is one buffer enough to split to two outputs?

[ode2no1]

if i were to build a simple single opamp buffer, could i run a guitar thru it and take the output and split it to two different pedals, then thru a summing amplifier to run in parallel? of course i know i CAN do this, but would the single buffer be enough to split the signal sufficiently? or would running a dual buffer to split the signal be significantly better? i feel like the initial buffer would be enough to split the signal, but if i'm wrong i'd like to gain the knowledge as to why.

[R.G.]

It'll be fine.

Your intuituion is backed up by the fact that the output of most single opamps can drive loads down to about 2K just fine, and that most pedals have input impedances of 1M or so.

The only thing more "splitting" does is to keep a fault in one load from affecting the performance to the other one. You could put a 1K to 10K in series with each output to prevent that as well if you wanted.

It'll be fine.

[ode2no1]

thanks R.G. i always appreciate when you chime in. i have actually wondered about the resistor in series thing though. it seems to me that doing so would rob you of gain. so if i added a series resistor to each output would i be hitting each pedal run in parallel with less signal?

[PRR]

> the resistor in series thing though. it seems to me that doing so would rob you of gain.

Do math.

R.G. asserts that "most pedals have input impedances of 1M or so". Let's assume "or so" may cover 500K inputs.

He suggests 2K to 10K series. Take 10K.

You have a Voltage Divider with 10K and 500K. The output is 500K/(10K+500K) or 10K/510K or 0.98 of the input signal.

I assert that you do not play with 2% accuracy of amplitude.

Also that you can not hear 0.98 gain change.

0.98 is -0.17dB. Generally we assert that less than 1dB is no-difference. Skilled listeners can hear smaller changes. 0.25dB however is vanishingly small.

Note that simple buffers (one-transistor followers) may already have a gain a hair under unity, and we usually do not care.

Opamp buffers strapped for unity will have a gain like 0.999, or so near unity that nobody can prove the difference.

Opamp buffers can be rigged for a gain of 1.02. This plus the the 0.98 loss in an output resistor and typical load results in dead-unity gain.

Remember that your amplifier has a gain/volume knob. This typically dials overall gain from zero to 1,000. Unless you do only dimed-out headbanger playing, you probably run the gain around 250 or 500. If "stuck" with a loss of 0.98, you will turn-up a micro-hair to give amplifier gain of 255 or 510.

[pinkjimiphoton]

related question kinda,
couldn't you use a transistor like a tube, and take output off of collector and emitter at the same time? or would they be out of phase with each other?
sorry for polluting your thread, just wondering this while reading it. i still have decades of millenia i need to learn

[GibsonGM]

Out of phase jimi, yup.
 
The "top", common emitter output would be out of phase with the input by 180 degrees.
The "bottom" emitter follower, or buffer, would be in phase by definition, as it 'follows' the input.   Good stuff.   They used to actually USE this property to make a form of phase inverter, back in the good old days!!   Same deal as with tubes, plate vs. cathode  (paraphase amplifier?)

The CE config. is a gain stage, and would likely do the job fine if you compensated for that.

The follower would, OTOH, be the ideal driver for something like a splitter....hi input Z, low output Z, and able to source lots and lots of current. Gain a bit less than unity unless you play with it to bring it back up (why bother?).

The OP may well wish to make a follower and use THAT to send the output to 2 other inputs, and save an opamp.   Kinda the same thing...small difference.

[pinkjimiphoton]

thanks for the explanation, mike.
that makes sense. i wonder if you could use it somehow to "phase out" or notch noise or certain frequencies.
it's basically how a phase shifter works? sweeping between the in and out of phase sounds, or would that just modulate the volume?
seems like it could be exploited for my idea of sweeping fuzz filters maybe.
sorry for the hijack again.
fascinating topic.
hoping to get my workshop together soon, i've been doing a lot of repairs on amps and guitars, still have the flying spaghetti monster on my breadboards
all these months later... but... getting closer..

[GibsonGM]

Good idea, Jimi, and ya, there are some phasers out there like this!  But 'set' BJT stages like this seem to be 'locked' into what they are...0 degrees, 180 degrees.  You COULD blend them together, and you will get more or less cancellation....I suspect it would sound a lot like a strat wired "out of phase", tho, and not that much like what we call a phaser.

I've seen at least one where they use an LDR to allow 'more or less' of the emitter signal to blend with the collector, and it seems to work.  I believe a member named "Mac" posted one on here years back...search "bjt phaser"...

More commonly, FETs have been 'the way' to alter phase in a ton of things, so that's what my instinct tells me to look at for phasing.  But hey - this is a READY MADE FOR MOD thing, man!  Get to it!   

[ode2no1]

how funny that the issue of phase came up in this thread cause i was just about to post a question about it. say you took something like the splitter blend:



instead of using a switch to flip the phase back and forth, how easy would it be to have it be variable instead?

[R.G.]

Easy.

[Eddododo]

how funny that the issue of phase came up in this thread cause i was just about to post a question about it. say you took something like the splitter blend:



instead of using a switch to flip the phase back and forth, how easy would it be to have it be variable instead?

if you mean varying phase, then youd have a few more steps. like a prettyverymuchly different circuit.
if you mean varying the amount of cancellation... maybe.

