Wiring of speaker
UC
series III. These are mirror-image speakers with three drivers and
pots for the mid and tweeter levels. I used to own these same model
speakers, only those were the first series, which were basically the
same, except the speaker alignment was not mirrored.
After hooking them up, I noticed a distinct lack of bass response,
which I thought was odd, as I remember them having excellent bass. I
figured I had accidentally reversed one of the leads from the amp, but
checking this I determined that the leads were connected correctly. It
seemed nonetheless that the speakers were out of phase. I tested this
by placing them face to face and then listening to a recording with
prominent bass whacks (track 6 of Enigma MCMXC A.D.). When I wired
them out of phase, the bass was definitely louder. I deduced that the
woofers were connected out of phase with one another.
I unscrewed the woofers from the from of both speakers and determined
that they were indeed wired incorrectly. The red lead was connected to
the left terminal on the left speaker, but to the right terminal on
the right speaker (looking at the drivers from the front, the open
side). The mid-range units were wired with the yellow wire to the
right terminal and the black on the left; the tweeters were wired with
the pink lead on the left, black on the right. These again were the
same in both speaker systems. Only the woofers were inconsistent. The
soldering appeared to be original, from what I could tell.
Obviously reversing the leads on one of the woofers will bring them
into agreement with phase, but how do I determine which one is wired
correctly? I connected a single 1.5-volt AAA battery to the systems
and determined that the left woofer moves outward with a positive
voltage, and the right one moves inward with the same positive voltage
applied. This proves that the speakers are wired inconsistently, but I
have no idea which one to reverse.
Any ideas? Also, why are the tweeter and midrange wired in opposite
fashion?
Dick Pierce
> I unscrewed the woofers from the from of both speakers and determined
> that they were indeed wired incorrectly. The red lead was connected to
> the left terminal on the left speaker, but to the right terminal on
> the right speaker (looking at the drivers from the front, the open
> side). The mid-range units were wired with the yellow wire to the
> right terminal and the black on the left; the tweeters were wired with
> the pink lead on the left, black on the right. These again were the
> same in both speaker systems. Only the woofers were inconsistent. The
> soldering appeared to be original, from what I could tell.
>
> Obviously reversing the leads on one of the woofers will bring them
> into agreement with phase, but how do I determine which one is wired
> correctly? I connected a single 1.5-volt AAA battery to the systems
> and determined that the left woofer moves outward with a positive
> voltage, and the right one moves inward with the same positive voltage
> applied. This proves that the speakers are wired inconsistently, but I
> have no idea which one to reverse.
>
> Any ideas? Also, why are the tweeter and midrange wired in opposite
> fashion?
The standard convention is that a positive voltage applied to
the positive terminal of the speaker will result in an outward
movement of the woofer. By this convention, the one that's
moving inward is wired oncorrectly. It's altogether possible
that it was wired incorrectly from the day it was made.
As to the reason the midranges and tweeters aren't consistent
with the woofer, it's probably done correctly. The reason has
to do with how the crossover influences the phase of the
signal feeding the drivers. If, for example, the crossovers
were 2nd order designs, the result is that at the crossover
frequency, the woofer output has a phase of -90 degrees, and
the midrange output has a phase of +90 degrees. The difference,
at the crossover frequency, is 180 degress, and the result would
be an on-axis null in the frequency response. To prevent this
null, the phase of the midrange is reversed, resulting in a
consistent frequency response through the crossover region.
The tradeoff is that you have a flat frequency response but
now have a 180 phase reversal as you move through the crossover
frequency, or an all-pass function.
Flip the leads on the wrong woofer, and you'll be all set.
--
+--------------------------------+
+ Dick Pierce |
+ Professional Audio Development |
+--------------------------------+
UC
I had guessed exactly what you have written here (both points) but
kept this to myself to see what you would say. I figured the right
woofer was wired wrong, and that the midrange was supposed to be out
of phase because of crossover phase changes.
Does this produce any sort of response dip at the crossover points?
Dick Pierce
> On Oct 12, 3:34 pm, Dick Pierce <dpie...@cartchunk.org> wrote:
>>As to the reason the midranges and tweeters aren't consistent
>>with the woofer, it's probably done correctly. The reason has
>>to do with how the crossover influences the phase of the
>>signal feeding the drivers. If, for example, the crossovers
>>were 2nd order designs, the result is that at the crossover
>>frequency, the woofer output has a phase of -90 degrees, and
>>the midrange output has a phase of +90 degrees. The difference,
>>at the crossover frequency, is 180 degress, and the result would
>>be an on-axis null in the frequency response. To prevent this
>>null, the phase of the midrange is reversed, resulting in a
>>consistent frequency response through the crossover region.
>>The tradeoff is that you have a flat frequency response but
>>now have a 180 phase reversal as you move through the crossover
>>frequency, or an all-pass function.
>
> kept this to myself to see what you would say. I figured the right
> woofer was wired wrong, and that the midrange was supposed to be out
> of phase because of crossover phase changes.
>
> Does this produce any sort of response dip at the crossover points?
No, the phase inversion between the two drivers is done
explicitly to PREVENT a response dip at the crossover.
Try it again, assuming reasonably standard 2nd order
high- and low-pass functions, at the crossover point,
the response of each driver is at half its pass-band
power, half its pass-band slope, and half its pass-
band phase. A second order low pass' ultimate stop-
band phase shift is -180 degrees and -12 dB/octave,
that of a 2nd order highpass is +180 degrees and
+12 dB octave.
Thus, at the crossover frequency, the woofer will
have half its slope (-6 dB/octave), half its phase
(-90 degrees) and half it's power (-3dB). The midrange
(or tweeter, as the case may be) will conversely have
half its slope (+6 dB/oct), half it's phase (+90 degrees)
and half its power (-3dB).
Sum the two if in phase, the woofer at -90, the tweeter
at +90, the difference is 180, or cancellation, and you
have a response dip. Now, flip the phase of ONE of the
drivers (the midrange in your case), and now the phase of
the midrange is -90, which is now in phase with the woofer
and no cancellation and thus no response dip.
Assuming everying is working ideally, which, in reality, it
isn't. But it's close enough.