SB Acoustics monitors build

[DeZZar]

Quick update for everyone interested in this build. Measurements are in and I thought I would attach them here in case there is a guru out there that would like to try modelling something up.



Few challenges with this one. Woofer breakup hits nearly 20db above target and the 3K mess on the tweeter is diffraction.

[DeZZar]

I've come up with several crossovers so far and tried a few of them. The speakers sound nice and when you get the balance right the bass is really solid from such a small setup.

I haven't really decided which topology to go with yet as they are currently all sounding quite similar. Here is the measurement from my last attempt but not quite where I want it just yet.


This still sounds fairly bright on some material - jut about everything from Jaz to pop sounds excellent but when you get into some rock/metal the tweeter just sounds like it takes over - not really sure why given its level is suppressed and basically flat - several db below bass levels - I've had this challenge with this tweeter before and I'm starting to think maybe its just not a great tweeter for metal - I could be (and probably am) wrong.

Irony is my neighbor is completely stoked with what he hears so at this stage its just my fussiness that's stopping me from putting a bow on them!

I'll keep tweaking and post up my schematics shortly. In the meantime - anyone else is welcome to take a crack at it

[Steve Lee]

That 5.5 ~ 7.2 KHZ region is what is biting your ears - its too hot - try some EQ in that area with the Rock music again.

[DeZZar]

That 5.5 ~ 7.2 KHZ region is what is biting your ears - its too hot - try some EQ in that area with the Rock music again

You would think so, but I've slammed the door shut on that woofer with eliptical filters and still the same. There seems at this point to be little difference between a fourth order eliptical and a second order with tank.

For a lot of music, the tweeters not really working that hard. When you disconnect the woofer it surprises you how little it seems the tweeter is doing. However, when you throw on some metal, suddenly the tweeter really comes to life.

EDIT:
Sorry - miss-read your post. Thought you were referring to woofer breakup! Yeah next round I'm going to try leveling out that tweeter top end and see if that improves it!

[Geoff Millar]

How do the same rock/metal tracks sound on your other speakers? I ask because some albums, such as my pet hate sound-wise, the Red Hot Chili Peppers - sound quite harsh on our good speakers, but are almost tolerable on speakers with muted treble, such as our little 'Panasonic' bookshelf system.

Maybe your SBs are revealing harshness in the recording/mixing?

Geoff

[DeZZar]

It's sort of a similar story on my current two way floor standers, same tweeter. Most music sounds great, including a lot of rock, but then sometimes you play something that sounds pretty good on the sonos for instance, but drills a hole in your head on the big speakers. It's odd, you sort of just get hit with this wall of trebble that magically makes all the bass disappear leaving you wondering where all the warmth and depth went. But you look back at the measurements and its running a 5db downwards til from 100hz to the top end without any nasty peaks. Theoretically 100hz is the loudest point. But the tsunami of trebble from a lot of metal tracks just completely takes over. It's something I'm hoping to improve with version 2 of those. But maybe I'll try another tweeter as well. It'll take some experimenting. Should take a measurement of the sonos one day ...I'd love to see what frequency response it is that makes everything generally sound fine.

Maybe I'll pick out a few tracks and get some of you guys to play them and tell me what you hear. I might just be getting really fussy.

[rpb]

What stuffing is being used to absorb the woofer back wave?

Is the impedance "normal" above 20k? Any chance the amplifier is doing something like overloading? Maybe the impedance is a difficult load.

I may take a stab at a sim tonight. What xo point do you think is optimum?

How does the woofer measure at 45 off axis?

[DeZZar]

Hey mate, the entire cab has felt lining on all walls except baffle. Top section is fully stuffed with acoustic damping material (i.e. acrylic stuffing) and directly behind the woofer is a layer of the acoustic stuffing as well (it's about 2" thick). I attached out to 60 degrees for the woofer and tweeter below. It measures as you'd expect...nothing unusual from what I can see.

