"monitoring" guitar signal

[Thomeeque]

 Hi!

I'm not sure, if "monitoring" is the best term for what I mean, but maybe picture and some explanation will make it clear:



So, what I want basically is to get dry guitar signal mainly for always-on tuner and for envelope-detectors of envelope-controlled modulation effects (and maybe for some other use) at MONITOR OUTPUT with as low influence on guitar signal itself at NORMAL OUTPUT as possible - from the point-of-view of device connected to the NORMAL OUTPUT it should behave like if it is connected to the guitar directly (there should be really just wire connecting INPUT and NORMAL OUTPUT).

Is there a way? Probably something with very high input impedance. Did somebody experimented this way already (I assume that I'm not the first one with this idea)?

Thanks, Tomas

Note: It would be part of something bigger, so influence of INPUT and OUTPUT jacks can be ignored.

[Thomeeque]

sorry, wrong button (again)

[jkokura]

http://www.diystompboxes.com/smfforum/index.php?topic=36392.msg697630#msg697630

There's a perf layout for a buffered splitter I use to do the same thing you want. There's a dry out for tuner, and a main out for my pedal board/amp send. I mounted it into a volume pedal, but you don't have to - you can easily use whatever enclosure you can mount three jacks and a power jack into.

Jacob

[Processaurus]

Any FET input buffer (GGG jfet buffer for instance, or a FET opamp buffer (like using the popular TL072)) will have little loading effect on the guitar signal.  However, be aware that some vintage pedals, like simple fuzzes, can do strange things on the input signal when they're connected.  I've had it happen, splitting the guitar before a fuzz, where the clean branch sounds screwed up and muddy and sort of distorted because of what the input of the fuzz is doing interacting with the guitar.

[Thomeeque]

Any FET input buffer (GGG jfet buffer for instance, or a FET opamp buffer (like using the popular TL072)) will have little loading effect on the guitar signal.  However, be aware that some vintage pedals, like simple fuzzes, can do strange things on the input signal when they're connected.  I've had it happen, splitting the guitar before a fuzz, where the clean branch sounds screwed up and muddy and sort of distorted because of what the input of the fuzz is doing interacting with the guitar.

Yes, I guess that's the price I have to pay - natural interaction at NORMAL OUTPUT has definitely higher priority for me. Would you say, that there still remained enough information to derive usable envelope (for let's say auto-wah) at clean branch?

So there was no audible difference at fuzz output with and without this follower-buffer? Which one exactly you have tried?

Thanks!

http://www.diystompboxes.com/smfforum/index.php?topic=36392.msg697630#msg697630

There's a perf layout for a buffered splitter I use to do the same thing you want. There's a dry out for tuner, and a main out for my pedal board/amp send. I mounted it into a volume pedal, but you don't have to - you can easily use whatever enclosure you can mount three jacks and a power jack into.

Jacob

Thanks Jacob! This solution however I cannot use, as I don't want to put any buffer between the guitar and the effects like fuzzes and overdrives (and if I would put it after these effects, I would loose most of the original envelope information).

T.

[GibsonGM]

AMZ has a very nice splitter (and A/B/Y box) that would work great for this application!  Click AMZ link at top.....

[Thomeeque]

AMZ has a very nice splitter (and A/B/Y box) that would work great for this application!  Click AMZ link at top.....

I have problems to find it thru the AMZ link at top - do you mean this one: http://www.muzique.com/lab/splitter.htm (I did get there using google)? Thanks, T.

[jkokura]

If you don't want a buffer, than your option is to simply split the signal. Take two leads off of the input jack tip and then connect them to the two output jack tips, one to each.

However, if you don't buffer this you may run into some small tone loss (arguable), but as you say you don't want a buffer... I don't think you have another solution. Either you buffer split it, or your just split it. Are there any other options? Not that I know of...

Jacob

[Ben N]

Seems to me that you should be ok with just a very high impedance buffer (1M-10M) on the "monitor" signal (or side-chain); that would have an insignificant impact in terms of additional loading, almost "invisible.". Your regular signal path would be essentially unchanged (and unbuffered), and the side-chain would be isolated from the signal path. 

[ashcat_lt]

Seems to me that you should be ok with just a very high impedance buffer (1M-10M) on the "monitor" signal (or side-chain); that would have an insignificant impact in terms of additional loading, almost "invisible.". Your regular signal path would be essentially unchanged (and unbuffered), and the side-chain would be isolated from the signal path. 

This was my first thought, too.  A very large impedance (I'd go for 10M, to obey the 10:1 rule with ubiquitous 1M devices) in parallel will have very little impact on the total Z seen by the pickups.  The problem, though, is stated above.  The pickup doesn't know or care about the Z of either of the individual lines.  It "sees" one total Z made of the parallel combination of the two.  This value is what will determine the cutoff frequency of the low-pass filter created with the inductance of the pickup and the capacitance of the cable.  That is, if you plug in a FuzzFace or something similar which purposely rolls off the high-end on the way in, you'll have a significant loss of high freqs at the input of the "monitor" side as well.

Seems to me, though, that something like a tuner or envelope detection circuit must likely be low-passing the signal before it does much else anyway.  You might just be helping it do its job better.  You could test it:  Roll down your Tone knob and see if it has any adverse effect on the action of the circuit.

[Ben N]

Right, I doubt that high-freq content matters much to whatever the sidechain may be feeding, be it a tuner, an envelope follower, noise gate, whatever. As you suggest, if anything, eliminating hash would make those things work better.

[R.G.]

Any buffer with Zin>1M  will be fine. Realistically, an opamp or JFET is the simplest choice, and you'll have fewer problems with side effects if you use the opamp.

[GibsonGM]

Yes, I was talking about this one:  http://www.muzique.com/lab/splitter.htm

I've built the box, and when I use it to cycle thru 2 amps, I get no problems.  AND (for your use), when I tap off the "Out 3" output, I hear ZERO difference in my signal.  So you could just build one with 2 FETs - 1 as a 'passing thru buffer', and the 2nd as the splitter.  Should work awesomely.