A few crossover questions

[Seth Smith]

Hola amigos,

A few questions for the electrically inclined. My search turned up a whole lot of not what I was looking for so I turn to you for specific answers.

Would someone please explain the effects of the crossover on the time domain, (phase shift?). More specifically, is the shifting a product inherent to the electrical circuit or is it specific to the reactive components of a passive network. What I really want to know is will I still have a problem with phase shifting using a DCX-2496 instead of a traditional cap/ind passive network. All points will be LR4 so ideally, my shift should bring me back to 0 degrees, eliminating any need for reverse polarity wiring, but I'm worried about time smear.

I will be using all seas drivers in my 3 way project, W26, W18EX, Millenium. All fall within 1 Db of each other within my intended passband negating the need for any padding. It also makes my life alot easier considering it will be an active tower and individual channel gain adjustment is not a feature on my amp. Assuming I'm a cheapskate and bought a cheaper XO unit, say a rane 2 way, and used it to split woofer from mid/tweet, would the passive net used in the mid/tweet upper reduce the sensitivity of the drivers significantly?

Thanks for any help guys, I like to talk things out, its the only way I learn, so please feel free to write down any and all thoughts.

 

[5th element]

The phase is a part of the crossover etc, not the components used, at least i think thats right.

 

[sreten]

The phase shift of a network is inherent to its function, so passive
or active doesn't make any difference, in terms of filter functions.

Its actually the case LR2 is better in the time domain, even with
reverse polarity wiring, to avoid "time smear" you should be
looking at phase degrees vs frequency, as a constant time delay
implies more phase degrees the higher you go up, Bessel filters.

Personally IMO LR2's are near Bessels so .......

I'll also note with 3 drivers of the same sensitivity gain must
increase by 6dB from treble down to bass due to baffle step.

sreten.

 

[Seth Smith]

Thanks for the replies, keep 'em coming.

I think the Behringer has a provision for level adjustment of the channels but that leaves me with a problem yes? Simply adjusting the level of the woofer section crossed at 300 doesn't cover the whole baffle step, and increasing the level of the midrange will give me a nasal character, so all I can do is increase the woofer and play with the built in EQ to bump the mid as well as I can right?

More importantly though, what formula do you use to determine where baffle step will occur, or is this more a measure and see thing? Is it 1/4 or 1/2 wave along the front, and is it the whole width of the baffle or the width from cone center to baffle edge? Understanding how to figure it will do me alot more good than simply knowing some generalizations.

Also, back to the passive componenets, do high quality passive components present a significant parasitic load to the system to the degree that they reduce overall sensitivity by a significant margin?

So Sreten, using the LR2 net for the 3 way, I end up with less shift in the time domain of course and to take care of phase shifting I would wire the mid reverse phase right? Sounds like a good plan, hopefully I can cross low enough to avoid those nasty resonances of the W18, if not, I'm doomed to a 1/3 octave...Any thoughts?

 

[Dave Jones]

Any filter must necessarily have a time lag. That's because the output is a weighted sum or integral of the input over time. The filter has to "see" more than an instantaneous peek at the signal. There are designs that incorporate an "all pass filter" whose sole purpose is to delay one signal enough to match the delay on another.

 

[Seth Smith]

That was my original thinking with the LR4 network, it does have an all pass nature.

Now on to baffle step, can anyone answer my questions on that beyond the standard "use a .5 way" or mount a rear facing bass driver?