Fun! First real sub-build - two JL 12w6v2s. For SQ and LOW extension

[noneYet]

First off - I am not asking for help, yet. I feel that I have to do more work on my own before I have any right to bug you who already know stuff - I know that there are a lot more threads on audioholics and elsewhere that I need to read before I start asking stupid questions. At this point, I am just announcing the project because I am excited about it. But just because I can't ask for advice, does not mean I won't take it. Anyone who wants to weigh in right now is more than welcome - this stuff is fun, no?!
I will do my best to document the project as it happens (photos, measurements, etc.), for the benefit of future generations. And I will undoubtedly be asking for help later on.
I am pretty good with construction (carpentry and such), and I am familiar with home and car audio. I've read a bunch of stuff about speaker design over the years, and as a kid I built and played with many speakers. But this will be my first serious effort, so practically speaking I'm a know-nothin'-newbie.

Now's known details:
Goals / design considerations: To (have fun) build(ing) a sub for music and HT, with SQ and LF extension being priorities over SPL. I'd like it to play from around 15Hz - 35Hz. Within reason, I am happy to go big and odd with cabinet size, if it allows me to maximize performance. (For example, I have a corner where I could place a 36" * 18" * 48" cabinet. On the other hand, I cannot place a 50' long horn with a 50'^2 mouth - I would if I could, but I live in a rental.)
Right now I'm thinking I'll power it externally, mostly because I'm more familiar with conventional home amps than plate amps.

Known components: Yesterday I bought two JL 12w6v2s off ebay.

Associated equipment: My current mains are Definitive BP10s, which have respectable (and clean) output down to 35Hz or so. My receiver is an Arcam AVR280, which has 2 unused channels that can power surrounds or zone2, but sadly not LFE.
I also have a Klipsch RW12d that I got for a very good price from newegg. I've been hopelessly into home audio for decades, but the Klipsch is the first real sub I've owned. It's good enough to make me appreciate having a sub, but bad enough to make me want a good one. It adds body to the 30-40 Hz region, and rumbles when things blow up, but it's too slow for music, and below 25Hz the only real output is port noise. (Yes, I have a port-noise complaint. No one's ever made that joke before, have they?)

Areas in question:
Cabinet - Yep. Probably vented, probably big. I'd like to maximize output in the range of 15 - 35 Hz. I've played with cabinet design s/w, and will continue to do so. Then I'll ask for help from people who know stuff.

Electronics - Right now I'd be more comfortable using a standalone amp, for familiarity, flexibility, and cost. I aim to learn more about plate amp options, but ebay has lots of good monoblocks (Marantz, Adcom, whatever) for cheap, so that's pretty appealing. That said, I want to include filtering (low pass and infrasonic), and I imagine that phase control is good, perhaps even very important. I know that filter circuits are simple enough to build, but I don't know if GOOD ones are as easy. (I want to go as diy-cheap as I can, but I don't want to degrade SQ. I want the final product to be good.)
I recognize that using a fancy - I'll say equalizer - would probably be best, but again, for familiarity and cost, I'd like to go simpler. I will continue to read up on components like the recommended Behringer speaker manager. But they're costly and I have little experience with them. So, if I can achieve similar results with gain adjustment and a simple crossover circuit, I'd rather go that route.

Well, I think that's about it. Oh, in the interest of taking measurements - I have not been able to find documentation on using MeAud, so if anyone can point me to a manual, FAQ, or forum, that'd be great.

So, thanks to all the audioholics whose knowledge I've already gained from, and everyone here for making this such a lively, fun, and informative forum. I can't wait to have something real to share about this project.

 

[BoredSysAdmin]

I may be missed it, but do you want to build a single sub with two droverd or two separate boxes I'd go with second option if you could stand having two subs. Unfortunately I don't have sufficient knowledge to help you model the box, but few people around here can. Ask nicely for TLSGuy help.
As for amp and dsp. Best way to go is not consumer grade mono blocks, but pro D class amps. They cheap, very powerful and plenty of clean signal. One better value is Crown Xls series. Probably a single xls2500 could drive both subs

 

[BoredSysAdmin]

I may be missed it, but do you want to build a single sub with two droverd or two separate boxes I'd go with second option if you could stand having two subs. Unfortunately I don't have sufficient knowledge to help you model the box, but few people around here can. Ask nicely for TLSGuy help.
As for amp and dsp. Best way to go is not consumer grade mono blocks, but pro D class amps. They cheap, very powerful and plenty of clean signal. One better value is Crown Xls series. Probably a single xls2500 could drive both subs

And for dsp, minidsp is the product to get.

