In three-way crossover design, besides meeting points between tweeter/midrange and midrange/woofer, there’s a minor point where tweeter and woofer met each other.
For instance, the red arrow is the stated point. And it locates approximately 14dB below reference line (0dB).
Is there a rule (or a rule of thumb) suggesting how much dB below reference level should the tweeter and the woofer meet each other?
Cr. for picture: PassDIY
For instance, the red arrow is the stated point. And it locates approximately 14dB below reference line (0dB).
Is there a rule (or a rule of thumb) suggesting how much dB below reference level should the tweeter and the woofer meet each other?
Cr. for picture: PassDIY
You could expect about a 1.6 dB bump in frequency response at 1000Hz, if the curves were electrical and complementary.
As then, the passbands would be summing coherent signals.
For coherent signal summation https://sengpielaudio.com/calculator-coherentsources.htm
As then, the passbands would be summing coherent signals.
For coherent signal summation https://sengpielaudio.com/calculator-coherentsources.htm
No. Because what you are showing in the 3way crossover from PassDiy is not reality. In reality both, the woofer and tweeter have their acoustic fr responses, as they rollof naturaly outside of their useful range. Tweeter is not flat accross the midrange, its response goes down at mids, depending on tweeter offcourse. This adds to the actual response plus electric response of the filter.
Overlap between woofer and tweeter is completely irrelevant. There are many other parameters way more important for speaker design. Woofer and tweeter overlap is not any of them.
Disagree gents.. Even with optimal xover separation, like with xovers all a decade apart, there will be an audible effect....and for 2nd order acoustic.
Push xovers tighter together and audibility magnifies.
Bottom line is this imo...
Where do you call the line where out-of-band summations don't matter? How many dB down? It's the same question as when needing to knock down out-of-band peaks.
Or iow, how much response ripple are you willing to tolerate?
Push xovers tighter together and audibility magnifies.
Bottom line is this imo...
Where do you call the line where out-of-band summations don't matter? How many dB down? It's the same question as when needing to knock down out-of-band peaks.
Or iow, how much response ripple are you willing to tolerate?
3way crossover points depend on the drivers selected. Assuming small mid which can easily be crossed 3-5kHz, small tweeter can be selected. If mid can go as high as 7-10kHz, nice ribbon tweeter can be used.
Assuming 200-300 Hz woofer cutoff, and small tweeter early rollof at mids, they the woofer and tweeter may overlap 60dB below. Not 14dB like thread starter suggested. Not audible. I disagree with you unless you provide data on audibility between woofer and tweeter overlay in 3way.
Btw, they will always overlap. Just a matter of scale. Select -1000dB like some tb fullrange does, and the overlay is uuuge.
Assuming 200-300 Hz woofer cutoff, and small tweeter early rollof at mids, they the woofer and tweeter may overlap 60dB below. Not 14dB like thread starter suggested. Not audible. I disagree with you unless you provide data on audibility between woofer and tweeter overlay in 3way.
Btw, they will always overlap. Just a matter of scale. Select -1000dB like some tb fullrange does, and the overlay is uuuge.
Here's a simple best-case 3-way for this issue, imo. Xovers are a decade apart @300Hz and 3000Hz. 2nd order.
The woofer and tweeter both contribute at about only -21dB in the middle of the midrange.
In this ideal case, there is a +1.2dB broad tapered bump in mid-range response. That will be audible, and needs attenuation imo.
If the mids carried a wider passband than a decade, covering say 200Hz to 5000Hz, mid range bump goes down to a minor +0.5 dB.
Woofer and tweeter tails are more attenuated.
A narrower mid passband, say 700 Hz to 3000Hz, gives a broad +2 dB tapered bump across mid. Tails matter more.
Point is imo, the summation is case specific, and we need to know that it can cause audible issues....and not dismiss it based on some particular design we have in mind.
The woofer and tweeter both contribute at about only -21dB in the middle of the midrange.
In this ideal case, there is a +1.2dB broad tapered bump in mid-range response. That will be audible, and needs attenuation imo.
If the mids carried a wider passband than a decade, covering say 200Hz to 5000Hz, mid range bump goes down to a minor +0.5 dB.
Woofer and tweeter tails are more attenuated.
A narrower mid passband, say 700 Hz to 3000Hz, gives a broad +2 dB tapered bump across mid. Tails matter more.
Point is imo, the summation is case specific, and we need to know that it can cause audible issues....and not dismiss it based on some particular design we have in mind.
I thought i just did. You don't think you could hear a broad bump of 1+dB or so, across the midrange?
I know I easily can, and I have old ears...
I know I easily can, and I have old ears...
I've got the Weems book somewhere that discusses this and if I remember correctly [ and I may not] it talks about the need for rolling off the woofer a little early and the tweeter a little late to take this effect into account. So assuming ideal drivers and text book filters if the mid is filtered at 300 and 3000 the woofer rolls off at ~{300-x%] and the tweeter at {3000+x%}
You think ideal drivers exist? First, real drivers have their own rollofs besides the crossover, which adds to actual steepness of the filter.
Second, no drivers in 3way come with the same efficiency, as we do not live in ideal world and 3way speakers rarely use only textbook filters. Mid and especially tweeter will need fixed resistors or lpads, because they will likely have higher efficiency. Some may need tilt or notch filter.
Now when you two gave poor presscot the reason to worry about woofer and tweeter overlap, please help him make perfect 3way speaker with ideal drivers and textbook filters. I am out.
Second, no drivers in 3way come with the same efficiency, as we do not live in ideal world and 3way speakers rarely use only textbook filters. Mid and especially tweeter will need fixed resistors or lpads, because they will likely have higher efficiency. Some may need tilt or notch filter.
Now when you two gave poor presscot the reason to worry about woofer and tweeter overlap, please help him make perfect 3way speaker with ideal drivers and textbook filters. I am out.
Of course not.
It was a post for contemplation and interests sake. However some drivers come closer to ideal than others I am sure, although I'd never be able to afford such rare and highly desirable objects.
I'm so old and deaf now I can barely hear myself think let alone one dB boost in that range; it needs to be greater than 3dB at least.
However what I posted is a valid concept and could be taken into account when designing and building passive crossovers.
It was a post for contemplation and interests sake. However some drivers come closer to ideal than others I am sure, although I'd never be able to afford such rare and highly desirable objects.
I'm so old and deaf now I can barely hear myself think let alone one dB boost in that range; it needs to be greater than 3dB at least.
However what I posted is a valid concept and could be taken into account when designing and building passive crossovers.