Remote Mechanical or CMOS Switching

[liquids]

Okay....I've been pondering this idea...
Hammond's 1444 series enclosures can be huge and cheap for the amount of space you get, for one large circuit, or even multiple circuits housed in one unit.  
These are plenty sturdy to house circuits,  but flimsy as -or flimsier than-the old bent-aluminum EHX enclosures - which IMO not the ideal for mounting a real mechanical stomp switch to.

I see plenty of guys, even pro's with pedal boards that use a long Hammond 1411 series enclosure at the bottom, lined with 3PDT switches, LEDs and in/out jack for each pedal on the board, so that all the switches are in a row and the pedal board merely holds the pedals (with wiring to/from pedal to 1411).  That is typically cheap aluminum too, maybe they use the steel versions?  But still, unless the steel versions are solid and easily hand drillable, mounting a stomp switch to think aluminum is not what I think of as wise for stomping.  Steel, maybe, but aluminum is easy to drill, and familiar, so the thicker aluminum is my starting point for stomp switch mounting.

That premise, I got to thinking about alternatively using the 'standard' thick enclosure many of us use to build pedals in (like the 1590 series), but to put nothing in one (small or large) but jacks, stomp switches, leds.
That said, while in/out would use mono cables....a stereo cable could serve as a send/receive to a remote pedal, one jack...and if the remote pedal unit (such as a huge 1440 series chassis) was designed as such or modified as such, in/out on the pedal could take place on one stereo jack.

That said...apart from the need for buffers to drive the inherently extended runs of wire via remote runs of 'stereo' cable to/from stomp unit and circuit (got buffering covered on all fronts, trust me), would running a to/from guitar-level signal on a stereo cable just asking for crosstalk etc?  In my mind, by the time the signal is out of the circuit, it's back to guitar-level amplitude/volume...and a good stereo cable will be shielded, so the in/out will be sharing and an parallel with the same ground connection...

Does it just depend on the cable, or should running long lengths of wire with in/out signal in parallel a bad idea?

Stereo cable configurations only enter my realm of possibilities because  I can make them myself for fairly cheap, at my own desired length, etc.

Likewise, I haven't planned it out...but I imagine that it wouldn't take much work to utilize 9v and 3 of the 4 switches on a CMOS switch and either a simpler or some form of tactile stomp arrangement to do the switching, at that, maybe.  Whatever enclosure used to do stomp duty would probably have the space to spare for a small switching board.  I also thought I saw something like that on an old Craig Anderton design....as opposed to the Boss/Dod/etc FET scheme, which seems to work just fine, but not worth the effort for me to go to that length - everyone has to draw a line for themselves somewhere.  

Any thoughts, or better yet, experiences, are appreciated.

[R.G.]

Good idea.

Have you read the "steel studs" article at Geofex?

That said...apart from the need for buffers to drive the inherently extended runs of wire via remote runs of 'stereo' cable to/from stomp unit and circuit (got buffering covered on all fronts, trust me), would running a to/from guitar-level signal on a stereo cable just asking for crosstalk etc?

Yes.

In my mind, by the time the signal is out of the circuit, it's back to guitar-level amplitude/volume...and a good stereo cable will be shielded, so the in/out will be sharing and an parallel with the same ground connection...

The problem is that there is a long run of wires where the high impedance input of the effect is parallel with the low impedance and probably highly amplified output of the effect. The capacitance between the wires will let the output feed back to the input and will cause oscillation with some circuits. Others, with low gain or different phase responses will get away with it. And it may cause odd resonance peaks/valleys even when it doesn't oscillate.

Does it just depend on the cable, or should running long lengths of wire with in/out signal in parallel a bad idea?

It depends on everything, but in general outputs and inputs should not share long parallel runs of cable without being shielded from each other.  It's just one of those things. Mother Nature says.

[soffa]

Quote from: liquids on Today at 11:08:14 AM
That said...apart from the need for buffers to drive the inherently extended runs of wire via remote runs of 'stereo' cable to/from stomp unit and circuit (got buffering covered on all fronts, trust me), would running a to/from guitar-level signal on a stereo cable just asking for crosstalk etc?
Yes.

I have used Belden 8273 (load cell cable with 2 shielded twisted pairs) coupled with good grounding practices with good success in a similar application. I'm sure YMMV depending on what you're running through it!