 

[TLS Guy]

Are you planning to build two boxes, or put both drivers in one box?

Those drivers are primarily intended for sealed application and cars primarily.

It is a relatively high Q driver and will not produce a nice tight bass in a vented box.

Even in a sealed box, total Q is at the upper end of acceptable.

Now one thing you need to understand, is that for any driver Fs, which is the free air resonance of a driver pretty much sets the low end extension. In practice the F3 the point at which the sub will be 3db down will be a little above Fs. Fs for that driver is 26 Hz.

Now if you go to the web page for that driver, you will find that the F3 for a sealed sub is 41 Hz. Q is 0.79. Now another rule. The total system Q can never be lower that the Qts of the driver. In this case Qts is 0.463. I personally like to work with drivers with a Qts no higher than 0.35 or thereabouts.

Q is the quality or width of the resonance. Total Q in the range of 0.5 to 0.7 are considered acceptable. As Qs go above 7 reproduction gets over warm and boomy.

Below 0.5 a lot of people consider the bass excessively controlled and dry. I personally aim for a Q of 0.5 in my bass alignments. At that point bass reproduction is considered essentially non resonant.

Since you have these drivers, I would personally go for a sealed alignment, 2 drivers in one box, in opposed configuration. I would then equalize the drivers electronically to 25 Hz and then have a 12 db per octave high pass filter become operative. This will prevent useless driver excursion and limit the high distortion trying to operate the drivers below Fs.

These drivers are not overly sensitive and when equalized a lot of amplifier power will be required. For two drivers you will need an amp that can provide 800 to 1000 watts into 2 ohms. If you drive each driver separately you will need two amps each capable of 400 to 500 watts into 4 ohms.

The F3 for a vented alignment is 31 Hz. You can not equalize a ported sub below F3, but you can a sealed one, if the drivers have enough excursion and can handle the power.

For those drivers I would strongly encourage you not to build a ported box,
and plan for sealed.

 

[BoredSysAdmin]

^^^ Do what he said. Amp wise (per TLS suggestions) - Crown XLS1500 is rated for dual 4ohm at 525W each (not bridged) - should be plenty of powah
I am contemplating building dual opposed sonotube subs meself.

http://www.amazon.com/Crown-DriveCore-XLS1500-Power-Amplifier/dp/B003HZV2JS/ref=sr_du_1_map?ie=UTF8&qid=1423871557&sr=8-1

 

[noneYet]

Wow! Thank you both for so quickly giving me way more help than I have yet earned (even if you didn't tell me exactly what I would have hoped to hear)! My plan was to use both drivers in one enclosure. I had hoped that by using two, I could drive the extension a bit lower, but I get what you, TLS, are saying about Q, and about tuning a box to play lower than a driver can handle. (That condition is particularly offensive to my ears, and the prevalence of boxes that are tuned to go lower than they really should is one of the reasons it took me 20 years to buy a subwoofer.)
I will always value accuracy over output, so I guess the dream of achieving substantial output to 15Hz will have to wait until I can build a rotary sub. And I guess my first sub doesn't have to be the best one I'll ever build; it's ok for it to be just ok. At least I can take solace in knowing that I should be able to build something that way outperforms my RW12D.
You have given me much to study on, so I won't make my reply too hasty. Though I should point out that these are the w6v2s, not v3s, which have a Qts of only 0.45, and the much lower Fs of 25Hz. I guess when you say "in opposed configuration", you mean isobaric loading?
Again, you've given me more help than I deserve, already, so I will buckle down, do my homework, and come back when I have more mistaken ideas you can correct. (And I will study up on class d amps, BoredSysAdmin, so thank you for that, as well.)
Thank you both very much.
(Oops - can't post a link yet. The product page for this driver is just like the link in TLS Guy's post, but ends in 12w6v2-d4-car-audio-w6v2-subwoofer-drivers-92121. I appreciate the caution, btw. I just love the "want to market the sale and promotion your products by pretending to be a forum member?" bit.)

 

[BoredSysAdmin]

@TLS Guy , op has good point, he's got pair of v2 drivers, not v3
Correct link: http://www.jlaudio.com/12w6v2-d4-car-audio-w6v2-subwoofer-drivers-92121

Dear OP, Many good folks built speakers based on Tls designs. So far no one complained
Check out his home studio to get some idea

 

[noneYet]

@TLS Guy

Dear OP, Many good folks built speakers based on Tls designs. So far no one complained
Check out his home studio to get some idea

I have read a few of TLS' threads, and no doubt, I respect his opinions & appreciate his guidance. In fact, it is the expertise and helpful attitudes that I've seen demonstrated by a dozen or so AH members that has really drawn me to this forum. (I've been occasionally visiting the AH forums for years, and always found it a valuable resource.) If I sound like I'm questioning what y'all are telling me, it's only because if I just <i>blindly</i> do exactly what I'm told, I will feel like I'm cheating - making someone else do the work for me. (E.g. presenting my plan and having someone critique it, vs. just saying "Will someone make my speaker for me?")

