Modula NeoD CC build for fireplace mantel

[Scareurpasenger]

I am making progress on the subwoofer and learned a decent amount from that build. Moving on the weakest link in my system really is the center channel. I have an old Martin Logan low end center channel that likes to kill amplifiers..
HTguide member Concilian was kind enough to sell me the drivers and some parts to take this on. Thanks!

Some design goals:

• Fits below the TV on the fireplace mantel (~11" deep)
• Sounds great with clear dialog
• Integrates well with subwoofer and other channels

I did some initial modeling and decided it was best to get some input before building this time around.

The challenges I see so far is that port lengths may be a challenge to fit in at 11" deep outside dimension, that doesn't leave much room. So, is sealed the way to go since I have a sub? How shallow of an enclosure will work?
BSC or No BSC I assume no BSC since it is so close to the wall and on the mantel below the TV.

I used Zaphs driver specs to start with and will attempt to measure the actual units. I am not getting the same results Jon got from the original NeoD CC thread.


Being new to all this, I am not sure what QTC to choose. .6 makes it a huge enclosure!


Where did I go wrong..

[Bear]

The challenges I see so far is that port lengths may be a challenge to fit in at 11" deep outside dimension, that doesn't leave much room. So, is sealed the way to go since I have a sub? How shallow of an enclosure will work?

Sealed will give you the smallest box, dependent upon the target Qtc. There are three sealed alignments to pay attention to: Butterworth, Bessel, and Critically Damped. Plug '=1/sqrt(2)' for the wanted Qtc, then hit Start. Your target box volume should drop a bit as it gets increased.

Butterworth is 0.707 (1/sqrt(2)), Bessel is 0.577 (1/sqrt(3), and Critically Damped is 0.5. Butterworth is nominally the flattest frequency response. Critically Damped nominally has the fastest settle time, and Bessel has the minimal impact on phase within the passband. The optimizer utility embedded in Unibox may not yield a result, depending upon the driver and Qtc target (try an Accuton C79 and a critically damped target...). Also, because Unibox relies heavily on macros, the standard tools in Excel don't work (e.g., Goal Seek, Solver). One can crack open the VB module, but that has always seemed a bit rude to me.

One more quick recommended change: check the box next to FR affected by Le, Le2, Re2. That will show you how inductance in the motor acts as a low pass filter.

I am not getting the same results Jon got from the original NeoD CC thread.

Build version matters a lot. The D26 came in and out of availability, and some iterations were very different. The RS52 of today doesn't seem to match the 2010 version. It would make sense that HiVi has tinkered with their recipe, as well. It could also be operator error (this is DIY, after all). Be sure to run each woofer for a fair bit of time (many recommend 24 hours at 5.6+ volts) in order to break-in the suspension. Also, trace the crossover to add-up all of the series resistance to the woofer. Plug that into the Rs field. That will give you a more realistic perspective on the required volume. It also demonstrates why low DCR inductors are often a Good Thing.

Where did I go wrong..

There's nothing wrong unless you've infected your PC with a malicious virus buried in the macros. If you got Unibox from an email message from someone you don't know with a subject line along the lines of "Bild Speakrz Now!", then that was probably a mistake...

For your ported model, click on the "Include effect of port resonance" checkbox. You will then start to see how big ports make unwanted noise. Start by walking your turning frequency (Fb) towards the F3 frequency. Then play around with the box volume a bit. Monitor how the port length changes.

Also flip over to the Vented worksheet and look at the Cone Excursion graph and the Port Speed graph. Bigger ports have less chuffing (what typically happens when the speed goes above the red line). That allows you to play with the tradeoffs. Remember that port resonance check box, too.

Check out the excursion graph. You'll see the driver excursion plotted. Unibox makes a horizontal line equal to the driver's Xmax. Play with the applied power to have the top of the hump just reach Xmax. The look left. The excursion curve goes up quickly after the fall off from the hump. That will tell you where a limiter would be a good idea or a crossover to subwoofer. You can also play with Xmech instead, but that is a hard-to-get statistic.

There are a bunch of supplementary calculations that I've added to my version. And the concept of power modeling is really good for sealed enclosures, too. If you start playing with the PR section, then both Dayton and ScanSpeak publish data that facilitates modeling in Unibox. SEAS? Not so much.

That should be enough to get you going.

[chrisn]

No BSC is usually for in wall designs. I started reading through the original thread to find more info on it but got sidetracked. That center channel looks sufficient sealed or vented, especially with room gain and a sub.

[Scareurpasenger]

No BSC is usually for in wall designs. I started reading through the original thread to find more info on it but got sidetracked. That center channel looks sufficient sealed or vented, especially with room gain and a sub.

Found the answer in the original thread. With this being close to a rack type setup, the original crossover with its low bsc design should be fine. The dsp in the receiver will compensate as well.

[Bear]

I plugged in your numbers, and it looks like you've got more decimals than Unibox is set to show. Here's what a Butterworth sealed alignment looks like:



Note some of the extra performance metrics I pulled out of the underlying chart data. In 18L, they don't exceed Xmax until 45Hz, and even then, it's not by much. In other words, this is a pretty safe alignment when handing over to a subwoofer. If you want to crank it up, and excluding insertion losses beyond the 0.2 ohms, these can handle 100W with a 70 - 80 Hz Fc. At 200W, you need to be rolling off of them to a subwoofer no lower than about 95 Hz (assuming -6dB at Fc).

VENTED:
Here's what the vented alignment looks like, keeping your T/S parameters, but ticking a few more boxes...



