B&C 18SW115...A critical "Q" experiment!

Well when the "Q" is right, one must build, and while the result won't likely match what preceded it, such a result is not required this time around. I've always had a penchant for high BL-low qts drivers, chiefly because the motor force is sufficient to shoehorn them into impossibly small boxes but with equally devastating results. The heady days of the Adire Tumult and the TCSounds TC-3000 have faded but new players have emerged, even though the problem of vaporware seems a continued vexation. Indeed after looking at the TD18h+ and the Pro 5100 I once again resigned myself to the fact of spending $500 or more on a driver with the right "metrics".

I did find the SW for much less than it was on Parts Express. I only hope it is'nt DOA...I just might fall into conniption! This will be my first experience with an 18" driver and I chuckle to myself when thinking just a month ago that I would just put (4) 8" Tang Bands in 2.0 cu. ft and call it a day. It is a far cry from the 125 dB, 5.5 kilowatt 3x15 TC-3000 based system I had five years ago. However this time the room is much smaller (1200 cu. ft.) and the gain "that must not be named" will start much earlier.

So this is something of a reconnaissance mission to see if the reputation that precedes this driver is accurate and whether I can defy the limitations of its preferred operating range. I'm chiefly interested in it's excursion limited performance below system resonance, which hopefully ends up around 53 Hz or lower. Crossed at 40 Hz, the system resonance will be outside of the operating range for critical listening. It will only operate above in home theater mode. By comparison, the 3x15 system had a measured Fb of 33 Hz (Qtc. = .450) so I'm quite interested in discovering if the sound signature will be different.

Box size [3 cu ft.] for the SW is growing by the day because my obsession with critically damped systems usually wins out over my heightened reluctance to carpentry. The last enclosure was built [not by me] for nearly $200 and it equated to almost $1 per pound after the final weigh-in. This time I'm doing it myself and it's been only "politely" challenging so far.

I sold most of my measuring equipment but I still have my trusty impedance analyzer, the singular piece of electronic equipment I've ever built, and so it is prized very highly (thank you Dan Marx). Still have TrueRTA, REW and a Rat meter somewhere.

Looking around Home Theater Shack, AVS and a few other sites I can see that the term "basshead" has taken on new dimensions, as I see a myriad of mulitple 18" and higher projects. I'm sorry to see the Maelstrom and the FTW go discontinued but the costs of those driver sizes are ultimately too prohibitive I guess. Ricci's 21SW152 dual sealed was a quiet surprise, as is his website. I don't know whatever happened to Illka but it seems someone else has stepped up to measure these beasts of prey we so assiduously hammer together and that is always a welcomed duty.

My current space precludes any high decibel theatrics but it is likely that I will eventually employ an opposed 2x18 if the SW works out. For now, I await the driver...

Well I really wanted a TD18h+ because the qts is a really low .22 IIRC. I also really like the inductance value. It will be a little limited thermally and displacement-wise compared to the SW but it would be my first choice. I just don't have the heart to wait I'm afraid. I've been there too many times before. I sent out a message to AE and got the auto response update about the group buy.

One of the things I really like about these Neo drivers is the weight savings. The TC-3000 drivers weighed about 46 lbs. and I had two of them in a 150 lb enclosure so it was pretty heavy. And those were only 15" drivers, the typical non-neo 18" will run 60-70 lbs or more depending on the xmax of the driver. By comparison the SW is a svelte 26.5 lbs, give or take. I keep working on the cabinet, which is now at 3.5 cu. ft. and I'm actually worried that it might not be heavy enough even after the driver is put in!

I just checked the AE's forum and John relayed about a three to four week lead time since each driver is custom built. That does'nt seem so bad a wait but the SW is already on the way.

