Gavin's DIY Home Theatre (Hardcore)

[Wechnivag]

That roughly closes the chapter on the Fi IB ULF subs. I think most of us here are pretty comfortable with the idea that different sub drivers have different sound character.

What i have surmised after lots of experimentation with different drivers, and studying T/S parameters, is that there are roughly 3 types of woofers or subwoofers.

First, the ULF drivers, true subs. These are the UM 15/18, JTR, the Stereo integrity, the Harbottles. The Fi IB318. They excel in low bass and ULF, but their handling of mid and upper bass, the speed transients etc often are not as good.

Next the Balanced subs. I was surprised how often i see this particular set of parameter optimisation, across seemingly vastly different subs. Examples are the the Peerless XXLS 12, Eminence Lab 12, JL 12W7, 13W7. Lots of 12 inch subwoofer drivers from Scan Speak and SB Acoustics fall into this category. I use this set for the woofers in my Mains channels.
The bass sound, as the name suggests, is balanced between low bass weight, and midbass speed. With quality drivers, they are very very satisfying and most will be very happy with these subs. The exception being the hardcore bass lovers who crave that ULF.

Finally, the midbass drivers. These are usually pro subs from B&C, etc. There are some very unique pro sub drivers that approach the balanced sub driver characteristic. These subs excel in midbass attack and transients. They hit hard. Too hard, for some. Just nice, for others.

Here’s a sim, showing how their impedance peak (resonant frequency) affects how they draw power from the amp. For a nominal 100W input, at the impedance peak frequency, they draw as little as 10W over a broad frequency range.

The resonant frequency are resultant in-cabinet, whether sealed or vented. These are sealed, for simplicity.
The ULF subs will be in the 30-40hz range, for the weight. The balanced subs are in the 40-50hz range. Nearer to 40hz, better weight. Nearer to 50hz, faster speed and transients. From what i could tell, the MK sub drivers are very close to 50hz.

So what i did was to augment the LFE ULF drivers with a pair of 21 in pro subs, namely the SB Audience Nero 21. They are available from the local distributor for a very attractive price. Plus there is hardly any shipping cost, vs ordering from overseas.

https://www.sbaudience.com/index.php/products/subwoofers/nero-21sw1100d/

Here’s a sim of 8x18, and 2x21 all driven with the same nominal 100W. The pair of Nero 21 hits 115dB at 60hz, with 100W. Actually the 8x18 are already at almost 118dB, but supposedly the bass character is different. Anyway, i wasn’t taking any chances.

Here we can also see the inherent differences in the drivers. All things equal, 8 drivers has 6dB more output than a pair. At 30Hz, the difference is max, with about 7dB more from Fi than from Nero. But the mid and upper bass is where these pro subs really shine.

If you subscribe to this school of reasoning, PSA and their current range of subs using custom DVC B&C 18DS115 and 21DS115 pro subs deliver the goods.

So, i ended up with the LFE system of 10 drivers, 8x 18 and 2x21. The other benefit of the baffle wall SBA, is the delays and setup are trivial. They are all on a single channel of the SP1-6000, wired into an approximately 2 ohm load, which is ideal for full power from the amp.

[Wechnivag]

The 21 in drivers are very impressive when i unboxed them. They weigh 21kg each, vs the 14kg of the Fi IB318.

It has a massive 5 inch voice coil

My laptop for scale.

12in sub for scale, testing the baffle fit.

BTW, the 18in woofers on the bottom of my current mains, are the Nero 18, smaller sibling of the Nero 21. They share the same motor structure, so the 18 has serious drive too.

[sammy]

Awesome. I love how you left enough space to be able to get behind the baffle wall to help install stuff. Also the fact that the subs are detached from the structure and sit independently on the sorbothane mounts. I only recently elevated my sub onto some elastomer feet from SVS and that removed a lot of the vibration around the house. However, the rattles were fun from the rooms perspective.

[Wechnivag]

Hehe, i will outline my sorbothane experiments in the Center channel description later.

Rattles, it is mainly coming from the aircon housing… it is mounted on the wall, which is of a compliant construction and moves quite a bit. More on that during the room construction section… i will add some foam stripping to try and mitigate this eventually. Need to figure out how to remove the A/C covers…

The other noise i’m getting seems to be some baffle wall wood-sound creaking from the amount of flex due to the pressures generated by the IB… this was somewhat unexpected, as in i knew the wooden baffle wall would flex, but i didn’t expect it to make creaking sounds hahaha. Not sure what the fix might be short of an extensive bracing retrofit behind the structure.

