Signal Polarity

[Bill Mountain]

Do any of you guys worry about polarity of the signal as it travels through your circuit?

I used to get hung up on having the output phase coherent (is that the right term???) with the input but the more I study filters the more I realize that once you start adding eq, HPF's, and LPF's the polarity will change depending on frequency, gain, eq settings, etc.

I'm working on a simple boost stage and I think I want to use an inverting opamp stage so I can easily reduce gain below unity (I have a SUPER HIGH output bass).  I'm trying to convince myself it doesn't matter but I don't see inverting stages used that often.

I understand that there might be a concern with having a low input impedence but that's a topic for another thread.

Thoughts?

[DavenPaget]

You can do something different though with a quad opamp .
First you buffer the input with a non-inverting opamp and then you go into the inverting setup you want , if you want a signal in the same phase , invert it again but this time buffered .
Takes up some space but totally worth it .

[defaced]

I'm trying to convince myself it doesn't matter but I don't see inverting stages used that often.

Most tube and transistor stages are inverting. 

[Bill Mountain]

You can do something different though with a quad opamp .
First you buffer the input with a non-inverting opamp and then you go into the inverting setup you want , if you want a signal in the same phase , invert it again but this time buffered .
Takes up some space but totally worth it .

I don't agree on the "worth it" part.  That's a lot of work to keep the polarity the same.  I guess my question is, is there really a reason that I need to keep it the same?

[Bill Mountain]

I'm trying to convince myself it doesn't matter but I don't see inverting stages used that often.

Most tube and transistor stages are inverting. 

Yes, but there are usually an even number of stages.  Especially with tubes.

I was referring to inverting opamp stages.

[Mark Hammer]

Signal polarity only really begins to matter when two or more signals are combined.  So, if you decided to run your guitar to a delay with a wet and dry output, fed the dry to a distortion, and combined the distorted dry and clean delay at the amp, you would have problems at the amp if one of those was opposite in phase to the other.

In theory, that,s not really any different that running your amp to multiple speakers.  If they are all wired in phase with each other, it doesn't matter if they push or pull back in response to a positive-going voltage; what matters is that they're all doing the same thing together.

[Bill Mountain]

Signal polarity only really begins to matter when two or more signals are combined.  So, if you decided to run your guitar to a delay with a wet and dry output, fed the dry to a distortion, and combined the distorted dry and clean delay at the amp, you would have problems at the amp if one of those was opposite in phase to the other.

In theory, that,s not really any different that running your amp to multiple speakers.  If they are all wired in phase with each other, it doesn't matter if they push or pull back in response to a positive-going voltage; what matters is that they're all doing the same thing together.

I figured this would get mentioned.  You're right this is a good reason to worry about polarity but won't this only really help if there is no filtering at all?  Any filtering (which any pedal will have) will effect polarity some what.  This leads me to think that anytime you mix pedals (whether it claims to match phase or not) it's a crap shoot.  Or am I over thinking this?

[slacker]

Deleted I'd just said the same thing as Mark.

[PRR]

> an inverting opamp stage ...might be a concern with having a low input impedence

And noise-figure. (Tho your bass may be hot, elsewhere I found a lo-output bass and a reasonable preamp seems to hiss.)

> once you start adding eq, HPF's, and LPF's the polarity will change depending

The phase will change. Ultimately 90 degrees for a 6db/oct filter. 90 deg is not an inversion. 12db/oct and steeper will invert but you don't take steep boosts and a steep cut will be so far down in amplitude that you don't care about phase.

Keep it non-inverting just because some day you (or client) will do something strange and should not be surprised by polarity.

[bacanador]

I figured this would get mentioned.  You're right this is a good reason to worry about polarity but won't this only really help if there is no filtering at all?  Any filtering (which any pedal will have) will effect polarity some what.  This leads me to think that anytime you mix pedals (whether it clames to match phase or not) it's a crap shoot.  Or am I over thinking this?

The polarity does not change if you have a non inverting filter, there is phase shifting for certain frequencies (below or above the cutoff frequency), the frequencies of interest will be in phase with the original.

[Bill Mountain]

Thanks for the replies.  I wasn't thinking about noise and for some reason I had the idea that phase shifts would be noticeable when mixing signals.  PRR had explained this before when I was working on a mixable 2 channel bass preamp but I let it slip my mind.

Thanks again!

[defaced]

I'm trying to convince myself it doesn't matter but I don't see inverting stages used that often.

Most tube and transistor stages are inverting. 

Yes, but there are usually an even number of stages.  Especially with tubes.

I was referring to inverting opamp stages.

Opamp, tube, or transistor, it makes no difference, your question is applicable to any device that inverts phase.  The point I was driving at is that while you may not see many opamp inverting stages, inverting stages as a whole are used, so your original assertion (the one I quoted) doesn't make much sense.  You are correct, alot of devices do have an even number of inverting stages, but I think you'll also find there's just as many that don't, and yet, some that are both depending on what "mode" they're in.  For example, take an amp that has an effects loop.  Often the send will be a simple follower stage, and the return will be an inverting amplifying stage.  So simply turning on the effects loop on your amp inverts the phase. 

[Mark Hammer]

There is an axiom in organizational surveys that you never ask employees about anything you're not prepared to address or fix.  So, you never ask them if they're happy with their benefits/pay if you aren't prepared to improve those benefits/pay in the face of a negative response.

Does "group delay" (phase shift within frequency bands) exist?  You bet your bippy it does.  Can you DO anything about it in the context of a pedalboard?  Not really.  I mean there are things like the BBE units that attempt to correct for misalignment of fundamentals and harmonic content arising from group delay and misalignment of tweeters and woofers,  but by the time your signal has passed through a gauntlet of modifiers, lord knows what parts of the spectrum are lagging by how much.  So if you're not prepared to be able to do anything about it, don't ask about it.  Just treat it as the sound of rock and roll.

Cancellation, though, as a result of wholesale signal inversion, is a whole other ball of wax and somethng you CAN track and do somethng about.

[amptramp]

There are pedals such as the Orange Squeezer that modify the signal based on one polarity, so if the input goes positive first or negative first does matter.  I prefer to retain non-inverting polarity between input and output of any given pedal.

[merlinb]

There are pedals such as the Orange Squeezer that modify the signal based on one polarity, so if the input goes positive first or negative first does matter.  I prefer to retain non-inverting polarity between input and output of any given pedal.

That really isn't a problem, the signal envelope is symmetrical on average, so even if the phase is 'wrong' you're only going to miss the first half-cycle, which is too quick to produce any audible effect.

I couldn't give a monkeys about absolute phase. If I am ever going to mix signals in parallel, I'll just correct the phase there and then; only one stage required, instead of many more stages needed in all the other devices to force them to be absolute phase!

[Mark Hammer]

I will only add to what Merlin noted that attention to phase is a triviality when you're the one making the pedals, or at least have a schematic.  When the pedal is a commercial one of unknown design properties, I think folks have a right to want to know what they can expect in the way of phase.

Which naturally raises the question, do manufacturers routinely state when a pedal has an output inverted elative to the input?