[liquids]

I see what you are saying.

Does the fact that my guitar always sees a stand alone buffer/splitter, and ALL the pedals in my board, outside of my wah pedal (hmm, need to fix that), send a low-impedance output signal make it more reasonable, or is it still risky (I'm assuming the latter)?  

What I don't get is that I routinely have seen stereo cables that split to mono send/returns for outboard effects use on mixerboards.  The assumption is that everything there is driving long distances, and/or  what often look to be unshielded thin-wire, high capacitance, dare I say HOSA stereo 1/4" to 2x mono 1/4" jacks....why is that different, or is that just as risky/lucky not to result in 'mother nature' wreaking havoc upon the sound?

Okay...so the alternative is four individual, mono jacks, and two shielded mono cables - george L's - invariably running parallel between circuit and switching box, if all else stays the same....lets say I got 'wild' and put each pair of input/output jacks as close to each otehr as is possible, and then twisted the cables that ran between circuit box and switching box.  Is this likely to result in the same issues as two stereo jacks and one stereo cable, or is it even better as a 'worse-case-scenerio' since the two mono cables at least have insulation and shielding between them no matter how much I twist them?

[liquids]

Quote from: liquids on Today at 11:08:14 AM
That said...apart from the need for buffers to drive the inherently extended runs of wire via remote runs of 'stereo' cable to/from stomp unit and circuit (got buffering covered on all fronts, trust me), would running a to/from guitar-level signal on a stereo cable just asking for crosstalk etc?
Yes.

I have used Belden 8273 (load cell cable with 2 shielded twisted pairs) coupled with good grounding practices with good success in a similar application. I'm sure YMMV depending on what you're running through it!

Interesting....luck, mojo, or design?  The world may never know, seeing as you are testing mother nature 

Where did you get this cable?  It's kind of spotty finding info/pictures etc from a preliminary google search.

RG - not to get too personal, but the inside of my Visual Volume pedal uses some kind of RG-45 type cable to connect between board if I recall correctly...maybe input and output arent directly next to each other, but is that a specialized circumstance, more doable because of the short runs, or just a risky design?

I'm not beyond usin friggin' XLR connectors and cable to do this via one cable and two jacks...I hate the thought of using two individual cables to/from the same place...so if someone can guarantee a way to do it via one cable, two jacks, no unreasonable compromises, I'd be all ears.

[R.G.]

Does the fact that my guitar always sees a stand alone buffer/splitter, and ALL the pedals in my board, outside of my wah pedal (hmm, need to fix that), send a low-impedance output signal make it more reasonable, or is it still risky (I'm assuming the latter)?  

What makes it risky is the high impedance on the input cable. If your guitar is buffered before it hits any of the send receive cables, that helps hugely. If all the cables are driven from less than several k ohms on each end, then you may never see a problem.

What I don't get is that I routinely have seen stereo cables that split to mono send/returns for outboard effects use on mixerboards.  The assumption is that everything there is driving long distances, and/or  what often look to be unshielded thin-wire, high capacitance, dare I say HOSA stereo 1/4" to 2x mono 1/4" jacks....why is that different, or is that just as risky/lucky not to result in 'mother nature' wreaking havoc upon the sound?

It's the driving impedance.

Let's look at the setup. two small-gauge wires with typical vinyl insulation have a capacitance between the two that's from 1pF to about 5pF per inch, depending on the wired diameters, insulation thickness and dielectric constant of the insulation. This fact is used by radio amateurs to tune RF stages, where it's referred to as a "gimmick capacitor" and is tuned by snipping little bits off the length. So if the cables are 6 feet long, and you get 24pF per foot, you have a capacitor between send wire and receive wire of 144pF. It's actually not that simple, but this will do for now. If the send wire is driven by a low impedance (that is, below about 10K, which will hit almost all buffers worthy of the name and almost all pedals), and the receive wire has a 1M input impedance and is not otherwise driven by a low impedance, then you get a voltage divider of 144pF and 1M. The frequency where these are equal is F = 1/2*pi*(1M*144pF) = 1106Hz. Again, it's not really this simple, as I'm ignoring several things to make the point. But at that frequency, you'd get a loss through the send back to the receive of 1/2. 2.2kHz, you get a loss of only 1/4 (that is, 3/4 left) and at 550Hz, the loss is 75%.

