What slope should I use on the bass section of a 3 way speaker?

[fjhuerta]

Hi,

I was modelling some stuff in order to build the 3-way I've been thinking about, and had a question regarding slopes...

I can easily hit 4th order on the mid-tweeter, but I can't seem to find a suitable topology for a 200 Hz 4th order. Not even a 4th order electrical seems to help.

2nd order is easily achieved with a simple 2nd order electrical filter. But I worry about acoustical cancellation or other nasty effects on the high range.

I chose 200 Hz / 1600 Hz as preliminary crossover points.

Do you think I could get away with it?

[Jed]

Your question is a bit confusing because we don't know what drivers you are talking about in your design or why a 4th order filter would be beneficial in the design over a simple 2nd order. More details are needed.

Jed

[Dave Bullet]

From what I've seen - designers that use a 2nd order acoustic on the woof/mid crossover tend to have the mid/tweeter xo a little higher, so that the woofer is 40dB+ down. I suppose this is to minimise as you pointed out destructive interference.

Take a look at Dennis Murphy's and Roman's 3-way designs....

I therefore think a 200/1600 xo point "might" cause some perceivable interference if the woofer isn't phase tracking, since the woofer will be "only" 36dB down at mid/tweeter xo. An example where this might be handled well is Zaph's ZDT3 - where (from memory) the woof/mid XO is 800 with target 2nd order acoustic and mid/tweeter xo is approx 3500Hz. Of course this design may work well for other reasons I don't understand (or woofer / mid phase tracking is still good... or not relevant at 3500Hz).

Don't ask me what the impact is on vertical lobing.... Something I can't model and don't yet understand how to predict depending on xo type and slope.

The above of course is not knowing the drivers intended (presume they are Dayton RS?) - and whether breakup nodes will be sufficiently down with only 2nd order acoustic. They probably will be though at a lower (200Hz) xo point.

Cheers,
David.

[fjhuerta]

Thanks a LOT for your comments, Dave. They made a lot of sense, and are definitely helpful.

Jed, the drivers are the 225 / 150 / RS28. I've been looking at the WMTMW design with these drivers on the Missions Accomplished section, but I think I'd like to try something a bit different in the crossover. A lower X-over point is one of the things I'd be after, some sort of diffraction control in the baffle, and a flatter tweeter response off-axis would be the others.

[Jed]

To hit 4th order acoustic on the RS225 wtih a Fc of 200hz, try a 3rd order electrical cauer type filter with a zobel.

[Dave Bullet]

Hi Javier,

One more thing to consider is which driver will reach excursion first at those xo points. I presume the RS28 will (if 4th order acoustic @ 1600Hz). Might pay to do some tweeter excursion modeling, depending on your desired max SPL may not be an issue.

Cheers,
David.

[rc white]

The crossover needed for the WMTMW configuration is different from that needed for the WMT type because there is no lobe steering with the first configuration as this is taken care of by geometry.
It is therefore easier to design a crossover, since you do not need to take lobe steering into account.
There are also three way all pass crossovers, Bullock wrote and excellent AES paper about these but I haven't seen any popular articles on the net about them .
I would seriously consider bi-amping for such a low w-m crossover.
Looking at the rs28 specs it does not specify an xmax, with a 1600Hz. region crossover the tweeter is handling about one quarter of the average power input with a "typical" signal, with the output capabilities of a wmtmw that is a lot of excursion.
Audax make a metal dome tweeter with a rated xmax of .35mm, that is the largest I have seen for any advertised tweeter.
rcw

[Deward Hastings]

Have to agree with rc here . . . this is a candidate for an active crossover. Both the RS225 and the RS150 are well behaved on the “other side” of the proposed W/M crossover, so a textbook LR4 works just as expected (which is very well). If you are using PCD for your sims that’s over on the right hand side of the spreadsheet . . . just plug in the filter description and numbers and turn the filter “on”. Actually “building” it is just as easy.

Crossover cost is reasonable . . . $50/side (using the Behringer CX2310) and you get turn-the-knob adjustability of the crossover frequency and reusability for future designs. For an extra $10/side you can add a second LR4 for the M/T (with a Behringer CX3400), so it’s probably worth it (there are other “extra features” too) to get the additional capability even if you don’t initially use it. For twice that price you can get parametric and shelving equalizers, individual driver delay (or advance), and a slew of other features that the RS combo doesn’t need but which might be useful someday if you keep experimenting with other three-ways.

Odds are you already have the additional channels of amplification sitting around somewhere . . . if not the 60 Watts/channel needed (that’s all the RS150 and RS28 can handle) can be easily had used or DIY for $1/Watt.

The RS150 will sound best (cleanest) in the upper mids with plenty of stuffing behind it to absorb the backwave.

I’ve been surprised (shocked, even) to hear how much a number of different midrange drivers sound alike (even different cone materials . . . metal, poly, paper) when the signal to them is bandpass limited and the driver/crossover interaction is eliminated. I’m coming to believe that much of the difference we hear is because of behavior *outside* the intended passband, and we hear that only because of the inadequacy of the more common crossover designs to deal with it.

An aside to rc regarding tweeters . . . the Seas Excel T25CF-002 lists “Linear coil travel (p-p) 1.0 mm”, or an xmax of 0.5 mm.

[Dennis H]

The RS28 will work fine crossed at 1600. Plenty of projects around here use it. Active crossovers -- nothing wrong with them but you can do a passive that will work just fine and make things a lot simpler to hook up -- one amp, one wire, one speaker. See Jeff B's high-efficiency design with the OB mid and Dan and Jed's variations on the theme. Jeff used LR4 for the MT and LR2 for the WM.

[Deward Hastings]

I've used the RS28 with the RS150 in a two-way crossed at 1500 and it worked fine . . . but of course the whole thing is SPL limited to 102dB. In a three-way one might be putting a little more power to it. It would be nice to see rated xmax for it to get a better idea of what it should be able to do.

I've looked at the crossover BOMs for a couple of RS225/RS150 designs, and see as much as $100/side for the components attributable to the W/M crossover alone. And the crossover wiring makes two amps, two wires look not all that complicated by comparison. But of course it can be made to work . . . and high power amps that still sound good full range are no longer all that expensive. The rising cost of copper and foil and the falling cost of modest-power amplifiers of excellent quality has, however, laid rest the argument that active crossovers cost more to implement . . . they may even be cheaper in many cases. And the overall quality and performance arguments tend heavily to favor active . . .

[JRT]

Audax make a metal dome tweeter with a rated xmax of .35mm, that is the largest I have seen for any advertised tweeter.
rcw

http://www.seas.no/images/stories/excel/pdfdataheet/e0047_t29mf001_magnum_datasheet.pdf