BoredSysAdmin - Thanks for suggesting miniDSP - I had not heard of it, and it may be just what I want here. About your recommendation to use a pro audio, class-D amp. Do you suggest that solely for energy efficiency? Or are there other reasons that a class-AB consumer amp is less appropriate here?
Thanks again, guys

 

[BoredSysAdmin]

I'm not an expert on subject matter, if you think about it for a second - there's nothing magical about how amps work.
Pro amps just do it without the fuss. Thd might be a tad worse , but as long its below 100 db you have nothing to worry about.
Subwoofer is not like a sensitive tweeter. It needs gobs of power. It often produces it's own signal distortion which on occasion you might even like
Specifically D class is much more power and heat efficient and make much lighter boxes. Heavy is not always better.
I know at least one AH forum mod who's very happy with same amp xls1500 driving his top notch DIY speakers (statements btw)
If it's good enough to power full range speakers, why shouldn't it be great powering sub?

 

[TLS Guy]

@TLS Guy , op has good point, he's got pair of v2 drivers, not v3
Correct link: http://www.jlaudio.com/12w6v2-d4-car-audio-w6v2-subwoofer-drivers-92121

Dear OP, Many good folks built speakers based on Tls designs. So far no one complained
Check out his home studio to get some idea

The drivers are not significantly different, except that in their published sealed model Q is in excess of 8 which would be a ghastly thumper sub.

I will have to play with the box model tomorrow and see if we can get the total system Q down a bit. I'm starting to get a horrible sense that those drivers are not suitable for home audio at all, a lot of drivers from that company are not.

I will play with the model tomorrow and see if we can tame it.

If not the OP will have to repost the drivers and we will start again.

I was not thinking of an Isobarik solution,but one where the drivers are on opposite sides of the box.

The only advantage of isobarik loading is to halve the size of the box. However performance is the same as one driver, with the exception that 3db of sensitivity is lost.

One further note to the OP, using two drivers does not add bass extension, it only adds increased power handling and if the woofers are wired on parallel then there is a 3db increase in sensitivity.

 

[TLS Guy]

I have doodled with that driver.

If you put two in a box sealed you can get a respectable sub.

In a box with an internal volume (you need to add driver volume and brace volume to this number) of 1.76 Cu ft you get an F3 of 48 Hz and a system Q of 0.707. This is acceptable. There is enough xmax in a box of that size you do not run out of cone excursion. So with a drive of 1200 watts you can get an spl of 109 db. This is very respectable.

The box needs heavy fill with acoustic damping material.

So you need start Eq 12 db per octave at 50 Hz and then cut it off at 25 Hz with a 12 db per octave high pass filter.

This cabinet does several things.

The response is smooth without ripple.

Q is lowered to an acceptable range.

The box is small enough to control cone excursion at maximum power, so the driver is not damaged. This driver has a very high mechanical Q (sloppy suspension) and so it needs the resistance of air pressure in the box to properly control it.

This driver has to be in a sealed enclosure, or it will self destruct. JL have no business publishing a ported alignment, which in any case would sound awful with this driver.

Please see attached the design pdf.

The program does not allow the drivers to be shown on opposite sides of the box, but box volume and specs do not change.

Wire the voice coils of each driver in series to make 8 ohm drivers, and then wire the drivers in parallel to make a 4 ohm sub.

To maximize performance you need an amp that can deliver 1200 watts into 4 ohms.

Do not worry about bass extension to 15 Hz. All that is required is good performance to 25 Hz, which this gives you. Despite what you may read there is no point in chasing the infrasonics. In any event, even with the high pass filter engaged the sub will still only be down 6db at 20 Hz and with room gain it will be pretty much flat to 20 Hz. It will do the job for you.

 

[annunaki]

Having personally worked with this driver and heard the results, I can tell you it works well in sealed or vented applications.

Not sure if things have changed as it has been a long time but in a third party review, with Klippel analysis, it was the most linear sub Tom Nousaine had ever tested. His analysis was not limited to the motor alone but the suspension included (KMS) measurents.

I would have to dig up previous models I have done for it but venting to the mid 20's seems to be what I recall worked well in the past.

I am not stating any of this to cause argument, but having heard this sub first hand in both enclosure designs I can vouch for its credibility, capability and performance. It is among the best subs I have heard in a sealed configuration and in the top 5 in a vented.