If the pedal out there amplifying has a gain of 2 at 1106Hz, it can become self oscillating if the phase adds up right - or wrong, depending on how you look at it - it will sing. It's easier at higher frequencies where the attenuation is less, and the phase changes more. More gain makes it even easier to sing, as smaller caps can let through enough signal.

If, however, the return/input side is driven by a 1K source, and isn't just a 1M input, the attenuation goes up by a factor of 1000. Now the gain to oscillate has to be about 2000, not 2 at 1100Hz. And you're back at the same problem up at 1.1MHz, but it's unlikely the circuit has a gain of 2 at 1MHz. Well, OK, you have much better chances that it's dropped off. This is what makes FETs deadly for oscillation - their input impedances are high, and they have high gain right on out into the hundreds of MHz.

So if the cable signal wires are driven from low impedances on BOTH ends, there's not likely to be a problem, because the cable capacitance transfer is pushed too high in frequency. (Note: watch those FETs though!).

As a side note, the twisting of the cables together, if they're twisted pair, introduces inductance above the per-length inductance of straight cable. So the cables may act like a tuned circuit and help the electronics at the other ends sing.

But low impedance on each wire, both sending and receiving, can make the oscillation for small systems unlikely. This is why studios, and the POTS system, are optimized around 600 ohm twisted pair. The 600 ohms is low enough that it makes the capacitance/inductance problems a lot less, and in many cases the cables are terminated in about-600-ohms resistors on both ends to damp things down.

Okay...so the alternative is four individual, mono jacks, and two shielded mono cables - george L's - invariably running parallel between circuit and switching box, if all else stays the same....lets say I got 'wild' and put each pair of input/output jacks as close to each otehr as is possible, and then twisted the cables that ran between circuit box and switching box.  Is this likely to result in the same issues as two stereo jacks and one stereo cable, or is it even better as a 'worse-case-scenerio' since the two mono cables at least have insulation and shielding between them no matter how much I twist them?

Yes. Shielding each run means each cable has a capacitance to ground, but the shields "intercept" the capacitive field pickup before it can get to the other wire. It's not perfect, but it attenuates wire-to-wire capacitance hugely. This is why shielding wires to/from controls in an oscillating pedal works. The shields "eat" the signal that would otherwise couple to other wires.

RG - not to get too personal, but the inside of my Visual Volume pedal uses some kind of RG-45 type cable to connect between board if I recall correctly...maybe input and output arent directly next to each other, but is that a specialized circumstance, more doable because of the short runs, or just a risky design?

It's special because the lengths are so short. (... he says while putting on a saintly smile and casting his eyes heavenward...  )

There is some risk in all wiring setups. If the tolerances add up against you, it's up to you to devine what went wrong (i.e., what Mother Nature whispered in your ear and you ignored) and fix it. Some people calculate worst-case capacitive coupling and gain-phase margin, yada, yada.   I did the other. I designed for ease of service and manufacture, then tested the bejeezus out of it.

One thing that RJ-xxx wire does is to reliably put every single wire in the same place related to each other, every time. So if it works for one, chances are it works for all of them. It's not like a random-wire situation where wires may have any relationship to each other at all. Predictability is often worth far more than high performance. I also fed prototypes test signals up to about 1MHz and watched for humps in the frequency response; the circuit was supposed to have flat responses except where I wanted to kill it off, so a hump was Mother Nature telling me I'd forgotten to think about something. She was quiet about this, so I took that as a good omen.

Good designers know that one of anything doesn't tell you how the 10,000th one will act. It's an indicator, but no more. So you look for margin - how close to the edge of the cliff are you? In this case, I judged that I was far enough from the cliff. So far, I've been lucky - I don't know of any we've gotten back for oscillation, and The Gear Page hasn't told me of others we didn't get back.

There are some Golden Rules of wiring.
1. Keep input and output as physically separated as you can.
2. Keep wires as short and direct as possible, taking into account 1 above.
3. Watch where the currents flow, and keep combined currents from flowing where they matter.
4. Worry about wiring should increase as the square of the gain of the circuits being wired.
5. Worry about wiring should increase as the square of the frequencies flowing in the circuits being wired.

