What Makes Amps Out Of Phase?

Discussion in 'Amps/Cabs Tech Corner: Amplifier, Cab & Speakers' started by Teahead, Apr 30, 2008.

  1. Teahead

    Teahead Member

    Messages:
    1,378
    Joined:
    Dec 12, 2005
    Location:
    Northern Ireland
    I'm about to try running two amps at once with the band, a UK reissue AC30tbx 2x12 and a Cornell 18w Plexi 1x12. I already know there's an issue with humming and have a Lehle Little Dual on order to hopefully solve that.

    I do not yet know if there'll be any phasing issues and wondered if such a thing could be predicted according to the type of amps used? Or, do I just have to suck it and see what happens?

    Many thanks for any help,

    Tea.
     
  2. Swarty

    Swarty Member

    Messages:
    1,143
    Joined:
    Dec 16, 2003
    Location:
    Michigan
    The phase (polarity?) is determined by the number of gain stages. It changes 180 degrees at each stage. If you have 2 amps that are firing in different directions simply reverse the speaker leads on one of them.
     
  3. mark norwine

    mark norwine Member

    Messages:
    16,190
    Joined:
    Jul 25, 2005
    Location:
    NJ...GSP135
    Tim's right on.....the issue at hand is "polarity", not "phase". Phase is something altogether different.

    Phase involves time dwell between 2 signals.

    Polarity is simply the electromechanical relationship between 2 signals.

    That said, I've long ago surrendered this fight. The world wants to refer to polarity as "phase".....fine.
     
  4. Teahead

    Teahead Member

    Messages:
    1,378
    Joined:
    Dec 12, 2005
    Location:
    Northern Ireland
    Cheers for the help guys and sorry for my ignorance, Lehle call it Phase reversal in their blurb and I just ran with it. Like Leo caused with the whole tremolo bar thing!
     
  5. fullerplast

    fullerplast Senior Member

    Messages:
    6,792
    Joined:
    May 5, 2003
    Don't feel bad, they can both be used interchangeably in the context of an AC signal. For AC, polarity is virtually the same thing as a 0 or 180 degree phase shift, IOW a non-inverted or inverted signal. (For DC however, phase doesn't exist... only polarity.)

    Regarding your original question, yeah you can count up gain stages to figure it out, but you'll be able to tell right away when you listen to the amps what their relative phases are. As mentioned, you can swap the speaker terminals of either amp, or you can invert the guitar signal in front of the amp with an inverting buffer. Note that some pedals may also invert your signal, and some amps may have channels that are inverted from each other...like the EF86 channel on an AC30, vs the TB channel or the normal and verb channels in a Fender.
     
  6. VikingAmps

    VikingAmps Member

    Messages:
    395
    Joined:
    Feb 21, 2007
    Location:
    VA
    Turning one backwards.
     
  7. Groovey Records

    Groovey Records Member

    Messages:
    3,146
    Joined:
    Nov 19, 2007
    Location:
    Jackson Heights Birthplace of Johnny Thunders
    Alittle of thread but

    Roy Buchannan used to reverse the terminals intentionally to put it out of phase(reverse polarity) and fire one amp straight into the floor with a mike behind it. I saw him at the Village Gate in the mid seventies he played there for like a month and I would go as often as I had the cover. He had a switch on it, it would cancel out the low frequency and his sound was a high wire act.

    Amazing
    EnJoY ThE MuSiC
    GrooVey RecOrds
     
  8. Dave_C

    Dave_C Supporting Member

    Messages:
    14,147
    Joined:
    May 18, 2004
    Well, in the college I earned my BSEE in, polarity always meant a specific kind of phase relationship...that is, one in which two signals were either in phase or 180 degrees out of phase. So, there's nothing wrong with using the word "phase". It's the broader term and is actually probably more accurate in this case because with two different signal paths, there will be phase lag in one of them, so even with identical nominal polarity, there will still be a slight phase differential between the amps.
     
  9. Dave_C

    Dave_C Supporting Member

    Messages:
    14,147
    Joined:
    May 18, 2004
    And, yes, I'm referring to periodic waveforms, which is the context of this discussion.
     
  10. brad347

    brad347 Member

    Messages:
    4,813
    Joined:
    Feb 8, 2006
    Location:
    Brooklyn, NY
    I'm sorry, but I gotta quibble with you there.

    I disagree that they can be used interchangeably in the context of an AC signal, because one is a function of time and the other is a function of electrical polarity.