Things to note:

• A 2 inch port has a resonance above 1kHz. That's workable.
• Fb and F3 converge at 42 Hz, but you've got an Xmax of almost 11mm.
• The port volume is 0.6 liters. That's ID volume. You will need to add the outer diameter volume to the modeled box volume to get your theoretical total enclosed volume for the woofers.
• I did change some of the box/port assumptions.

Because it's vented, you generally need to check the Vented worksheet to see the details.



Excursion isn't a problem with 60W until you're in the mid-30s. These could be used full-range, but not for action movies. Let's check the port speed. Unibox wanted a port diameter of 7cm, so a 2" / 5cm port is small by its calculations.



Houston, we have a problem. This thing will huff and puff like it's a wolf wanting fresh pork. Air speed is very much affected by how much the cones are moving, which is a function of applied power and frequency. So, one needs to look at using a larger port (which will take up more volume - which is additive to the box volume) and which will generate a more pronounced resonance at a lower frequency. Or, one can reduce the power handling significantly (a bit less than 20W looks manageable).

It's all about managing tradeoffs.

[Scareurpasenger]

Thanks Bear your responses have been very helpful! Looks like the last 3 attachments didn't load though.

It is very tempting to build a cabinet that can be adjusted to the different alignments just to hear the difference. If only I had the spare time!

[Scareurpasenger]

So, not trusting the data on the drivers. I pulled them out as I should have and hooked them up to DATS v3.

This is the updated sealed model with calculated VAS from the MMS.

[Bear]

Images are now reloaded. Hopefully they stick around. When you see the "Graphs not updated, OBS!" message, click the update button. Otherwise, it seems like you're good to go as far as Unibox goes. Your sensitivity measurement looks incredulously low. I can't imagine Jon putting up with something with a raw sensitivity much below 80.

[Bear]

Thanks Bear your responses have been very helpful! Looks like the last 3 attachments didn't load though.

It is very tempting to build a cabinet that can be adjusted to the different alignments just to hear the difference. If only I had the spare time!

This where some dark magic comes into play called a Linkwitz Transform. This will be easiest to do with something like a miniDSP and a big AMP. You never get around piston volume limitations and power requirements, but an LT can make a Chebychev (Qtc = 1.0) act like it's critically damped -- just at a much lower SPL. TANSTAAFL

[chrisn]

It looks like VAS is .9 L? Have you broken in or flexed the woofers? Those woofers are lower sensitivity iirc, but 2 in parallel with reduced BSC should be fine.

[Scareurpasenger]

I got FS down to 58.8 after running them for an hour. I have not flexed them by hand yet.

[chrisn]

FS and QTS are fairly close to what Zaph's measurements show from 2006. Did VAS calculate as .9L or is the decimal in the wrong spot?

[Scareurpasenger]

It calculated to .9l in dats. I may try a different method just to be sure.

[Bear]

It calculated to .9l in dats. I may try a different method just to be sure.

I didn't catch that one. But VAS does affect sensitivity (you can trace the math in Unibox), and even something like a SS 10F has a higher Vas specification (2L).

[Scareurpasenger]

I found the issue and it is looking much better. I may try the added mass method as well just to be safe. I have seen the PE video on it and they measure with the cone faced up and have read that it shouldn't be done with the cone face up. I find it hard to believe PE would put it out there if it was incorrect. Also, wouldn't adding mass and measuring horizontal cause potential misalignment with the speaker?

[Scareurpasenger]

This is with the added mass method. Peak cone excursion is at 80 watts and 81hz when going for QTC of .707.

[Bear]

That looks much more reasonable. I assume that the 80Hz number is well below the red line at 45 degrees? If so, then you have a viable solution. If you start decreasing the box volume, then Qtc will go up. Basically, Frequency response will start to roll off pretty quickly below Fb. The higher the Qtc, the faster the rolloff. You will also start to see some ripple show up in the passband.

https://en.wikipedia.org/wiki/Chebyshev_filter

[Scareurpasenger]

There isn't that much room but considering that it should be fine at regular listening levels, I am thinking it should be an ok size. At the same time, the measurements seem to be changing a bit as I run them and I may try to break them in a bit before completing the model.

I am seeing a trend that most ported designs show over exclusion at lower levels. I assume these are not a concern due to the volume levels used and/or the recordings not going that low?

[Scareurpasenger]

So, I think this is where it will land based on break-in. Vented has more extension and is tempting.

[Bear]

Your closed box model has the SPL intersect the Xmax curve right at 80Hz. Since most subwoofer crossover slopes are LR4 (-24dB/octave; -6 dB @ Fc), you should be just fine blasting your ears at 108dB.

In the Vented model, you can clearly see what happens when the cone unloads below Fb. One either needs to be very cautious about content below ~40Hz (no explosions, no deep organ pipe music), or you are back to the subwoofer crossover. Since a Vented alignment is typically much larger than sealed, the additional extension only matters if you want to run full-range. There are opinions about what happens in the upper bass area for midwoofers designed to be vented but which are in closed boxes, but that's a discussion unto itself.

[Scareurpasenger]

The vented modula MTM seems to have the same issue no? Or is there a frequency at which this is no longer a concern?

[Bear]

Take a look at something like the SW223BD02 in ~30L tuned to ~20 Hz. Or the L26ROY in about 50 - 60 liters. You'll see what the power handling looks like in a vented enclosure, but then look at the port resonance vs air speed for any given ID of vent. TANSTAAFL. That's why many of these subwoofers that can also be woofers do best with a PR alignment (e.g., SW223BD03 in 30L with a RSS265PR at 425g - 500g).

[Scareurpasenger]

Crossover parts have been ordered. I will remeasure the drivers since I found an odd bug in the 3.11 version of DATS. One of my rs150-8 drivers was not responding to getting a FS measurement until I updated to 3.15.