And 6. Hope you stay lucky.

[PRR]

> 5. Worry about wiring should increase as the square of the frequencies flowing in the circuits being wired.

I would extend this to "square of the frequencies which could flow...."

Many "audio amplifiers" can not read their own name-tag, don't know they should only do audio, and happily amplify low radio frequencies. I built a power-amp that in-out barely did 15KHz, but the final stage would do 2MHz plate-to-cathode. Then there are all the 8MHz opamps which obviously can't turn 8MHz but sure can sing 1MHz even if the intent is a 5KHz low-pass for distorted guitar.

[liquids]

RG - thanks for your response.  

To more or less repeat what you said, you're testing with the VV pedal shows your knowing what COULD happen, you analyzed if it WAS happening in real life, felt it was far enough away from a problematic issue/failure rate to say 'good enough' and then crossed your fingers after shipment...and over time have not received a significant number of complaints/returns due to this issue cropping up, as you both tested for and prayed for.  

I think, while from a technical perspective (you latter need to cross your fingers, specifically), while that's not the 'ideal' in the theoretical world where over-engineer and not even coming close to having any of these issues ON PAPER, let alone need to test if they crop up in real life,, I'm more in line with a willingness to shoot for a REASONABLE compromise here, and still take SOME risk with high quality material that should mitigate the factors for the sake of the potential payoff benefits.  Then, I can work backwards if I encounter an issue....and only abandon ship if I encounter crosstalk issues that aren't remedied by disassembling the problematic cable, snipping a new length of the same "could be good enough' cable, re-soldering to new plugs (just to isolate that variable too), and getting the same issue....

Keep in mind, this is just for me, not for mass production, micro production, or a customer.  Likely NEVER to even be utilized in a live environment.   The needs or risk of failure is more time and effort based and mental.    I just need to get something to work for ME, that will not end in utter failure and pulling out my hair in trying to get it to work without serious, intermittent, or repeated issue.  

Running parallel to this (pun intended), is the thought of having a switcher box with, say, 3 switches remotely switching 3 circuits (housed in the same big enclosure) with 6 mono cables between, in which my low level trichotillomania starts to kick in at the thought of....seriously.  

I'm sure someone or many won't get why I would want to go to great and risky lengths, "just to avoid 3 cables, 6 plugs and 6 jacks" in this theoretical (but potentially real) scenerio, but it just 'is' - and it's not a financial thing, might I add.

More questions.
1) if the cable says it's twisted pair, isn't that going to make the difference?  Or are most 1/4" stereo cables twisting their conductors as the assumption?
2) Soffa mentioned a cable that I believe has 2 pair of twisted pair....and ground sheild....if in/out conductors are run from one conductor of two separate twisted pair, and then each of the 'spare' conductors of two twisted pair are grounded with the shield, shouldn't that cover it??

3) This article: http://www.geofex.com/article_folders/cd4053/cd4053.htm    I read this article yesterday. I've read most of the ones on your site, but this one was either new or previously didn't interest me.  I didn't realize there were SPDT CMOS switching chips.  Joy.  However, it annoys me to need to reference all 3 parts of each individual switch to a VREF at all times.  Is there another, otherwise similar but 'improved' switch these days?

[liquids]

Or PS....for the switching mechanically vs CMOS realm of my plan, should I just learn how to use relays?  If so, where is a good place to start...is that in TAOE?  Is there a relay 'bible' that won't fly over my head with math?

[R.G.]

To more or less repeat what you said, you're testing with the VV pedal shows your knowing what COULD happen, you analyzed if it WAS happening in real life, felt it was far enough away from a problematic issue/failure rate to say 'good enough' and then crossed your fingers after shipment...and over time have not received a significant number of complaints/returns due to this issue cropping up, as you both tested for and prayed for.  

Designers with experience, if they're honest with themselves, will admit that this is always the case.

Mother Nature has one final trick waiting for everyone. After you learn Her Rules as best you can, you run into Her propensity for doing infinitely many things. No one can know everything which can possibly happen. The Rules are often boundary conditions; for instance, the First Law of Thermodynamics says in effect "you can't win". Matter and energy can't be created or destroyed, so you can't get more matter-energy out of a process than went in. The Second Law says "you can't break even," as there are always some losses. However, neither of these rules says how good or bad the situation is. That part is up to you to deal with in your designs. There's a lot of range between 99% of the energy doing what you want and 100% being wasted. 