Im not an expert on anything, but everytime I've tried to 'blend' purely out-of-phase signals, be it with FETS source vs drain or op amp stuff, I always find there is little-to-no effect until there is [suddenly] total or near-total cancellation.

For in-betweens you'll likely want a different approach than this polarity-reversing setup.

Im not sure what youre looking for, but if its a phaser, youre probably barking up the wrong tree. If youre looking for variable-phase blending, i have found the following to be true- in the case of phase-aligning a mic+pickup combo, a subwoofer-style variable-phase circuit is often 'good enough' for my devices (i DO usually low pass one of them, so couldnt say much about higher phase wierdness)



edit: RG has spoken:p

[ode2no1]

i was messing with this earlier, and knowing that i probably would not succeed, i replaced the switch with a 1M pot which had one leg tied to ground...hoping i could basically go from open to closed thru the sweep. using only my ears, and not an oscilloscope, it seemed i could get the out of phase sound just fine on one end of the pot, but sweeping back the other way i could get most of the way back to the in phase sound, but it wouldn't quite get there. it was lacking a bit of volume and low end. i then removed the ground connection and got the same results except when i got to the very end of the "back to in phase" side of the sweep i would hit a spot in the pot's rotation where i then got a decent volume boost, but it sounded in phase. clearly i don't know all the theory behind these things, but i'm trying to learn if you guys would be so kind as to enlighten me. i think it would be great to be able to use a pot to go from 0-180 degrees in/out of phase.

[CooperL]

Is there a way to vary the phase without storing the signal somewhere and delaying it?  RC networks have a phase response, but the phase change is different at every frequency.  You could of course sample and store the signal then delay play back slight to adjust the phase but perhaps that approach is better suited for the DSP thread..

[deafbutpicky]

An allpass with variable R should do that.

[CooperL]

An allpass with variable R should do that.

An allpass  would only work for one particular frequency. What about the entire signal?

[ode2no1]

i'm definitely not looking for a phaser type of sound. i would just like to be able to vary the phase 180 degrees with a pot if possible.

[pinkjimiphoton]

the problem is i believe in the nature of the phase shift... you want a fairly narrow band of frequencies, and then shift that against itself to creat a notch that moves around i'd think. i'm still just in the dawning of understanding iotas of this stuff... cuz like you said bro, if ya go !100% in/out of phase you get a cancellation of the audio itself.
you can definitely hear the phasing happen... it sounds almost like counterpanning two pan controls while listening to a stereo track. but it's not deep enough for that quasi leslie, make your drunken audience seasick thang.

pure conjecture, i was just thinking maybe if ya take a deal like this, like the old phase inverters in a tube amp in the driver stage, but used some kinda ramped lfo to sweep between 'em using photoresistors and led's or something. but i'd assume that would just end up being tremolo, not phasing.

maybe make a "contour" kinda control, heck, a big muff pi tone stack even, and sweep that against a buffered dry signal?

way above my paygrade still   

[GibsonGM]

pure conjecture, i was just thinking maybe if ya take a deal like this, like the old phase inverters in a tube amp in the driver stage, but used some kinda ramped lfo to sweep between 'em using photoresistors and led's or something. but i'd assume that would just end up being tremolo, not phasing.

maybe make a "contour" kinda control, heck, a big muff pi tone stack even, and sweep that against a buffered dry signal?

way above my paygrade still   

Well, I think his forum name is/was "mac", already did this, man! He came up with a BJT phaser that used the collector AND emitter signals, which are out of phase, and blended them using LDR's and an LFO.  I have the schematic somewhere, not sure if I should post it tho.
I haven't seen him around in a long time, and can't find the post again.

Never built it, but it looks VERY cool!   From the looks of it, it probably sounds good, too.    Exactly the idea you're talking about...

[ode2no1]

with a clean signal i was for sure getting pretty near total cancellation, but running two dirt pedals in parallel and reversing the phase on one end i still had plenty of volume coming thru the amp (obviously only due to the amount of gain from the pedals)...like pretty much the same volume as them being in phase, only bright and nasty sounding. thought it might be cool to be able to dial in different tones by being able to vary the phase on one side assuming both sides are running relatively high gain dirt pedals. i guess full bandwidth variable phase adjustment isn't very commonplace huh? i don't want the phase adjustment to be like a phaser, because i don't want to alter the tone of that side of the circuit itself, only the way it interacts with the other, static phase, side.

[pinkjimiphoton]

pure conjecture, i was just thinking maybe if ya take a deal like this, like the old phase inverters in a tube amp in the driver stage, but used some kinda ramped lfo to sweep between 'em using photoresistors and led's or something. but i'd assume that would just end up being tremolo, not phasing.

maybe make a "contour" kinda control, heck, a big muff pi tone stack even, and sweep that against a buffered dry signal?

way above my paygrade still   

Well, I think his forum name is/was "mac", already did this, man! He came up with a BJT phaser that used the collector AND emitter signals, which are out of phase, and blended them using LDR's and an LFO.  I have the schematic somewhere, not sure if I should post it tho.
I haven't seen him around in a long time, and can't find the post again.

Never built it, but it looks VERY cool!   From the looks of it, it probably sounds good, too.    Exactly the idea you're talking about...

wow, figures mac would have done it already.... i love his ideas, too.. what scares me is i'm seeming to understand this stuff...lol
i'll send mac a pm and ask about it... thanks mike