 

[lsiberian]

I'm with Annunaki on this one. I know the numbers may not suggest it, but this sub can definitely work in a vented enclosure. It was even considered a viable alternative to the Kappa Perfect. This is what we call an outlier. You really do have an excellent sub that works wonderfully either way. Sealed is much easier for a first time build of course.

 

[noneYet]

Wow! I say again. Thank you gentlemen for all of your help. TLS Guy, I feel like you've put really put some time and effort into this. I am a little confused by the group delay - when I put the numbers you provided into winISD, it shows me Group Delay of ~16 - 20 msec - 4 to 5 times what your BassBox graph shows.
I follow what you are saying about needing the spring of the box to control the driver - is that represented by Qtc?
I have been playing with the numbers in winISD, and have come up with one configuration that looks ok to me (with the exception of high group delay - but maybe that's coming out wrong for some reason). The figures I've used are much different than what TLS Guy provided, and I would love it if you good folks could tell me what I'm missing - I imagine there's something flawed with my configuration, and I would love to know what it is.
This is figured with a volume of 350L (yes, I know how big that is), a Qtc of 0.574, 1st order HP at 14 Hz, 2nd order LP at 45Hz. Total output is similar to what TLS Guy modeled (though I figured this one with only 600W - not 1000). Cone displacement peaks at about Xmax, at about 17 Hz - so output at 20 Hz is achieved below Xmax. I have read many conflicting opinions about whether a higher Qtc is "tighter" or "looser." "Tighter" would seem more intuitive to me, as it means we have a stiffer air-spring, no? But that is the opposite of most of what I have read. I'm attaching a few graphs from winISD - for that matter, I could attach the project file, if anyone cares to see it.
Again, TLS Guy, I am in no way discounting the efforts and advice you've given me - I am grateful for both. I also want to try what seems reasonable to me, and then hear what's right or wrong with what I have done. (And just to reiterate, I am only interested in response <40Hz - my mains already carry down to 35 or so reasonably well.)
Thanks again, everybody.

 

[annunaki]

I will try to model the sub again to come up with an optimal approach. I would do two separate enclosures personally. You may find that crossing over at 60hz or even 50hz may offer better results. I say that as a 12" driver can reproduce output at such frequencies more efficiently than smaller ones. That you will have to experiment with.

 

[noneYet]

...crossing over at 60hz or even 50hz may offer better results...

Thank you for the suggestion. I do see, right off, that raising the F of the HP can lower the Group Delay at the high end of the range (though it also raises it at the low end). The BP10s' pseudo-transmission lines do a pretty good job with that medium-low end, and I've always feared letting "big, slow" drivers go up too high. But I guess 50 or 60 Hz isn't so high, as long as I balance it carefully. And it actually might blend more easily if the transition is further from the BP10s' threshold.
And I hadn't even considered doing this build as two separate cabinets - I think I really like the idea of doing something a little more unusual. Nevertheless, I shouldn't leave so obvious a stone unturned. (I'd like you to know that reading about your beautiful "Pythagoras" sub is what finally pushed me into doing this. I started looking for those Acoustic Elegance drivers. Couldn't find them, but by then the ball was rolling...)
I might add, on the subject of multiple subs, that I thought I might keep my RW12d active. I thought it could help carry the transition region (40 - 50 Hz) at a low volume.
You gentlemen must be either compulsively helpful, or really into acoustic engineering & design. Either way, I love it. I have clearly come to the right place. I'm just sad that it took me so long to finally join this forum.
Thanks again!

 

[annunaki]

I got a chance to model the driver in question. Overall things look rather decent from an anechoic perspective.

The frequency response is rather linear across the operating range (I used a 60hz low pass 3rd order). I also used a high pass infrasonic filter at 17hz 2nd order. If the infrasonic filter is not applied the below recommendation would need to be altered. I also modeled with 400 watts of input power.

I came up with a 4.35ft^3 enclosure tuned to 21hz using a 2.25" x 14" slot vent 39.43" long. The configuration puts the -3db point at 20.4hz and the -10db point at 16.5hz.

Excursion stays completely within limits through the entire operating range. Although a cascaded high pass may help at the lowest of frequencies.

The vent velocity stays well in check peaking at just 15m/s at about 18hz. Additionally group delay looks well controlled until you get at or below tuning which should not be of any significance.

To me it looks like a pretty healthy vented system.

 

[lsiberian]

I got a chance to model the driver in question. Overall things look rather decent from an anechoic perspective.