I think, while from a technical perspective (you latter need to cross your fingers, specifically), while that's not the 'ideal' in the theoretical world where over-engineer and not even coming close to having any of these issues ON PAPER, let alone need to test if they crop up in real life,, I'm more in line with a willingness to shoot for a REASONABLE compromise here, and still take SOME risk with high quality material that should mitigate the factors for the sake of the potential payoff benefits.  Then, I can work backwards if I encounter an issue....and only abandon ship if I encounter crosstalk issues that aren't remedied by disassembling the problematic cable, snipping a new length of the same "could be good enough' cable, re-soldering to new plugs (just to isolate that variable too), and getting the same issue....

That's a reasonable attitude. Some people can't get there. We do get a lot of posts saying, in effect, that they've done everything right and it still doesn't work, so something other than them must be the problem. Sigh. Nature is a Mother.

1) if the cable says it's twisted pair, isn't that going to make the difference?  Or are most 1/4" stereo cables twisting their conductors as the assumption?

Hmm. Probably not. The inter-wire capacitance gets worse as more wire length is added. Twisting adds more lenght. Twisting cable pairs is most useful for wires carrying equal and opposite currents - differential signals - so that external influences affect each direction the same, and since they're opposite, the externals tend to cancel. Different signal situations.

2) Soffa mentioned a cable that I believe has 2 pair of twisted pair....and ground sheild....if in/out conductors are run from one conductor of two separate twisted pair, and then each of the 'spare' conductors of two twisted pair are grounded with the shield, shouldn't that cover it??

Cover it is probably too strong a term. It's certainly better. If the grounds are wired right, this gives you the situation that you have equal-and-opposite currents flowing in the signal/ground pairs and unique to the pairs. The ground wires and twisting certainly force the signal wires to be further apart, which all by itself reduces the wire-to-wire capacitance a lot, as they will only touch insulation at the odd places where the twisting lets them. This reduces the capacitance a lot, and that moves the frequency/impedances out to where they may not be a problem. It's probably enough for the situation as you've described it where you're willing to tinker if things go wrong, and there are not several thousand of these out in customer hands that have to be fixed under warranty.

3) This article: http://www.geofex.com/article_folders/cd4053/cd4053.htm    I read this article yesterday. I've read most of the ones on your site, but this one was either new or previously didn't interest me.  I didn't realize there were SPDT CMOS switching chips.  Joy.  However, it annoys me to need to reference all 3 parts of each individual switch to a VREF at all times.  Is there another, otherwise similar but 'improved' switch these days?

It is annoying, isn't it? Sadly, Mother Nature seems to demand it of the simpler and cheaper CMOS switches. My experience with CMOS switches is not encyclopedic. There are certainly newer CMOS switches these days, more all the time, and every one of them is "improved"; just ask the maker. 

Or PS....for the switching mechanically vs CMOS realm of my plan, should I just learn how to use relays?  If so, where is a good place to start...is that in TAOE?  Is there a relay 'bible' that won't fly over my head with math?

Well, there's "Relays Basics" at geofex.     Relays are not that tough, and if you're just using them, not much math applies. A relay is an electromagnet that pulls a switch into position or releases it. Over and done. If you pull enough current through the coil, it switches. Release the current (for single-side-stable relays) or pull it the other way (for latching relays) and the switch flips the other way. There is no electrical connection between the magnet and switch, only the unavoidable capacitance between everything and everything else in the universe, and this is usually "small".

For your purposes, it might make more sense to put a relay in a small box right out at the pedal, or in the pedal itself. See the Technology of Boss and Ibanez Bypassing (quote]http://geofex.com/Article_Folders/bosstech.pdf[/quote]) at geofex for one way to insert a relay into the pedal. Any way you pull current through the relay coil probably works. See "A Remote Footswitch Bypassing System" (http://geofex.com/Article_Folders/remoteftsw.pdf) for one way to do it. There's another variation shown in "Remote relay effects switching ", http://geofex.com/FX_images/relays2.gif.

Put the relays right at the pedals and the relay controls out where your foot is. The problems with running audio on long cables go away with the long cables.

