim E's SSWS, String Ringer (Lovetone Ring Stinger workalike; almost done, in box, needs wiring up and testing) to finally the ARP Quadra Phaser, Jurgen Haible version.
I really liked the samples Mike Irwin had on the forum ( http://www.elixant.com/~stompbox/smfforum/index.php?topic=35776.0 ) and figured it will probably sound nice with a 2 octave down square wave from the rocktave.

Anyhow, the pcb is soldered and is just waiting for the CA3086's (backordered) to go in their sockets.

All info on this build is here: http://home.debitel.net/user/jhaible/additional_schemos.html

Schem: http://www.oldcrows.net/~jhaible/tonline_stuff/hj_qph_1.gif and http://www.oldcrows.net/~jhaible/tonline_stuff/hj_qph_2.gif

Overlay and layout (including some additional supply decoupling not shown in the schem):http://www.oldcrows.net/~jhaible/tonline_stuff/hj_qph_pcb.gif and http://www.oldcrows.net/~jhaible/tonline_stuff/hj_qph_comp.gif

So far I've decided to house this thing in a half width 19"-rack box for a number of reasons. The pcb and internal transformer takes some space and a stompbox of the required size will be too high for my liking, half rack is at least a standard and I haven't decided where exactly it is going in the effects chain yet.   

Since it is a stereo effect I will probably run it into a tc fireworx (for delay/reverb) to ss poweramp at higher than stompbox levels.
Note, the stereo aspect, as I understand it, is this: one channel has the phase shifted signal mixed with a non-inverted clean signal and the other has the phase shifted signal mixed with an inverted clean signal. So using whichever output in mono will work but it will probably shine in stereo. This leads to my questions:

How would you bypass this thing? One design criterion would be that the existing on board +/-12V psu is to be used, and I don't want any extra spikes or clicks on those lines. Another design criterion is that it will be bypassed by either a latching make contact at the box or a latching make externally (by a separate footswitch, external midi-relay switcher etc). As a fan of Tim E's circuits, it would be nice with a really simple solution...wouldn't it always be

I have read all about relay and cmos switching at GEO and the wicked switch article at Andrew the tone god's site and considered both options. Right now I'm leaning toward buffer the front end with an opamp, split to input to pcb and to a CD4066 that switches either the clean buffered signal to both outputs (clean stereo split) or the effected outputs to its respective jack (acting as a DPDT along these lines http://www.geocities.com/thetonegod/switches/switches.html ).

Any problems with this line of thought considering the input stage of the quadra phaser, signal levels, psu, clicking when switching etc?
Am I doing this more complex than it should be? How would you do it? My mind is sort of stuck in a rut here...
Any input is most welcome!

Best!
Pelle Garpebring

[Mark Hammer]

Believe it or not, the simplest way to bypass the *effect* is probably to simply to lift the allpass network connection to the mixer stage at S2.  That's essentially what commercial manufacturers do with their FET-switch schemes - everything stays in place but the allpass connection to the mixer is interrupted.  The Arp Quadra is obviously a more complex circuit but the combination of allpass and clean remains common to this broad category of effect.

[Gripp]

Thanks for the comment Mark.
Yes, I've actually though about that too. A possibility that only requires a hack on the pcb and some switching element.
The hack itself won't be that easy though due to the layout and the "clean" will still go through the compander, possibly not a bad thing either.