[Wechnivag]

I originally planned for the nero 21 to go into ~80L sealed cabinets in the baffle wall. Then I read a comment from a IB owner. He has the same low bass, midbass IB setup. Compared the midbass in sealed cabs with IB and found the IB midbass to sound much better.
This decision change came about very late, as I was in the design and planning for the mounting for the Nero 21.
They go into a double layer 1inch thick baffle. Which isn’t much considering how little baffle ‘meat’ is left after cutting the holes.

There’s not much behind it but some framing to support the assembly, sitting on sorbothane pads. Here’s a test fit, after cutting both holes and building the mounting structure. Center channel work also in progress.

Edit: Found another photo before the lift up. This one was also a 2 person job due to tight clearance, and needing to sit exactly on the sorbothane pads.

Midbass next to low bass

[Wechnivag]

For the Nero 21, there is no opposed force cancellation. So vibration may be a significant issue. The amount of reaction forces and motion is determined by the ratio of masses. The cone moves forward, and everything else (frame, cabinet) moves backwards.

The Nero 21 cone is 300g. When it has a 10mm excursion one way, everything else has to move backwards according to Newtons laws. How much it moves backwards depends on the mass. If it is 3kg, vs the 300g, it moves 1/10 or 1mm backwards. If it is 30kg, it moves 0.1mm, which seems a reasonable level. The sorbothane mounts should handle that.
The driver itself is 21kg. With the 50mm baffle, frame structure, I added about 6kg of steel shots to each assembly for good measure.

Steel shots further dissipate vibration energy by a mechanism known as Particle Impact Damping - PID for short. There’s another thread here somewhere on it. It works extremely well for subs and vibration control, without having to resort to insane masses to counter vibration.

Particle damping

[Wechnivag]

Next, is the center channel. We always hear about how the center channel is the most important channel in the HT, anchoring the dialogue and some say over 90% of the soundtrack.

Realistically, in terms of setup, budget allocation etc, it is often a significant compromise due to space and layout constraints. And maybe centers are just not as sexy from an audiophile gear perspective. It often ends up the (literally) middle child, with less attention that it deserves.

For projector users, it is a major early decision and most go with a non AT screen, which significantly limits the options for center. From previous experience, and I guess everyone who has been to the cinema can tell, that center behind an AT screen gives the most natural and immersive presentation.

For this build, I wanted it to be a no holds barred approach. The sky is the limit. Where do we start?

It is a 2.5 way, with a compression driver in a large horn, a 12in midbass and dual 12in woofers bringing up the low end 0.5 way.

But actually, there is a constraint. I bought 3 sets of the JBL 2384 horns a long time ago, back in 2017. And I have been itching to use them…
These are from the JBL Cinema 4722 speakers, there are quite a number of HT builds with them. Including 2 owners who are in SG. *wink

With this being a JBL horn, naturally I had to use a JBL compression driver. This was an easy decision, the D2430K driver, used in the M2, the 4367 and most recently the flagship SCL 1 in the Synthesis range.

SCL-1 | 2-Way Dual 12-inch (300mm) Custom LCR Loudspeaker

[desray]

I can sense the excitement in the months ahead.

[Wechnivag]

Actually, with the D2430K, the natural choice is to use the M2 waveguide, and the 2216ND 15in woofer. Both are available to the diyer, effectively making it a M2 clone. In fact, my fellow kakis bought the WG and woofer.

I used the big 2384 horn because I already have it. Also, there is some differences in the directivity control. Naturally the modern design M2 wg is technically superior, and with the modular design, I can still switch to it if the 2384 doesn’t work out.
I just stubbornly want to use the 2384, ain’t it so badass, such a massive horn. It is also more fun, to (foolishly) go where no one else has gone before.

Regarding the 2216ND, it is a supremely competent driver. From my experience, a well sorted 2 way horn loaded speaker has coherence that is unmatched by a 3 way, especially in the bass.
The down side is it wants a large cabinet. At least 150L ported. Better if 200L. It would also be ported, and I preferred everything sealed in this setup.

So an easy choice would have been a M2 clone setup for LCR. I just itchy backside want to try it a different way.

[kennyluck2000]

Love your attitude, never try never know

[Wechnivag]

I choose to use a 12in midbass (vs 15in), and dual 12in woofers, to keep the baffle and height dimensions manageable, and the cabinet volumes manageable. The horn TW level is at 1.1m, with the design axis just below it at 1m. Right about seated ear height.

