[Anthony]

Hey all,
Just wanted to update you all on my center channel build.

For reference, it's a 3 way design using Dayton RS 8" woofers, HiVi B3N mids, and a BG Neo3PDR tweeter with an open back.

Crossover points are currently 500 Hz (2nd L-R), and 3kHz (4th L-R) on a 12" tall baffle with wings of 4" on top and 6" on bottom. MTM section is vertical D'appolito configuration flanked by the woofers (sideways speaker).

It's been in my system a while now and honestly, I like the sound for dialog only. It is a tad resonant, due mostly to a peak-null-peak in the high bass, low mid region. However, for any sort of music, the weaknesses show. It always sounds like it is holding something back and upper vocals seem to alternate between damped and shrill.

I liken this to the bad baffle design. I am a firm believer that this is the trade of all open baffle designs: what is a good baffle for bass is awful for mids and vice versa. The tweeter seems unphased by the baffle design, as it is pretty flat baffle response from 4k on up.

What I seem to be hearing is a peak right before the woofer hands off to the mids, then the mids have some alternating peak-null-peak throughout their region. If a note hits on a peak, it sounds shrill, in a null it sounds dead.

So, this got me thinking, what if I had two baffles, one for the bass section and one for the MTM. Make the MTM baffle as narrow as possible to eliminate the sine-wave respnose effect of the large baffle, and have a large baffle for the woofs. Essentially, it becomes a sort of "dog bone" configuration for the mains, although in a center, this would look considerably weirder (with the MTM section vertical).

So I have a test baffle to try this out, but haven't finished cutting it out or wiring anything up yet. I feel good about this approach, though.

So pics to come once I get this thing in the test phase.

Anthony

 

[BoomieMCT]

So I have a test baffle to try this out, but haven't finished cutting it out or wiring anything up yet. I feel good about this approach, though.

So pics to come once I get this thing in the test phase.

Anthony

As someone who visits your house, eats your food and listens to your speakers - get your rear in gear and git er done!

Also, just for grins and giggles, let's do a FR outside or somewhere else in the room. Even though you are measuring fairly nearfield I still feel like there could be some room effects in there.

 

[Anthony]

The results have been matching up too close to Edge simulations to be room effects. Besides, I have Magnepans to compare to, and they do not have these FR discrepancies.

I cleaned my shop during the weekend, so I should be able to get some actual cutting done this week.

 

[BoomieMCT]

The results have been matching up too close to Edge simulations to be room effects. Besides, I have Magnepans to compare to, and they do not have these FR discrepancies.

I cleaned my shop during the weekend, so I should be able to get some actual cutting done this week.

Post pics of what you are talking about pls.

 

[Anthony]

First off -- no pics yet. My digital camera died under a known recall, so I will have to wait for the repairs to get pics.

But, we have a winner! I completed the crossover last night and did some testing. Reasonably flat response (given suboptimal room to measure in), but most importantly, sounds great!

The details:
Dayton RS 8" shielded woofers, crossed at 500 Hz, 2nd order LR (see more below)
TangBand 3" Bamboo midrange, crossed at 500 Hz, 2nd order LR low, 3.5kHz 4th order LR high.
BG Neo3PDR in dipole mode tweeter crossed at 3.5kHz, 4th order LR (3rd order electrical with a big notch filter for the top octave)

In room extension down to 65 Hz, although they were designed for 80 Hz, so I'm happy there.

Design ended up being completely passive with no padding. The B3N's just weren't efficient enough, so the Tangbands came through in a big way.

The woofer was way too sensitive, but I had to borrow a trick from Adire to get it matched properly: use a low pass crossover tuned to a much lower frequency in order to basically trade efficiency for low end extension. For my particular baffle, this meant tuning the crossover below 200 Hz for a 500 Hz target. It ended up being something like 5.3mH inductor and a 50 uF capacitor, but it hits the 500Hz, 2nd LR slope almost exactly.

Everything balacned out so well, that the only resistor in the entire crossover is in the tweeter notch filter.

30 degree off axis was spot on, except for a 3db rolloff at 18kHz instead of 20khz. My room was not big enough to test 60 degrees.

The problems:
Cheap pine baffle is still putting some coloration (null) at 1khz, which corresponds to the baffle vertical "width" (height for a center channel). Final will be either baltic birch or veneered MDF with more bracing, so I imagine this problem will disappear.

Impedance drops a tad too low for my taste at 1kHz and around 8kHz. 2.2 Ohms at the lowest, but for an average, the speaker sits right around 4 Ohms across the frequency range. I'm not sure if I should worry about this or not.

I do not have a Zobel on the mids or woofers, so I may try that as well (although I don't think that will help my impedance issue).

Finally, these will end up being expensive speakers. $50 crossover, easily, $170 in drivers, + baffle. So probably $300 a piece when all is said and done. Not awful, but it may scare some people off.

FR plots and pics to come, as soon as I get my camera fixed and get off my lazy butt.

Now to build the mains . . .

 

[BoomieMCT]

I was over at Anthony's yesterday to hear the "final" setup. He's not joking - this center is *nice*. Very flat, relaxed and natural sound. Excellent dispersion. The clarity on dialog in all seats in the theater (two tiered couches with a riser) is top notch. The critical mid-bass is quick, tight and seemingly effortless. I dare say this speaker put his fancy pants Magnepans to shame.

I want to emphasize the elegance of the crossover. Getting a design like this to work without having to use any padding is no mean feat. Considering this is a 1) dipole, 2) W(MTM)W and 3) center channel this isn't an easy project to perfect. Anthony has done a great job getting the baffle, crossover and drivers to all mesh together properly. I look forward to seeing what the final product will look like.

 

[Anthony]

Here is the offending impedance plot if anyone wants to comment. I'm not sure if I should be concerned and if so, what to do about it (because I really like the crossover the way it is now )