An "order switcher."

[asmith]

I was reading up on Beavis Audio's ambitious 'FuzzLab' project, (beware the link - your bandwidth is likely to be eaten by a large image) and noticed that the builder decided to go down the route of using patch cables to play around with the order of his fuzz modules.

I thought about using a piece of kit like this, and decided to try and find a tone, I'd have to play something, unhook the patch cables, think about which inputs to put the jacks in, put them in, turn up the gui... wait, I've forgotten what the tone sounds like.

Pah. Not nutz enough. So I came up with this.



Enough connections for ya?

That's (n) DPDTs to determine whether the modules are being bypassed, and (n) DP(n)T rotary switches, to determine where in the chain the modules are being placed. If two modules are on the same setting, they're in parallel with each other. In this example I've used n=4.

Is that right? Does this work? The diagram is merely the product of an exciting dream.

Also, I'd prefer to have those "In"s and "Out"s of the modules (grey boxes) be quarter-inch jack inputs, so that the switcher could be housed in a box, as an external circuit, and have a multitude of effects jacked in at leisure. But of course jacks have their own "In"s and "Out"s. How could the diagram be modified so that that idea could actually be schemed into the design?

[ashcatlt]

Looks like it works. The DPDTs switch on and off whichever pedal is wired in that particular position in the order. These must be in bypass mode unless something is actually assigned to that bus. For example, if you had something assigned to #1, 2 and 3, but not 4, you'd get silence at the output unless #4 is bypassed. Same for all the other positions, it'll break the chain. That might cause confusion, but I don't see any easy way around it.

There's no reason you can't use jacks. You could use NC switched jacks for the Sends (going to the input), with the tip's switched lug connected to the tip of the associated Return (coming from the output). Then, though, if it's set to share a position with some other pedal, it'll act as a short around that other pedal, and you probably won't get much out of there. Probably better to leave them unswitched and assign the unused modules to some bypassed position.

Oh, but then we've got the issue of mixing the outputs of these pedals when two or more are switched to the same position. The output impedance of these things could vary rather wildly, are not likely to match up very well, but are all likely to be fairly low. You're going to have one loading the other. There will likely be tone and/or volume loss, interactivity of the volume or level controls, and all kinds of other strange interactions. We usually want to put at least a resistor in series with each of the outputs on the way to the summing node. That would mean right there at the Return jack on the way to the rotary. Of course then you're adding unnecessary resistance in the series connections... I'd think that's less likely to cause problems with most buffered pedals, but maybe sg or JohnH can advise better on that.

[asmith]

Mar 22, 2011 22:56:02 GMT -5 ashcatlt said:

Oh, but then we've got the issue of mixing the outputs of these pedals when two or more are switched to the same position. The output impedance of these things could vary rather wildly, are not likely to match up very well, but are all likely to be fairly low. etc.

Oh, poo.

[newey]

Oh, poo.

Don't be so easily dismissive of this. The drawing alone gets a +1 from this corner! How did you do the drawing, it looks much better than anything I could ever do , , ,

And it is a nutsy solution to the problem of being able to easily do an A-B comparison. But it would be a whole lot easier IMO to just record the different set-ups you wanted to check out, then do your A-B at your leisure with playback.

[jcgss77]

Would putting a buffer at the end of each module help with matching up the impedance?

[ashcatlt]

Not really. It will at least let you know for sure what the out-Z of each module is, but they'll still be all the same. You'll be "violating" the (at least) 10:1 Rule (of thumb), and they'll still load each other down. I think the series resistor (47K or 68K, or something) will be the best thing for it.

Where we might want a buffer is at the very front of the chain, at the main input. Assuming that this box is the first thing the guitar sees, if position #1 ends up with one module in parallel with one or more others it will start to divide down the in-Z which the pickups see. In the case of this four banger, if each pedal had 1M input, and all 4 were in parallel, the guitar would see 125K = total tone suck! In the interest of tonal consistency, a buffer at the input is probably best.

I guess it's also worth noting there that some pedals (FuzzFace and similar, some old wahs) kind of depend on loading the pickups with a low-Z input to get some "free" pre-EQ. Putting one of these things in any position but first (before we add the buffer) will change the way they work and give possibly undesirable results. With the buffer, it'll be weird even in position 1. These pedals are almost always best when put as the first active stage in the chain. Probably best to just plug that thing in before this switcher box and use this for your other stuff.

[jcgss77]

How about this?

www.beavisaudio.com/techpages/PedalHacker/OrderSwitcher.htm

I referred this same circuit to sbgodofmetal in his thread in effects devices, but it seems to be useful here too.