Funny you mention Bag End as I have read just about everything on their integrator, even schematics to build one. However experience with the 3x15 system showed that a very low qtc. value coupled with EQ (with variable Q) shelving above and below Fb can produce a very similar result. In the case of the original Infrasub-18 they sacrificed a lot of efficiency to run flat (in-room) to 8 Hz and the proof is the many instances of reviewers finding that it was a little short on the output side in larger rooms. Obviously even an 18" driver is limited in how much power it can absorb under a boost curve designed to perfectly mimic the typical rolloff of a sealed subwoofer, not to mention the suspension issues that would likely arise long before that. Moreover, I think we both know that it is quite quixotic and questionable whether 400 watts can deliver the requisite visceral impact of an 8 Hz fundamental. I don't know, maybe if smallish concrete bunkers are your thing!

What is different about this design and my last one is the box frequency, which came in 5 Hz lower after polyfill was added. Winisd Pro was saying 38-39 but the analyzer said 33 Hz. This design looks to be different and Winisd is saying 55-56, but I will be satisfied with 53 or lower, which puts the qtc at or below .5. With the bottom room mode occurring at a rather high 40 Hz, it may be that eq might not be needed at all...unless I want a "theatrical" bottom end.

Glad to hear about ilkka, he was really into the measurement side of things. I still have a folder full of graphs on subwoofer tests conducted by him. Genelec is fortunate to have him.

Hello sir. I have the "bright" Klipsch RB-35's and Klipsch SB-2's for surround. I expect the sub to extend as low as my previous sub did only this time the room is much smaller [1250 cu. ft]. I had a 22 foot wall previously. That is now reduced to 14 feet, which raises bottom room mode from around 25 Hz to 40 Hz. The box frequency is a combination of both the Free air resonance of the driver and the enclosure itself. It can be measured with an impedance analyzer. It can also be manipulated "somewhat" with varying degrees of polyfill.

Hello Noah. Yes I'm aware of the Krevosky article having read it some while back. An interesting article but even it acknowledges the role that the room will play in the final analysis. It's not just about the timbre of sound, which is highly personal I'll admit. For me its about footprint, soundstage, rolloff, FR, non-theatrical bass, simplicity and maybe a tinge of ego [i'll admit] thrown in for good measure. But I can assure you of this, there is nothing anemic about a system with a qtc. of .5, especially not at 125dB. When it has something to say it states it eloquently, otherwise it keeps its big mouth shut. That, I can appreciate.

Building a subwoofer with a Q value of 1.0 exacerbates the inherent problems of extension that plagues sealed subwoofers, which is why most designs call for a Q of .707, which in itself is insufficient and requires equalization. If it were even physically possible to put the SW in a Q=1.0 then I guess there really would be a chance to fry a 4.5" voice coil rated to 3.4 kW. The response on the bottom end would be that bad! Hence the low Q enclosure.

I appreciate your observations Noah.

Hello Dan, thanks so much. I've been away so long that I missed Ilkka doing this, and with the LMS no less! I've always harbored a similar suspicion that a driver fighting an overly protective air spring was detrimental ever since the days of the Tumult when they were being put in smallish boxes and coerced in various ways. I also remember that 2700 watt Sunfire in a ridiculously small box!

The proof is in the pudding and the measurements speak for themselves. I don't mean to boast, just merely to say that I have always thought that a driver should be able to breathe...The sweet thing about high motor force/BL and its consequent low qts is that you can have both a smallish box and a shallower roll-off, which is somewhat paradoxical in the way one would normally think. The best part is being able to do it with 15 & 18" drivers. You will still have to apply some equalization but the box environment is much less hostile to your efforts. Ilkka's results certainly confirm this as well as his observations about power compression.

The SW has a voice coil size that will likely render power compression irrelevant for my application/environment. I've got the driver on a 10 Hz sine to stretch everything out but real power probably won't arrive until next week. I vacillated for a while but decided on 900 watts from QSC.

I just can't help thinking about two of these opposed in about 6.5 cu ft.

First Burn-in...

After 6 hours on a 10 Hz sine...Fs was measured at 25 Hz, which is much lower than the manufacturer specified 32Hz...the sine was turned off at 4pm...

