Inverting op amp out of phase?

[RobertE]

If I use an op amp in an inverting config which moves the signal 180 degrees out of phase, does this cause a problem if both channels are inverted this way? I mean, should/must I use another inverting stage to correct this? If I don't what happens to the audio quality ?

I don't have the second stage right now and it seems fine....

 

[wakibaki]

In general, no. It has no effect on the quality. There is no way of knowing whether the signal has been inverted already, anyway.

w

 

[BobEllis]

Some people believe absolute phase is important. Sometimes I think I can hear phase on percussion, but most of the time not.

Try flipping the speaker connection so that the plus terminals are connected to the amp negative and speaker negative connected to amp hot and see if you can tell a difference.

 

[ChristianThomas]

Some people swear by absolute phase and I know that there are some amplifier manufacturers out there who simply wouldn't contemplate an inverting amp. I haven't tried to tell the difference for years but I would always try to maintain absolute phase.

As for simply inverting the speakers, this may not do precisely what is desired. The layout of some crossovers - for instance a star earth - does often expect to have a ground on the normal terminal. I'm not even sure they always end up making the same circuit if the expected ground is floating around. (it's a bit hard to think through properly on a Saturday - one would expect them to be the same but I'm not sure I like the look of it that way round, esp when it comes to the amp's feedback loop).

I love inverting inputs, but I would be inclined to conform in order to keep the number of variables down. Incidentally you don't always need an extra inverting stage. If you have a differential input somewhere you can always flip the wires over.

 

[analog_sa]

Some people swear by absolute phase

It seems absolute phase audibility has been confirmed even by double blind tests. This really surprises me as i find other effects like those due to passive components or wire to be far more audible.

While i certainly hear a difference when the speaker connections are reversed i am not at all convinced this is indeed due to absolute phase and not to the directionality of speaker cables.

To further complicate things it is not that simple to create a transparent 180 deg inversion unless it is a balanced sytem. Quite often the digitally implemented inversion is faulty and the analogue one involves an extra stage.

To make it even worse it appears recordings are not particularly consistent in absolute phase polarity and unless one can remotely and transparently change system phase, there is just no point having a correctly phased system.

 

[RobertE]

I have a pretty critical ear, but this is a car system, so lots of other noise issues. The bass seems really tight - more so than with the stock radio, but it may be that I am filtering out more of the low signal so the bass speakers don't work as hard - now, there doesn't seem to be as much bass either, but the overall tone sounds better. Hmmm. Using the op amp in an inverting stage mode was required originally due to the single supply. Now that I have dual supply it seems just as easy to run it in a non-inverting mode, but I wasn't going to stress about it.

I agree that unless every other stage from recorded track to output maintains absolute phase it's hard to know what the "real" phase is...

Very interesting.

 

[wakibaki]

I agree that unless every other stage from recorded track to output maintains absolute phase it's hard to know what the "real" phase is...

I did anticipate that there might be some disagreement about this, but IMO it's just not worth stressing over.

The DB tests that reveal the audibility of absolute phase have, to my knowledge, been conducted with synthesized asymmetrical waveforms of an unusual character.

Most 2- or 3-way speaker systems have an inversion at the crossover to the tweeter, this is intended to compensate for the difference in arrival times between drivers.

You pays your money and you takes your choice.

w

 

[CanisLupus]

Here's my 2 cents . The original sound(s) that were recorded emminate from the source through the air in the form of waves of different pressure (compressions & rarefications) . When the compressions are re-created , the speaker cone should be moving toward the listener . when the rarefications are re-created , the speaker cone should be moving away from the listener . All audio equipment should be engineered to this requirement . I'd rather not use the term phase for this issue . Direction is all that comes to my mind (so far) .

 

[paulb]

... The layout of some crossovers - for instance a star earth - does often expect to have a ground on the normal terminal.

How can a speaker 'expect' a ground? There is no ground reference, just the two terminals. Just because there is a common point where most of the components connect, does not mean the speaker 'expects' a ground. Come on.

 

[CanisLupus]

Polarity could be another word for this issue .

 

[Gareth Connor]

To get back to the original question.......

If I use an op amp in an inverting config which moves the signal 180 degrees out of phase, does this cause a problem if both channels are inverted this way? I mean, should/must I use another inverting stage to correct this? If I don't what happens to the audio quality ?

I don't have the second stage right now and it seems fine....

Consider the actual recording and mixing processes and the equipment used to originate the music we all listen to:
Take the single most complex piece of equipment, the mixing console, and assume it is analogue. One important aspect of mixer design is that all inputs to, and outputs from, the mixer are in a coherant phase. Within any signal path between an input and an output there are often several phase inversions. A good example of this is a 4-band parametric equaliser, which will have 4 phase inversions through it: coherant phase is maintained.

In other circuits, inverting buffers will be added to maintain an in-phase signal between the input and output of a circuit block.


Phase inversions will only cause a problem if one channel is out-of-phase with respect to another channel. As long as both channels have the same number of phase inversions, I can see no reason to expect problems.

Regards,
Gareth.