The frequency response is rather linear across the operating range (I used a 60hz low pass 3rd order). I also used a high pass infrasonic filter at 17hz 2nd order. If the infrasonic filter is not applied the below recommendation would need to be altered. I also modeled with 400 watts of input power.

I came up with a 4.35ft^3 enclosure tuned to 21hz using a 2.25" x 14" slot vent 39.43" long. The configuration puts the -3db point at 20.4hz and the -10db point at 16.5hz.

Excursion stays completely within limits through the entire operating range. Although a cascaded high pass may help at the lowest of frequencies.

The vent velocity stays well in check peaking at just 15m/s at about 18hz. Additionally group delay looks well controlled until you get at or below tuning which should not be of any significance.

To me it looks like a pretty healthy vented system.

I haven't tried modeling it yet, but that definitely is what I was expecting.

 

[noneYet]

Immaterial update - (I have to drive about 70 miles for materials, so I'm not building just yet. But, having done lots of reading, pondering, calculating, guessing, and playing around with Sketchup) I have come up with what may be my box design.
For the retro-grouches out there - historically, I've been a pencil and paper kind of guy. But for this, I'm glad I tried some CAD. Here are two major reasons why. First, it makes precise scaling easy. Second, it makes it easy to view and change the box as a collection of 3d objects - especially handy for doing bracing without having to do lots of re-drawing. There is a bit of a learning curve, but I found Sketchup to be pretty intuitive. And even with only a pretty rudimentary grasp of it, I was able to muddle through my model, and improve as I went.

Now on to the design itself. I've decided to go the sealed route, with a net volume in the neighborhood of 110 liters. According to winISD this gives me a Qtc of 0.737, with fair and flat output within my range of interest. A volume of ~110 L also gives me decent values (¿maybe?) for these ratios, which I am led to believe are best when they are close to 1:1:
Fc/Qtc (55.51) : Fs/Qts (55.56), and
Qtc/Qts (1.638) : Fc/Fs (1.636) : (α+1)^0.5 (1.575). I temporarily forgot that I've always been biased toward sealed enclosures. That's not to say fairly, or rightly, but biased nonetheless.

I have settled on the form of a right-triangular prism, with the drivers in opposition. This form has several qualities I'm looking for. First, it's a little unusual. (Maybe superficial, but so what - it's mine and I want it that way. And besides, I think it's important to try things differently, to see what happens.) Second, it largely eliminates parallel faces. Third, I like the idea of mounting the drivers in opposition, which allows some mitigation of cabinet and basket resonances. And finally, if I added some crap to the front, it could double as a magazine rack. Ooh - or LPs, which could add a heap of mass!

There are two parallel sides - the sides that have the drivers mounted to them. I have dealt with that in two ways. First, the opposing faces are heavily braced, and tied together - particularly around the circumference of the drivers. Second, the material of the mounting baffles is double-thick. I chose to do this to make lap joints around all driver-loaded corners. Plus, it raises the resonance frequency of those sides. Up near the top - the acute vertex - I figure I'll stuff it with fiberglass (I figure that up there, waves should be at high frequencies anyway).

I do have a few lingering questions (which I will continue to research, myself - I did just buy a copy of the Loudspeaker Cookbook, which I will read thoroughly - but which I would gladly read forum answers to).
First, a small matter. I kind of like the idea of including a "bleed valve" - a small (maybe 1/4") hole, stuffed tightly with felt, just to ensure that pressure in the cabinet doesn't drift away from 1 bar.
Second - a multi-parter. If I run this through a mini-dsp, do I still need hardwired high- and low- filters? Since these filters dramatically increase group delay, I wondered if it would be ok / beneficial to forego them, for the digital filter.
Third, those aforementioned ratios. I read what they mean, and calculated their values. But I don't know how important they are, or how close is close enough.
So I am definitely still interested in any opinions that anyone wants to offer. If you notice something stupid about what I've presented (whether you're right, or just opinionated), I will be genuinely happy to read about it. And to you vetted individuals who've already offered your 200,000¢, I hope you won't feel insulted that I have not straight-taken your suggestions. I definitely value your expert opinions, it is just important to me to tweak things and make them my own.

I have attached the sketchup project file for anyone who wants to view it. Thanks again for all the help, past, present and future. Woohoo! Oh - nope. it doesn't want me to attach the Sketchup file. I will rename it to jpg to see... nope. no good, either. Well, I will explore the forum, and find out if and how I'm allowed to share these project files. Good night, everybody.

 

[noneYet]

Stereogram! Just for S & G... and because I can't upload the Sketchup project file... or start construction... here is a stereogram of the box (for all you 2 channel afficionados ). The front is hidden. I spent too long doing this, and it's pointless. But... well, y'know.