15 hours later...the Fs has risen to 27 Hz but undoubtedly, as the following website posits, QES QMS and QTS have all fallen a little from manufacturer specs:

http://www.gr-research.com/burnin.htm

I will check it again at the end of today then put it back on the 10 Hz for 12 hours and then see what shakes out after that. Eventually I will want to do 24 hours total before installing the motor.

The aforementioned website has some fascinating details on the mechanical and electrical burn-in phenomenon. Dan Wiggins himself chimes in with his thoughts as well! ... a good read.
Edited by kramskoi - 2/28/13 at 6:16am

First I have to say that this was a very intense burn-in, with the driver going for up to 9 hours non-stop at 25mm p to p...There were however moments of terror when TrueRTA suddenly locked up and shut down, causing the driver to go still... which caused a frantic grab for the multimeter to test the driver!! It is after all a nearly $600 sub...


Driver new [no measurement before burn-in]

6 hours/sine 10 Hz

25 Hz hot measurement [1] day one
==================================================
off [15hrs]

27 Hz cold measurement [2] day two
9 hours/sine 10 Hz
=====================================================================
off [14 hrs] no measurement, short across two resistors inside the analyzer...had to repair...
2hours/sine 10 Hz [17 hrs total]

Fs = 24 Hz [measured 3 min after sine off]

return to 10 Hz sine for 7 hrs to complete burn-in day 3

24 hours complete...estimate final Fs = 26 Hz

Fs = 26 Hz after 3 hours off
===============================================================
15 hours off day 4

Fs = 26 Hz final cold measurement after 24 hour burn-in


I don't know what the Fs was when I first received the driver but the specs list at 32 Hz. I did see a drop from 27 Hz to 26 Hz for the cold reading with lowest hot reading of 24 Hz on day three, immediately after turning off the sine. The driver really never got warmer than about 100 degrees, I compare it to really warm skin. That was about it, even after 9 hours continuous [25mm], although it is a 4.5" VC rated for 3.4 kW. Very quiet excursions at 25-30 mm peak to peak.

I did notice a 2 Hz rise between the hot reading and the next day cold reading...this happened daily. Day three showed the lowest reading, after 17 total hours.

There is no doubt that qms, qes and qts are lower, with Vas surely higher and I'm severely tempted to buy a woofer tester to confirm it. There is no way the stock values can be the same with a Fs of 26 Hz cold.

Hello Dan, well you can surely get great bass from low qts drivers. The two TC3000 15" were in a dual opposed sealed 5.6 cu ft. enclosure, with a qtc of below .5. However bass-reflex is... in the words of Sherlock Holmes..."not my area." That is not to impugn the exquisite edifice of evidentiary findings on the quality that is possible with vented enclosures. In a large room, the SW would not be the best choice, unless it was in multiple configurations or vented. That will also depend on your decibel requirements. If you are looking to do a vented application then I would say that the only way to ensure precision is a woofer tester because after burn-in the driver parameters are certainly changed. At 24 hours the SW had a warm Fs of 24 Hz...and maybe the spider has'nt completely relaxed...GR posits 40-50 hours, but I'm sure they were'nt using the same "methods" that I have.

I'm pretty sure that with a bottom room mode at 40 Hz, Q = .5 and eq-shelving, the sub will extend below 20 Hz. I've modeled it in winisd pro the same way I modeled the TC3000, shelving filters and all. The only question is output ... but its only 1200 cu ft. [14x11x8]

What the GR site showed was that a driver's parameters do change, resulting mostly from "cracking the spider" as it were. Looking at the SW, an Fs of 26 is quite a difference from the average run data of 32. It means that qms, qes and qts are all lower than specified...how much is debatable but the GR study showed a 17% drop in qes/qts along with an improvement in sensitivity as well, owing to the increased acceleration of a more compliant spider...

Well I don't know what to say other than the fact that if an Fs of 32 Hz is the result of a burn-in then I'm somewhere in the twilight zone with what I've measured so far. Actually, for the last cold reading it was difficult to tell whether it was 25 or 26 Hz...I went with the higher number but the multimeter is analog and both frequency readings looked almost identical.