With the last crossover under test Impedance is min 3.4 at 200hz, and does peak at the xo point (30) and has a kind of rounded hump as it heads to 20k sitting around the 10ohm region.

[rpb]

Unless I messed up, the posted tweeter frd file is incorrect. I messed up ! File is fine.

[Wolf]

2.3-2.5K should be better than 2kHz, considering the HD spectrum of that tweeter.

I'm wondering if it's not the polar response of the tweeter. It's a ring-radiator, and off-axis suffers quite a bit. There might be a peak off-axis. However, the polar response of the same tweeter in the Missing Link application does not have a peaking response. It just rolls off in both vertical and horizontal. I really don't think it's the tweeter.

The question is whether the elliptical suppresses the breakup to -25dB or greater. If it is not suppressed enough, it can be the woofer and not the tweeter at all, even if both the tweeter and woofer breakup are seemingly well below the level of the bass range. The diameter is a 17cm, so as long as you're not xovering at 3k it shouldn't beam badly either.

Maybe I'll take a crack at this...
Wolf

[DeZZar]

Here's 0-60 on the tweeter. I have all the way to 90, but upload restrictions....

I usually cross this tweeter under 2K. In a two way with a 6inch plus woofer I just cant see it working any higher. 1.5K the tweeter has 1.3% THD (my noise floor isn't great BTW) @ 2.83v. 2.5K is only a little less at 1.1%. I agree Wolf, that it would be ideal to go a little higher as things do rapidly deteriorate under 1.5K.

[DeZZar]

And here's the 0-60 on the woofer. These have all been merged with the woofer+port combined nearfield.

[DeZZar]

This is the difference in breakup suppression between a 2nd order filter (with tank notch) and the elliptical. The latter suppresses the first breakup node to -50db. However even the second order simpler network gets that first breakup (4.5K) to -25db and the rest to under -38db.

[Wolf]

I think you need to go 3rd order on the woofer. This will bring up the dipping range as it can give you a better blend at the knee before rolloff without scooping out the FR.
Does the Elliptical include the tank as well? I would try with both present too. I consider -40dB to be important, and -25dB to be bare minimum.

Not saying the tweeter can't go to 1.8Khz, and I've seen several designs that do this. However, the rising HD that I experienced when Coop and I measured them got ugly pretty quick below there. It is just plausible that moderate outputs and using it lower, as in metal compositions, could likely make it shrill. That response below the tweeter being drooped might make the treble stick out like a sore thumb.

Wolf

[rpb]

Here's what I came up with. Used .025 for z offset.

[DeZZar]

Thanks mate - I'm unfamiliar with the spreadsheet so forgive me. Tweeter starts with 3.5R, followed by third order 6.96uf, 0.27mh, 8.32uf with an 8ohm+2.4ohm Lpad?

The woofer has third order as well? What's the value for L2? 18uf & 0.1mh for the rest of the third order and then a series notch 1mh, 50uf and 3.5ohm?

[rpb]

Tweeter is a damped 3rd order. See the 1 ohm resistor? I forgot to round up values, so adjust as needed.

The woofer is damped 3rd order with a tank on the first larger coil. (.4mH, and 2.4uf) It's moved to line 55 in order to add the cap in parallel. That's why L2 is empty.

[rpb]

Here's a lower xo version.

[DeZZar]

This is the second order that produced the above result (post #32).



And this was the elliptical (?) I tried out which produced the best result on the woofer (-50db on all breakup) but needs more work getting the tweeter aligned.

[augerpro]

Are you sure it's the tweeter misbehaving? Try a resonance trap, some tweeters just really don't tolerate the Fs getting tickled. Try 2.0 mH, 22 uF, 2.0 ohm trap on one speaker and see if that makes a difference. Model it first, that is for a ScanSpeak 9130 but the Fs is similar to the SB29 you're using and should get you close.

Off axis modeling is essential, off axis flares in the 2khz area can really make for a hard/harsh sounding speaker.