The cabs are ~50L sealed, about 15kg each. Loaded with drivers, about 20-25kg each. Still very manageable from an handling perspective. I faint everytime I see those 200 or 300 L ported cabs on Avsforum.

With my exhaustive T/S parameter screening, I ended up with a list of candidates for the all important midbass /midrange driver. Not all pro drivers are made equal. Some are more musical than others, while some are just built like tanks to take kilowatts of power, rain and other abuse for years. Sound quality not usually high on the requirements list.

The winner is a Faital Pro 12FH520 neodymium midbass.

12FH520 (8Ω) by FaitalPRO

The woofers are dual SB Acoustics SB34NRXL75-6. Very very very nice woofers. With shipping cost savings from the singapore dealer, it is an unbeatable value.

https://sbacoustics.com/product/12-sb34nrxl75-8-norex/

I love it when a driver has such a large motor that you can’t quite judge its size, based on the proportions. Did I mention it has a large motor?

[Wechnivag]

A little insight into the madness of the midbass selection process…

I have a few criteria.

1. High Qms/low Rms for low loss suspension. This gives a very intimate sounding driver that excels at low level details and micro dynamics.
2. High motor force BL relative to the driver’s moving mass. BL2/Re is a common measure among subwoofer drivers for the strength of motor. However, to get a strong motor, one needs a large voice coil and this increases the moving mass. This is a fine balancing act.
3. Shorting ring in the motor. This is straightforward, surprisingly this is relatively uncommon. More than half the drivers in the shortlist do not have shorting ring, and in the end this was one of the major deciding factor.

I also have a pair of the SB Audience Rosso 12 on hand, it could be the twin to the Faital except for the lack of shorting rings.
Eminence Kappalite 3012HO would have been a no brainer choice, have heard plenty of eminence drivers, and they always have that x factor and sound great. Lack of a shorting ring has never stopped them, especially the neo motor ones.
If i just focused on the strongest motor spec, it could easily have been the 18 Sound 12MB1000, or the RCF L12P110K. But they both have higher loss suspension…

It was still a gamble, the final sound also depends strongly on the paper cone geometry and breakup. So far, no complaints on the sound of the 12FH520. It does its job well and hasn’t called attention to itself for the wrong reasons.

[rock123]

@Wechnivag super hardcore there! Lots of info sia cheers. Never try never know !

[Wechnivag]

The woofer went thru a similar selection process. Here’s a quick look at the candidates in case anyone is looking for nice woofers.

Looking at the list now, looks like i prioritized low loss suspension as the single most important criteria. More important than having an uber strong motor. Power is nothing without control, eh?

The ones in Yellow are the other drivers that i also have on hand, for various reasons…
i have 4x NS12, 4x XXLS12, and 8x L26ROYs… These will eventually be deployed for the other channels. More on this later.

Major benefit of using large, high sensitivity drivers is the system efficiency. Doesn’t take much power to hit target SPLs.

At nominal 10W, the midrange is at 101+dB at 100hz, going up to 106dB in the upper ranges. The pair of woofers is at 103dB at 100hz. For comparison, i have simmed a SB satori MW16 in 20L sealed at the same 10W. It has about 96dB at 100Hz, going to 98dB in the upper ranges.

Actually that is no slouch from the 6.5 in satori. It is however already at its limits, and will likely sound strained at 20W (hits XMAX) Paired with the right woofers for bass, it can definitely get the job done for HT. But i think for the purposes of my build, we are comparing apples and oranges.

Sidenote:
Actually, i wonder about using amps with 150W, 200W or the purifi and hypex with 300- 500W for regular speakers. For the majority of content, even with hifi speaker (not high efficiency systems), at ~70dB average levels (which is actually loud), we are technically at 0.1 W average…

(1W = 87dB. 10W = 97dB, 0.1W = 77dB, 0.01W = 67dB, 1miliwatt = 57dB )

For the more nuanced sounds, they are easily in the 50dB, or 60dB range, which means they are being driven by miliwatts of power from the amp.

Whatever improvements we are hearing with a 500W amp vs a 150W amp, it is definitely not at 500W levels… But we digress…

[Wechnivag]

hahaha, did i forget to mention that this is a hardcore build

Updated the title

[Wechnivag]

Coming back to the JBL 2216ND used in the M2.

In Green, 80L sealed, 10W. 1x 15 actually more than matches my selection of 3x12. Maybe i have really been foolish.

However, the other part of the story is in the impedance load, how it draws power from the amp.