[soffa]

What I did with the cable I mentioned was to use each pair as a ground and signal in each direction. I connected the shield to ground on one end only. 

if you wanted to make something analgous to two guitar cables in a single sheath, what if you connected the two wires in each pair together (or just used one) and then used the shields as your ground? (The shields for both pairs are not insulated from each other.)

I thought that the first situation was preferable, but maybe there is some advantage to the second in this case?

[wavley]

Put the relays right at the pedals and the relay controls out where your foot is. The problems with running audio on long cables go away with the long cables.

I actually did this  with my EHX Wiggler and English muffin, it worked quite well.  Puretube's site has a few of his pedals done like that.  The result was that I put two pedals that take up a lot of real estate on my board somewhere else and made room for more home made stuff.  My eventual plan is much like yours.  I was going to use RG's steel stud idea (I went to my lowes and they don't stock them anymore) and relay switching like in Puretube's stuff (and RG's article) to put all of my bypasses in a nice neat row.  I had kicked around using a PIC to do presets, but decided I would be just fine with a neat line of switches, I don't mind some fancy footwork in fact it's part of the fun.

Point is, I'm a pretty big fan of remote relay switching now and someday I'll find the time to implement more of it.

[liquids]

This thread may be heading in a good direction.  Thanks.  I have to learn a bit more about relays - they make no sense to me, and I only had about 25 seconds to read your post, but I'm hoping and currently optomistic that the relay implementation is not beyond my amatuer hobbyist electronics hack, who tries to take at least one datasheet or a techie webpage with every meal and visit to the men's room...figuratively and literally, depending on the day.   

[R.G.]

I've speculated on and tossed around with the boss at my day job the idea of putting another jack on our pedals that is a "remote bypass" jack, much like the remote pedal switching on guitar amps for switching reverb, tremolo, etc. The idea is that you could cause an engage/disengage remotely by grounding the end of a cable on that jack. Frankly, every Boss/Ibanez pedal can have this done to it for the cost of adding a jack. I didn't mention it in the "Technology of..." article, but it's pretty simple to just parallel a remote switch with the built-in footswitch on these pedals; no relay needed.

On pedals that don't have the Boss/Ibanez style switching or some other electronic flipflop driven setup, you'd have to build in an electronically switchable bypass.

[R.G.]

I thought to myself, "I really ought to just go update that article to show remote switching", and found that I had already done it, just didn't post the last page of the article.

You can read what I had in mind on Geofex now under the "Technology of Boss and Ibanez...", on the last page, Mods.

Switching logic signals is much simpler than switching audio.

[R.G.]

Another possibility is to use the single coil latching relay with a capacitor-dump circuit. I finally convinced myself that this circuit can be made to work reliably. 

So one could insert a single-coil-latching relay inside a pedal with an electrolytic cap, and bring out a three-wire cable to a footswitch. The footswitch is SPDT (or DPDT if you want a remote indicator too) and flip the relay from the end of the cable. The cable wires are +9V, ground, and "flip". The wiring for indication is done with an LED at the footswitch end of the cable, and may not work in sync with the pedal for a few flips, because the cap has to charge up correctly to make the relay flip.

But it can be done. A multi-footswitch needs one power, one ground, and as many flip wires as you have effects being controlled. This also works with non-latching relays, but uses more DC power. Probably not an issue with a power adapter.

[electrosonic]

My favourite tube screamer knockoff comes with a remote bypass (late 80s vintage I think)



Andrew

[R.G.]

Yep, pretty much any pedal that has electronic bypassing with a simple momentary switch can be converted to remote bypass by just remoting the switch on a jack. I'd guess that is what is in the TC.

[liquids]

hmmmmmmmmmmmmmm.....

Well, i only have a schematic for the TC sustain parametric...
Same 'ext bypass' jack, possibly the same circuitry....



Schematic is on that 'other site' if I'm not mistaken...

Shows a dual comparator chip used with discrete components as a flip flop, driving the controls of a 4007UB(? time to open CMOS Cookbook).  Bypass footswitch is SPST, EXT bypass attaches via the ungrounded flipflip side of the 'bypass' switch via the anode of anotehr diode, cathode of that diode sees the assumed SPST ext bypass mechanical switch...

