Print

Please login or register.
1 Hour 1 Day 1 Week 1 Month Forever
Login with username, password and session length

[Ivan Beaver]

Seems like reversed polarity on the horns would cause cancellation on a broader ranger of upper frequencies than just a 1.25k dip?

That is another one of those "it depends" on the particular situation/product.

I have seen situations that reversing the polarity will cause a pretty wide cancellation notch and other times a very narrow deep notch, and other times reversing the polarity only produces a dB or two "wiggle" (that probably would not even show up on an RTA).

HOWEVER the PHASE takes a 180° shift-and that can really screw with the other cabinets in the system. 

Of course an RTA will not show that either (the phase anyway)-but the interaction it would show.

I had a situation a number of years ago that had 2 cabinets in a center cluster (passive cabinets).  There was some weird interaction going on.  So i flipped the polarity on one cabinet and the area in the middle got better-but all the bass was gone.

So I pulled out the measurement rig and noticed that one of the cabinets had a 180° phase shift right around mid/high crossover.  AH HA-THAT is the problem

So we dropped the cluster and found that one of the HF drivers was wired backwards.  PUt it back right-reflew the cabinets and all was well.  Nice and smooth down the middle.

So just because it makes sound and shows "flat" in the amplitude response-DOES NOT mean that it is correct.

[Kevin McDonough]

Seems like reversed polarity on the horns would cause cancellation on a broader ranger of upper frequencies than just a 1.25k dip?

Well not really, as it would have nothing to cancel with.

Within the single cabinet, each driver is playing their own set of frequencies and the only place where cancellations can happen is at the crossover point, where for a brief set of frequencies they're both playing. The cancellations happen from the small set of frequencies from the LF driver combining with the same frequencies from the HF. Once you get out of that small band then its back to just a single driver playing. To me it sounds like the two drivers are out of phase, at least partially, at this point causing the cancellation.

If this were a biamped cab with an active crossover it would be a two second job to apply a little delay to the HF and bring it into line with the mids.

However as its passive, the only option it so try and swap the polarity on the HF.  Depending on how far out of phase it was to begin with it may bring things more into line, or it may make them worse. We cant measure ahead of time so the only way to know is to try and see. However it shouldnt make any difference at any other frequency as, when a single cab is playing, there will be nothing else playing the upper frequencies to cancel with.

HOWEVER the PHASE takes a 180° shift-and that can really screw with the other cabinets in the system. 

Very true, but I had assumed that when the swapping of the HF polarity was tested in a single cab, if it did prove to be an improvement to the speaker then the other cab(s) would have the same done to match.

I have been able to rule out the mic as a problem.  I measured a different set ofspeakers with the same RTA and got very close to a flat response with no issue at the 1.25 level.  I also measured each individual speaker 1 at a time at different locations and still get basically the same result.  I really was not wanting to have to open the speaker up.  I think I will just leave things as they are since the sound does not seem to reflect any extreme negative affects from the problem.  Thanks to everyone for your help and ideas.

The only thing I would suggest, short of opening up the cab, is lifting it outside and measuring there. Even though you've tried measuring the response by moving the mic to different locations, there is still the possibility that you could be measuring a node or acoustic anamily from the position of the speaker within the room. Its unlikely at this point, but the only way to rule it out for sure would be to measure outside (or in an anechoic chamber) where there are no reflections.

While I can understand your reluctance to open the speaker up if you've never done it before, it really is a simple job. a few screws will hold the grill into place, and then a few more will hold the horn in position. Once you've undone these the whole horn and compression driver assembly will slide out and you'll be able to get to the leads and try swapping them. with a power screwdriver or drill-driver, 60 seconds tops    However if your happy just to leave things as they are then all the best 

k

[Ivan Beaver]

Well not really, as it would have nothing to cancel with.

Within the single cabinet, each driver is playing their own set of frequencies and the only place where cancellations can happen is at the crossover point, where for a brief set of frequencies they're both playing.

Not always true.

In some designs there is a good bit of overlapping of freq going on-sometimes as wide as an octave.  that is why is it important to look "outside" of the intended bandwidth.

One of the "misunderstood" things in audio is that the crossover "stops" the freq reproduction at a particular freq.   The "crossover freq" can mean different things-depending on how you look at it.  In many cases the electrical and acoustical crossovers are NOT the same freq.

And in any case-it is a slope (that has varying amounts of reduction)-based on both the electrical response and the acoustical response of the drivers in question.

As usual-when you start to dig a little deeper to get to the REAL answer-it often gets a bit more complicated.  And that is why so often the "simple answer" is simply wrong.

[Steve Roth]

I have tried the microphone in different positions and have still obtained basicially the same result.  Thanks - Brian

by different positions, if you mean different positions out in the far field, you are perhaps missing something. measure real close like 1' in front of the speaker, then compare that to 'out in the room', to get a feel for whether you are measuring the speaker or the speaker/room/nearby reflecting surfaces combination - this may help you to narrow down what you are seeing

[Kevin McDonough]

Not always true.

In some designs there is a good bit of overlapping of freq going on-sometimes as wide as an octave.  that is why is it important to look "outside" of the intended bandwidth.

One of the "misunderstood" things in audio is that the crossover "stops" the freq reproduction at a particular freq.   The "crossover freq" can mean different things-depending on how you look at it.  In many cases the electrical and acoustical crossovers are NOT the same freq.

And in any case-it is a slope (that has varying amounts of reduction)-based on both the electrical response and the acoustical response of the drivers in question.

As usual-when you start to dig a little deeper to get to the REAL answer-it often gets a bit more complicated.  And that is why so often the "simple answer" is simply wrong.

No, your right very true.  when I say a "small band" of frequencies in my post I should have been more clear in that I mean this as small relative to the overall operating frequency of the whole speaker. But your right the actual width of this band will vary a lot depending on the types of crossover slopes/settings/electronics used etc 

[Tim McCulloch]

No, your right very true.  when I say a "small band" of frequencies in my post I should have been more clear in that I mean this as small relative to the overall operating frequency of the whole speaker. But your right the actual width of this band will vary a lot depending on the types of crossover slopes/settings/electronics used etc 

The *acoustic* crossover may well extend over an octave on either side of "center."  I do out-of-band correction for at least 2 octaves above/below crossover.

[Doug Fowler]

Seems like reversed polarity on the horns would cause cancellation on a broader ranger of upper frequencies than just a 1.25k dip?

No, only where the interaction between pass bands is greatest, meaning at crossover and when levels are (mostly) equal.   Reduce the level on either and the null is also reduced, assuming it is indeed crossed over @ 1.25 KHz.

[ProSoundWeb Community]