Those large impedance peaks (from low loss suspensions, strong motors) means the driver is extremely efficient at its resonant peak. At 10W nominal, the midrange is drawing 0.5W between 70-80hz. I believe this is related to the magic of how it handles midbass across that range. Similarly, the woofer (red) has its magic between 40 and 50Hz.

Subjectively, these characteristics, together with their overall high efficiency is aimed at delivering bass, midbass and upper bass that is detailed, nuanced, tight, dynamic. Strong motors with lower moving masses will stop all motion after the signal is gone. No overhang or slop.

These are all characteristics that many crave, but actually have no idea how to achieve.

I may be dead wrong with my theory and reasoning, and all these mean absolutely nothing… hahahaha

The JBL 2216ND, is a supremely well balanced driver, that is optimised to do everything well. From this measure, it supposedly has less magic than the drivers i selected. The differences are minimal, and may not actually be noticeable/ audible.

But that’s the set of drivers i have chosen for the Center.

Thanks for those who have stayed with me through the long stories. We can now move on to the cabinet design and construction for the center.

[Wechnivag]

Will try to be a bit less long winded in the next sections.

The cabs are sealed built from 1 inch regular plywood. The part I focused on was the bracing design.

A closer look at the internal bracing, before the front baffle was added.

Seems relatively straightforward, done?

Haha, there’s probably more. So I was really inspired by kef’s LS50 white paper where the showed the effect of their damped bracing on the cabinet. Apologies for the random link. Kef seemed to have stopped hosting the paper?

ls50-white-paper.pdf

This diagram shows the effect of bracing (red) on and undamped cabinet (blue). The bracing merely moves the cabinet vibration peak output to a higher frequency, without reducing the magnitude.

Then Kef shows the effect of damping on the panel vibration output.

The damping compound (red) is applied between the cabinet panels and a regular cross bracing.

Anyone who has done the knuckle rap test on the LS50 can attest to its effectiveness.

Since i have 3 cabinets to build, 2 woofers and 1 midrange, i knew i had 2 tries with different types of damping compound, before selecting the best one for the midrange, which is most critical.

The first one, I tried white acrylic sealant. This dries a bit too stiff , so it has minimal damping. Still probably better than a completely stiff joint with wood glue , some damping will reduce the peak of resonances.

Next I tried some bitumen strips. Bitumen is a compound commonly used damping, eg on automotive panels.

I apply a strip between the bracing and the sidewalls.

Finally, I tried silicone ‘magic tape’ as the damping compound.

Silicone tape on the bracing before attaching back wall.

Actually there’s a saying - big speakers have big problems. Cabinet panel resonance control is definitely one of them. This was a really interesting part of the build that I really enjoyed.

[Wechnivag]

With the center speaker mounted on the same baffle wall structure as the monstrous LFE system, I was very concerned about vibration transfer and effects on the center channel.

Research suggest sorbothane to be very effective. Some quick calculations for the size, based on the durometer rating of the sorbothane, and the mass of my speakers.

Those 4 pads per speaker went between the cabinet and the supporting frame structure that is attached to the baffle wall.

The wooden frame structure that the center speaker cabs will load into.

Using the phone accelerometer app, I monitored the vibration readings on the speaker cabinet, when it is isolated on the sorbothane pads. It is remarkably effective.

Center isolation test 1

Another view. Apparently videos in portraits default to Shorts in YouTube?
Anyway, I had to bash quite hard against the wall structure to register a faint trace on accelerometer app, with the sorbothane isolation.

Color me extremely impressed…

[Wechnivag]

As shown in the into photo, my 2384 horns came with a nice stand. However, I had to modify it to get the right position sitting inside the baffle wall cutout. I also added automotive bitumen damping on the horn walls, and a wood frame around the perimeter of the horn mouth, with damping strips sandwiched in between. Also we see the JBL D2 driver mounted on the horn. It’s small and compact, another reason I like it. Some JBL monster compression drivers are like 3 or 5kg…need a whole mounting system designed for it.

Did I mention damping is important? Also probably a good opportunity to differentiate between damping and dampening.

Damping is the technical term when used to describe vibrations and oscillations. The car shock absorbers are called dampers. We add damping to systems that we want to control vibrations.

Dampening, on the other hand, is to make wet…

Anyways, a quick fit check of the horn in the cutout. Note the very little clearance between the speaker cabinet and the baffle wall…

Once that was OK, it was time to cut into the baffle wall for the other speakers. Mark out the dimensions. Also cutting it very close to the upper right manifold!..

This was a stressful and dusty occasion. I had little helpers to hold the vacuum cleaner nozzle.

[desray]

wow bro, you in sabbatical leave ah? Every day can see your posts.