Purely acoustic crossover?

[maggiesnmacs]

Has anyone used a sealed enclosure as a high pass filter? I'm looking at maybe using 2 Dayton rs100-4 as extended range midranges in a 3-way. My goal is to cross low to a woofer where the crossover won't be too noticeable. I was thinking, what if I built a sealed enclosure that was about 1.5 liters for both drivers. This would lead to an fb of 144 hz, a total q of .9 and f3 of 120hz. So wouldnt the response essentially act like a second order 120hz high pass with a slightly higher q than a standard lr crossover?

 

[giralfino]

Without a crossover you'll run out of excursion on the midrange very quickly. Have a simulation with unibox.

Ralf

 

[system7]

This sort of small sealed back idea is actually used in some cone tweeters and mids:

Phenolic Ring Tweeter Replacement for AR-4X CTS Marantz and More 8 Ohm

M 10 - 8 Ohm

In practise you always use at least a capacitor to take out the low frequencies. And IMO, they sound better with a second or third order high pass electrical filter. Nothing wrong with a proper crossover!

 

[maggiesnmacs]

Ya I think what I may end up doing is the sealed back crossover idea with a first order cross

 

[adason]

...or you can use active crossover and biamp it...you get flexibility of crossover frequency and slope

 

[maggiesnmacs]

I do have an active crossover but I was specifically trying to make a speaker that can be single amped with the option to bypass the crossover if biamping is what you wanna do

 

[bigbottom]

Has anyone used a sealed enclosure as a high pass filter? I'm looking at maybe using 2 Dayton rs100-4 as extended range midranges in a 3-way. My goal is to cross low to a woofer where the crossover won't be too noticeable. I was thinking, what if I built a sealed enclosure that was about 1.5 liters for both drivers. This would lead to an fb of 144 hz, a total q of .9 and f3 of 120hz. So wouldnt the response essentially act like a second order 120hz high pass with a slightly higher q than a standard lr crossover?

You can do this as you describe. Some problems are:

If the mid and woofer are both level response, not hp/lp slopes, then you can get a highbass peak as both drivers sum ~ 120Hz

At bass you are driving both in parallel so the system is low ohms at bass.

Very low xovers sound like a good idea in theory [say when drivers are closer than 1/4 wavelength] but in a passive xover it can throw up problems, like having a messy upper bass, expensive parts, or not finding a sensitive enough low bass driver to match mid levels.

 

[AllenB]

You could use a capacitor to cross below fc, and a second order filter would improve power handling etc. Closer to fc second order is a little less sensitive to the impedance peak. Filtering the peak would help more, but a cap at the input to the amp would be just as good.

 

[system7]

This is one of those interesting, off-the-wall ideas that actually has something going for it.

Eminence build a sealed back midrange that does exactly what maggiesnmacs is thinking about.

Speaker Detail | Eminence Speaker

In practise you are going to end up with a sharp high Q knee at the bottom. The theory says you will hear that resonance. Once you filter it with a crossover you end up with a sharpish rolloff. This overall sharp bend is called a Chebyshev filter. TBH, it's not often employed in speakers, I think I prefer gentler smoother curves, but OTOH, until you build it and give it a listen, who's to knock it?

I show you the sim with a 5" midrange in 0.5L and 8L below. You'd obviously modify the filter for the smaller midrange enclosure.

 

[midrange]

Why do you want to use that size of midrange enclosure? What are the advantages of the peak at 144Hz? At the levels I listen at in my small UK rooms, excursion would not be a problem but a lot of forum members listen at high levels in big rooms. I think the point bigbottom makes of low impedance at bass frequencies is very valid.

 

[sreten]

Hi,

The reasons it won't work well are the same as why you
can't use the 2nd order roll-off of a tweeter as the x/o.

rgds, sreten.

 

[Bob Richards]

I've done it. No reason why it necessarily won't work well enough. The most likely problem will be cone excursion of the mid drivers if you ever turn it up loud. Just make sure the woofer rolls off fast where the acoustic outputs of the mids become significant.

 

[system7]

I'm all for running an idea up the flagpole and seeing if anyone salutes. But after a brief look I really don't like the midrange in a small volume idea.

Let's look at this rather lovely 4" scanspeak mid:
https://www.madisoundspeakerstore.com/approx-4-fullrange/scanspeak-discovery-10f/8414g-10-4-full-range/

I don't know quite what Troels is doing with it here, but it looks like about 4L of enclosure.
ScanSpeak-3W-Discovery

If you go smaller on enclosure, you end up with a bass peak that just goes away totally from the Steen Duelund theory. The whole idea is smooth curves that integrate well and magically line up phase. Impedance looks quite benign except for the usual reflex bass impedance horrors. Maybe closed box might suit a valve amplifier better.

FWIW, I found that offsetting the mid and tweeter to the cabinet edge was essential to avoid some unpleasant diffraction dips.

 

[bwaslo]

If you want to use the ported box and make it typical tube amp friendly, you can put a RL network across the whole speaker, something like 3mH in series with 10 ohms or so. Will bring those impedance bumps right down and make the speaker better behaved with different amps or long cables.

 

[midrange]

Those drivers get you towards FAST territory. 1.86 litres each driver, total Q .707 F3 113 Hz. How about just adding a sub that goes high enough? Unless the level of output you need will overload the mid drivers.

Two 4ohm drivers in parallel, and then another bass driver in parallel is not amplifier friendly at all (or were you planning on using the midrangers in series?)

 

[Bob Richards]

The first speaker system I built (when I was 12) was a 12 inch woofer and a 6 inch full range driver in a big cabinet. I put a cap in series with the 6 inch driver, which may have been in a sub-enclosure (can't remember). I had no idea how to get the cabinet volume right. I didn't know the math for calculating the cap size. I just looked at what other speakers used for a cap, and guessed from there. I did jam a bunch of fiberglas insulation in the cabinet to reduce resonances.

It was horrible from a technical standpoint, based on what I know now, and yet I enjoyed the result. I guess the more mistakes you make, the more you get to learn. That was about 100 speaker projects ago. It's been a fun hobby. Now I'm 60 yrs old and still doing it.

Designing and building a very excellent speaker system is actually very complicated if you strive for perfection. It could overwhelm a newbie. Having good tone controls in the preamp can help make a poor speaker much more listenable.