Only thing I don't get is, ignoring the buffering in and out that are always present, what look to be the 4007 switches - which I assume are always in opposite states, one seems to make/break direct connect from input buffer to output buffer, but the other seems to make/break half of what appears to be an audio r-c network with some level of complexity, and yet part the network is always connected to the input of the output buffer...

Maybe there are better TC schematics from that period to look at...
But maybe this is not far from what RG was suggesting?  

I'm sort of lost, sort of on to it...but in theory/in my head, this shows that I could, via one quad comparator one a 4069 cmos switch, and two spst stomp switches (momentary?), have the ability to switch in/out up 2 circuits with a two conductor cable that connect to nothing but a remote make/break mechanical switch to ground the conductor or leave it open.  

Then, add just one standard  'audio in' jack on the box with regular cable to it,  one 'audio out' jack to the next pedal onthe big aluminum enclosure housing the circuits.

That'd get me down to 3 cables, two of which are inherent to any circuit, plus one cable between circuit enclosure and remote footswitch....?

With a 4 conductor cable, 4 switches, 2 quad comparators, 2 4069 switches, and a 4 conductor cable, I could switch 4 such circuits with just one cable and one jack between circuit enclosure and remote switch enclosure with up to 4 SPSTs?

[R.G.]

Same 'ext bypass' jack, possibly the same circuitry....
Shows a dual comparator chip used with discrete components as a flip flop, driving the controls of a 4007UB(? time to open CMOS Cookbook).  Bypass footswitch is SPST, EXT bypass attaches via the ungrounded flipflip side of the 'bypass' switch via the anode of anotehr diode, cathode of that diode sees the assumed SPST ext bypass mechanical switch...

Only thing I don't get is, ignoring the buffering in and out that are always present, what look to be the 4007 switches - which I assume are always in opposite states, one seems to make/break direct connect from input buffer to output buffer, but the other seems to make/break half of what appears to be an audio r-c network with some level of complexity, and yet part the network is always connected to the input of the output buffer...

Time to take a step back to block diagrams. A bypsss switch switch driving a pedal bypass setup always has certain elementary things it does.
(1) there is some way for the human to tell the setup "do it now!"
(2) it drives a flipflop, which toggles every time you punch the button. This is a one-bit "memory" which keeps track of whether the actual switch should be effecting or bypassing.
(3) The actual switch itself. This can be made of JFETs, MOSFETs, CMOS inverters, CMOS single switches or CMOS multiplexers.

Mother Nature lets you choose any means of doing all this. A normal DPDT stomp switch has a **mechanical** flipflop which drives a **mechanical** switch section.

You can use a mechanical "flipflop" (latching switch) to drive an electronic switch. The electronic switch can be a relay or a JFET, MOSFET... switch of some kind. If you use a momentary electrical switch, you can drive a mechnico-electric flipflop and switch; this is what a latching relay is. You can use a momentary switch to drive an electronic flipflop. This is what Boss and Ibanez do. They use NPN transistors to make the flipflop to remember state. You can use CMOS logic to make a flipflop out of inverters (i.e. 4069 or 4049 ICs) or make a flipflop out of the "spare" MOSFET transistors in the 4007, which is a kind of DIY package with CMOS inside.

You can make the switch itself out of mechanics (mechanical switch or relay) or JFETs, or discrete MOSFETs, or MOSFETs inside an IC (4007) or a pre-packaged switch like the CD4066, or a higher-integration package like the 4053 and others.

In all of this, you have to remember the three parts: (1) human actuator (2) one-bit memory (3) actual signal switch. You have to pick what you want to use, then fill in the blanks between them to get them to talk to each other reliably.  What's confusing you is that there are lots of ways to split this up, and some variations in the stuff in the filled blanks.

Get out a blank piece of paper. Draw on it three rectangles; label them 'actuator', 'flipflop', and 'signal switch'.  Now draw how many of these you need to do your switching. There will be one 'wire' between each of them indicating what controls what. There may be a power and ground going to each one if they're electronic.

Once you have all your "architectural" level stuff figured out, you have the minimum number of wires you'll need. And from there, you can fill in the spaces with how you're going to make each actuator, flipflop, and switch. Each choice restricts what you can choose to hook up to it.