Anyways I'm pretty happy that the Fs is 26 and not 32 as this means that both qes and qts are lower, which suits my aims.

I did another 9 hours on the sine today and will test cold again tomorrow. So that puts the driver at 33 hours.

Hello again Dan...Tested the Fs this morning after 9 more hours yesterday...25 Hz appears to be the bottom while cold. Taking a measurement hot after a few hours might show 23 Hz. Further progress will be small so I will discontinue. So that is 33 hours total.

The winisd pro driver data shows a theoretical 126 db, which will of course be a fair ways above the final Fb [box resonance]...As for low frequency output, the only data I have to go by is Ricci's testing of a 21" SW152 in a sealed Bessel alignment [5.5 cu. ft. qtc. 577]. The SW 21" got 101.5 dB at 20 Hz outside. The Dayton 18" got 102.5 dB. The final qtc. should end up around .5 but I don't know how much difference that will make. In the end, it only matters what I get in room, though I won't be able to compute the THD because my ecm8000 is long gone...I'm slowly beginning to realize how much audio equipment I've sold over the last four years. Hopefully I can get 100-105 dB at 20 Hz corner loaded in-room. I'm counting on the pressure vessel effect from a 14 x 11 x 8' room. I guess we will see how this shakes out on Tuesday or Wednesday maybe...the amp arrives Monday...it is a QSC USA 900...so 900 watts at 8 ohms bridged [32dB voltage gain].

I can tell you that I spent MUCH wasted time on the notorious Linkwitz Transform. That was before experience and testing showed me that it is not needed in a critical Q alignment. You really want to use "it", and Bag End's integrator-type boosting apparatus for drivers in boxes too small for their own good. The shelving I indicated is what I used for the 2x15...I had an fbq 2496 which allowed me to go wider with the Q [5 octaves] I set +6 dB @ 20 Hz and another at 80 Hz -7 dB same width. It had the effect of shelving down the upper bass while raising the lower octave. One could also simply go with a single parametric with a much smaller Q value [at 20 Hz] but depending on the room resonances, you may be boosting an already boosted frequency. The dsp1124 I ordered only allows up to 2 octaves but the 10-20 Hz octave will still be affected, especially at 15 Hz, which is a small hot-spot of the excursion profile that runs almost dead flat to 10 Hz...the room will do the rest. In any event, the SW has an xlim of 60 mm peak to peak so there is protection without the need for any subsonic "attenuation", which I find dreadfully distasteful in a sealed alignment.

Finally, I must say that my goals for this design is modest from an output standpoint. I really want it to play deep and well most of all. But if I can't get Saint-Saens [big fan of organ Toccata's in Minor] to come through with some "presence" then this project will be deemed a failure. Period!
Edited by kramskoi - 3/3/13 at 12:17pm

Preliminary findings:

No polyfill ... bare cabinet.

Fb = 55.5 Hz

Fs = 25.5 Hz

qts. of driver at 25.5 Hz ... unknown

qtc. of system unknown without new qts @ Fs

likely between .51 and .53
==============================================
Update

3 lbs. of polyfill

Fb = 53.5 Hz

The sub should be right at qtc. = .500 or thereabouts but it is impossible to tell without the new qts value for the Fs = 25.5 Hz

Okay LT... I was looking at something similar in regards to compliance...Back and forth with both Win and Unibox. I know the Fb is 53.5 Hz, which is nowhere even close to Winisd's 64 Hz in a 3 cu. ft. box.

How does it sound?

I did some low level organ reconnaissance ...Dallas Wind Symphony "The Vikings", "Suite Gothique" and Vierne's Toccata in B flat...All I can say is that the driver is just sitting in the box not even screwed down...yet even at a low level the organ stabs come through. I don't even have it corner loaded yet. Got some schoolwork to do but I will try and get it tamped down and put in it's proper place by midnight.

I do think that output will not be a problem and I will deal with more extension when the dsp 1124 arrives by the weekend... if any is actually needed!