    Flipping the polarity is NOT "virtually the same" as 180 degree phase shift (not even close!), EXCEPT in one specialized case: That is when you are talking about two sine waves of the same frequency! The misconception is fueled by the fact that the sine wave is often used as an example when discussing these things, but pure, isolated single sine waves are rare in actual music.

    A complex musical signal will comb-filter if an identical signal is phase-shifted, and will disappear entirely if polarity is flipped on an identical signal. A pure sine wave will cancel completely in both cases, hence the misconception.

    Fourier's theorem specifies that any periodic waveform can be reduced to the sum of multiple sine waves.

    However, if you have a signal with multiple frequency components (i.e. anything other than an isolated sine wave), when one of these components is "180 degrees out of phase," then ONLY that frequency component will be 180 degrees out of phase... no other frequency will be.

    For example, if you have a signal with a 1kHz component and a 2kHz component, the 1kHz sine wave has a physical wavelength of twice the size of the 2kHz wave. Therefore, when the 1kHz wave has completed 1/2 cycle (180 degree shift) the 2kHz wave has completed the full 360 degree cycle. So if you had two identical signals and shifted the second signal 180 degrees with respect to the 1kHz fundamental, the 1kHz tone would cancel (zero amplitude) but the 2kHz overtone (second partial) would reinforce and actually have double amplitude!! This cancellation/reinforcement cycle would repeat every other octave in like fashion. This is what comb filtering is. It's called that because the peaks and valleys on a spectrum analysis visually resembles a comb.

    If you took an identical signal and reversed the electrical polarity, on the other hand, BOTH components (1kHz and 2kHz), as well as any other signal component, would be opposite one another in the two signals and cancel, leaving silence.

    If you have two signals, flip polarity and are left with some audio output, it simply means that the two signals you are summing are not identical. It may even sound comb-filtered. It does not mean that you have caused phase shift. Phase shift may be present, but it will be a result of the time relationships between the two non-identical signals. You may have brought that existing phase shift or comb-filtering a little more to light with your polarity reversal. Reversing polarity on two similar (but unlike) signals can change the characteristics of the comb-filtering induced by the initial phase shift, making it either less- or more-obvious. That's why sometimes people flip polarity on one microphone when they have a mic on either side of a drum, for instance.

    With respect to the OP, yes each gain stage inverts polarity. To say it inverts it "180 degrees" is a misnomer, however, because polarity is binary... it's either one way or the other. To say "180 degrees" implies that there is a cycle with points in between, and there is not. The way the speaker is connected can also invert polarity. So someone above was correct-- if your two amps have opposite polarity and are canceling each other resulting in a thin tone, reverse the speaker leads on one of them.

    With regards to why some people call it "phase," there's a reason for that snafu, as well.

    As I understand it, old-school EE types from the dawn of audio amplification used to call it "phase" if it was referring to audio signal, in part to distinguish audio signal from power supply AC. They thought it was a fair compromise. Power supply AC polarity was indeed referred to as "polarity." It turns out the compromise has caused generations of misconception and mis-labeled equipment. Now I just sort of roll with it, but it does add an extra step when you have to parse what someone really means when they say "out of phase." Usually you can figure out which they are referring to from context clues, but it is an extra step of thinking which is needless.

    Remember that audio signal is just alternating current. If you plug an old series-heater amp in and have 120VAC on the chassis, would you say that you got shocked because you plugged in your amp "out of phase?" Nope. You'd say you plugged the polarized plug into the wall backwards.

    :)

    That said, I'm not too picky about what people want to call it. But we don't need to spread more misconceptions about what it is! People can call a "phase inverter tube" a "phase inverter tube" and I will call it a "phase inverter tube" because we all know what we really mean. But most phase inverter tubes do not intentionally use time shift as a component of their goal. There will be some time shift present with most designs, but the idealized goal is to send two identical signals of opposite polarity to the class A/B output section (and a transformer phase inverter will do just that).

    :)
     
    Last edited: Aug 22, 2008
  11. fullerplast

    fullerplast Senior Member

    Messages:
    6,792
    Joined:
    May 5, 2003
    Nice dissertation, and absolutely correct.:D

    I should have specified an AC sine wave, as you cannot shift a complex waveform 180 degrees, for the reasons you have clearly described.

    What I should have said is: In the context of determining the number of inversions in a particular amp if one is using a sinusoidal AC test signal, then a 180 phase shift is the same as a polarity inversion. Thanks for keeping me honest.:eek:
     

Share This Page