Need a smple guitar signal splitter

[ZillaG]

I'm new here. I stumbled across this site looking for a signal splitter.

I want to build a stomp box that does this


Guitar ----->  signal splitter ---> signal A ---------------------------> |
                                                signal B -----> another circuit ---> | ---> output

signal A and signal B will be identical within tolerances. Essentially, I want to process signal B further, then sum it back with the original signal. Can you point me in the right direction please?

[GGBB]

Welcome to the forum. Search for Run Off Groove Splitter-Blend.

[antonis]

Welcome also..

Just beware of possible signal B phase reversal via "another circuit"..
(in case of other than that proposed by Gord..)

[ZillaG]

Thanks.

[Ben N]

Just beware of possible signal B phase reversal via "another circuit"..
(in case of other than that proposed by Gord..)

The Splitter Blend has a phase switch for just such an occasion, but if you are only switching between loops, it may be more complex than you need. There's another recent thread with a similar switching question that you may find enlightening.

And howdy!

[ZillaG]

Ok I downloaded LTSpice and simulated the splitter blend circuit, just the part that actually does the split. I put a 100mV/440Hz signal in. Why am I only getting a ~60nVpp signal on the outputs?

[GGBB]

Ok I downloaded LTSpice and simulated the splitter bend circuit, just the part that actually does the split. Why am I only getting a ~30nVpp signal on the outputs?

Why do you need to sim it? (What question are you looking to answer?) And according to the graph you are actually getting close to 60nVpp.

It's not the circuit - probably thousands have built it.

EDIT - in my sim software (I don't know what you are using) "1m" resistance is one milliohm. For 1 megaohm you need to enter "1M". Case matters with metric prefixes.

[ZillaG]

I sim'ed it as a learning experience (I don't have the parts). I used LTSpice (on a Mac). I put "1M" on the output R values and it kept setting it to "1m". Let me look into that.

[ZillaG]

Ok I downloaded LTSpice and simulated the splitter bend circuit, just the part that actually does the split. Why am I only getting a ~30nVpp signal on the outputs?

Why do you need to sim it? (What question are you looking to answer?) And according to the graph you are actually getting close to 60nVpp.

It's not the circuit - probably thousands have built it.

EDIT - in my sim software (I don't know what you are using) "1m" resistance is one milliohm. For 1 megaohm you need to enter "1M". Case matters with metric prefixes.

Yep that was it. I had to put "1Meg" for the R values, and now get the expected results. Thanks!

Also in addition to using LTSpice to learn, i'm also simulating the "another circuit" I have per my post. So the split blender circuit feeds into that.

[PRR]

....For 1 megaohm you need to enter "1M". Case matters with metric prefixes.

Actually: SPICE goes back to days when not all terminals had upper/lower case. (SPICE core is very much a punch-card mindset.) (And original-flavor FORTRAN is normally coded ALL UPPER CASE; C-spice is much later.) The ambiguity milli/Mega was resolved by requiring "Meg" (meg, MEG). While the ADM3a terminal (with lower-case ROM) was available about the same time, the Meg business stayed written in stone essentially forever.

When unsure and in a hurry, "1000k" is usually equivalent to the Meg.

As ZillaG says, some software is snarky enough to think it can correct you, so you have to keep a close eye on it.

[GGBB]

....For 1 megaohm you need to enter "1M". Case matters with metric prefixes.

Actually: SPICE goes back to days when not all terminals had upper/lower case.
...
The ambiguity milli/Mega was resolved by requiring "Meg" (meg, MEG). While the ADM3a terminal (with lower-case ROM) was available about the same time, the Meg business stayed written in stone essentially forever.

Interesting. The interfaces we have today in front of SPICE obviously vary. As I mentioned, in the software that I use only "M" is necessary - this is TINA-TI - but it gets displayed as "MEG". And "1000k" does get converted.

[antonis]

As I mentioned, in the software that I use only "M" is necessary - this is TINA-TI - but it gets displayed as "MEG". And "1000k" does get converted.

Not to my TINA-TI..
(in case you type m, it will stand for 0,001 Ohm..)

[GGBB]

As I mentioned, in the software that I use only "M" is necessary - this is TINA-TI - but it gets displayed as "MEG". And "1000k" does get converted.

Not to my TINA-TI..
(in case you type m, it will stand for 0,001 Ohm..)

Yes - I mentioned that earlier.

[ZillaG]

....For 1 megaohm you need to enter "1M". Case matters with metric prefixes.

Actually: SPICE goes back to days when not all terminals had upper/lower case. (SPICE core is very much a punch-card mindset.) (And original-flavor FORTRAN is normally coded ALL UPPER CASE; C-spice is much later.) The ambiguity milli/Mega was resolved by requiring "Meg" (meg, MEG). While the ADM3a terminal (with lower-case ROM) was available about the same time, the Meg business stayed written in stone essentially forever.

When unsure and in a hurry, "1000k" is usually equivalent to the Meg.

As ZillaG says, some software is snarky enough to think it can correct you, so you have to keep a close eye on it.

I'm actually an electrical engineer by trade when I started my pro career in the rat race, but have been doing software development for so long that I've forgotten the theory. I did use SPICE in those days. I'm excited to rejuvenate my interests in circuit design again.