It seems to me to only help with 2 effects, but does this help with the loading issues and the order of things? Perhaps 2 of these for 4 fuzz boxes? Of course, there would be some rearranging until the proper sound was achieved.

[ashcatlt]

Those are all series connections. You'd still have the issue i mentioned with FuzzFace et al, but that's true anytime you're using that type of pedal and is just something you have to learn to work around. The problem here is in the parallel connections.

I'm still hoping that one of our EEs will jump in here with some more expert advice, but I do think the resistors will be the best answer.

[asmith]

Mar 23, 2011 10:30:27 GMT -5 newey said:

How did you do the drawing[?]

Corel Photo-Paint.

J. C. G. S. S., class of '77:
I'd seen and like the idea of the Beavis Switcher, but didn't think it was enough! I like the idea of choosing a place for your effect in a chain with a rotary selector. I like the idea of the feeling of dialling in your specific order.

I can see now why effects in parallel would be a pain. This has been a quick learning experience. I'd be completely willing to forego the entire idea of having two effects in parallel. In fact when it boils down to it, I'd be willing to forego the entire idea. Of course, this is like saying "Just forget about it and don't bother," and that doesn't seem to go down well on this board at all. Except for:

Mar 23, 2011 10:30:27 GMT -5 newey said:

But it would be a whole lot easier IMO to just record the different set-ups you wanted to check out, then do your A-B at your leisure with playback.

Captain Sensible, eh? You have the right idea mate.

[sumgai]

Mar 23, 2011 10:30:27 GMT -5 newey said:

t is a nutsy solution to the problem of being able to easily do an A-B comparison. But it would be a whole lot easier IMO to just record the different set-ups you wanted to check out, then do your A-B at your leisure with playback.

This from the guy who gave us:

The First Law of GNutz2: "Leave no lug unsoldered."©


(copyright The Beta Particle Bombarder, 2010)

Turn in your merit badge, now!

There'd be no chance of knowing that the recordings were the same in all respects, if you made several passes into the recorder, each with the desired pedal-order.

But if one were insistent on this method, then one need record only one pass, with a bare unalderated guitar into the recording device in question. Then it becomes nearly child's play to feed that playback through the various pedal-orders, at will, and instantaneously. Or as Chris would have put it, one could simultaneously observe several instantiations......

~!~!~!~!~

More.........

Sorry ash, but the exact purpose of a buffer is to electrically isolate one component from another, or one sub-circuit from another. The intent is to do exactly as the passive resistors you called for - keep the output of one device from loading the output of another device, to the detriment of both. But I do agree with your suggestion about a buffer at the input, for all the reasons you mentioned, that's a good one for Ace to think about.

Of course the whole of the device is now bigger and a bit more bulky, plus it requires a battery (or a phantom supply). But it's still a viable idea. I can see where it might be useful on stage, for some folks anyways. But in the practice room.... no doubt about it, this would please a big portion of the Tone Junkies out there, I'm sure. ;D

HTH



sumgai

[ashcatlt]

Okay, so we stick 4 buffers in there at the returns. Now all the loops are returning to a known, predictable impedance. The pedals themselves won't be loaded. But what happens when we start combining the low-Z outputs from the buffers in parallel with one another? We're still gonna want mixing resistors. And then we just might want another buffer at the final output, to keep the overall out-Z consistent and predictable. That comes to 6 buffer stages altogether if we're talking about 4 loops.

This is actually probably the best answer, honestly. It would allow the use of those switched jacks I was talking about so you could leave one loop empty and have the option of bringing the clean sound in parallel with the effected signal.

BTW - Corel pwns! I usually use Draw for this kind of thing, then convert and manipulate in PhotoPaint.

[newey]

Turn in your merit badge, now!

Mea Culpa. I certainly didn't mean to dampen any Nutzy speculation here.

"I'm not sayin', you know, just sayin'". Or sumpin'.

[sumgai]

ash,

Technically speaking, you wouldn't need those "extra" mixing resistors - isolation is isolation. However, in some respects, you always see them anyways - they're usually disguised as the channel's output level controls!

A few years ago, someone here blathered about some new IC's that were a vast improvement over the more-or-less standard TL-07x and Tl-08x series. I don't recall now who it was (perhaps Channelman?), but the net effect was/is, all six buffers can be done in two chips that draw extremely low power at idle, have extremely high bandwidth, and are no more expensive than the average bear.

I suspect that if one were to look closely, one could figure out a way to install a pair of quad-amp chips into a very small space. A printed circuit board wouldn't even be necessary, although it would make things easier to assemble, but a chunk of Vero board would work just as well, to my way of thinking anyways.

When that's all done, I'm gonna seriously consider awarding a Karma point here......


HTH




sumgai