What’s best practice for mixed Xovers / driver polarity?

tajanes

I.e. 2nd order LP on mid with 3rd order HP on tweeter?

VituixCAD modeling leans me one way (although not near as definitive as outcomes of inverting a driver with a non-mixed Xover), measurements at /near listening area are pushing me the other way.

???

Steve_Lee

.

Wolf

12/18 elec yields 4th acoustic, and should typically be IN PHASE or same polarity.
There are however companies that wire drivers backwards to the terminals like JBL does. If you measure the impulse response and the peak goes down with positive to red wiring, then the polarity colors are reversed.

tajanes

@Wolf said:
12/18 elec yields 4th acoustic, and should typically be IN PHASE or same polarity.
There are however companies that wire drivers backwards to the terminals like JBL does. If you measure the impulse response and the peak goes down with positive to red wiring, then the polarity colors are reversed.

Thanks. That is what I’m seeing in actual measurements- positive / additive to the higher end of the Xover point (vs a bit of destructive interference when inverted), but a bit of negative on the low side of the Xover point.

tajanes

...Or maybe just best not to mix odd with even filters due to phase being 90 degrees off? Can deal with 180 (flop + with -), but 90 and 270 problematic.

I'm going to look at comparing the 2nd order LP / 3rd order HP with both; a 2nd order LP / 4th order HP, and a 3rd order LP / 3rd order HP, measure and look at the level of squigglies (technical term-ha), i.e. phase issues constructive and deconstructive interference around the Xover range.

Wolf

Usually the 90deg off is required due to driver Z offset.

tajanes

I tried a few different Xover formats, keeping with the current 2nd LP/3rd HP component values, just adding an additional inductor (of differing small values) where appropriate to change the format / orders.

Comparing to the initial 2nd order LP / 3rd order HP (non-inverted), I set up a 3rd order LP with the 3rd order HP, as well as the 2nd order LP with a 4th order HP. As such, trying to minimize the shifts (although the roll offs of course changed a bit) with the idea of just focusing on the affects on the (squiggles) interactions due to phase differences of the Xover formats. In the end, their was little difference (at least from measuring from a distance- near the listening area) in the outputs. While the squiggles moved around a bit (up / down frequency) their range was little changed. The biggest shifts were in looking at results with vs without inverting one driver.

Sticking with the current 2nd/3rd format, and moving to a less critical 'smoothing'

Eggguy

I have also been in the crossover twerking stage and I am trying a new strategy of giving priority to above axis response. My project has a sloped baffle but it is a short floor stander. I am currently listening to what I consider the best so far, a third order electrical on the tweeter and a second order on the woofer. I do not recall the tweeter polarity because I use a toggle switch, but one is much better above tweeter axis so that's what I' m definitely leaning towards.

Eggguy

When you tweek while working, its called twerking.

tajanes

@Eggguy said:
I have also been in the crossover twerking stage and I am trying a new strategy of giving priority to above axis response. My project has a sloped baffle but it is a short floor stander. I am currently listening to what I consider the best so far, a third order electrical on the tweeter and a second order on the woofer. I do not recall the tweeter polarity because I use a toggle switch, but one is much better above tweeter axis so that's what I' m definitely leaning towards.

Interestingly, my smaller 3ways will be sitting on very short stands with a ~ 5 degree tilt (just under 31 in. ht), so similar positioning. My last umpteen projects have been woofer(s) with mid/tw coaxials (or just woofer with broad rangers) so maybe seeing and comparing measurements of separated mid and tweeter to my previous builds is what was giving me a go-around.

jr@mac

It is very common to have a 2nd order electrical on the woofer/mid and 3rd order on the tweeter, when targeting 4th order LR anyways.