Comparing LF drivers

[Michael John]

Here's a, probably naive, question for the seasoned speaker designers lurking here...

Recently I've spent way too much time looking at spec sheets for 15", 18" and 21" drivers - JBL, B&C, BMS, 18sound etc. - and one way I've tried to compare the drivers is to plot their 1W freq responses scaled up by 10xlog10 of the AES rated power. (Yes, I know this doesn't account for suspension and temperature compression effects, and final cabinet design.)

With the large number of multi 1000 watt drivers out now, I expected to see major differences between these calculated SPL freq responses. Instead I find they're all more similar than different, and the 21" drivers have only fractionally more output at the low end than some of the 18" drivers.

However, one stand-out is the old JBL 2242H (18", 800W, ferrite) which shows a clear 4 to 5 dB higher output below 100 Hz than many of the newer drivers. It has a much lighter cone at about 158 grams: compare to the B&C 18SW115 (18", 1700W, neo) cone at about 275 grams.

So I'm wondering, what are the major advantages of the multi 1000 watt drivers? A clear disadvantage is the need for much higher power, and higher voltage (for 4 and 8 ohm) amplifiers.

 

[Phil Graham]

Re: Comparing LF drivers

Here's a, probably naive, question for the seasoned speaker designers lurking here...

Recently I've spent way too much time looking at spec sheets for 15", 18" and 21" drivers - JBL, B&C, BMS, 18sound etc. - and one way I've tried to compare the drivers is to plot their 1W freq responses scaled up by 10xlog10 of the AES rated power. (Yes, I know this doesn't account for suspension and temperature compression effects, and final cabinet design.)

With the large number of multi 1000 watt drivers out now, I expected to see major differences between these calculated SPL freq responses. Instead I find they're all more similar than different, and the 21" drivers have only fractionally more output at the low end than some of the 18" drivers.

However, one stand-out is the old JBL 2242H (18", 800W, ferrite) which shows a clear 4 to 5 dB higher output below 100 Hz than many of the newer drivers. It has a much lighter cone at about 158 grams: compare to the B&C 18SW115 (18", 1700W, neo) cone at about 275 grams.

So I'm wondering, what are the major advantages of the multi 1000 watt drivers? A clear disadvantage is the need for much higher power, and higher voltage (for 4 and 8 ohm) amplifiers.

Michael, its Saturday morning on the east coast of the us, and my "honey do" list is long today. Let me a stop you from further effort down this line of reasoning and say you can't look at it this way. To explain why requires a lot of typing time that I do not have this weekend on account of needing to write an article draft for FOH.

If you pm me, and have a reasonable way to call the US, I'm happy to pm you my mobile and talk it through.

 

[Art Welter]

Re: Comparing LF drivers

However, one stand-out is the old JBL 2242H (18", 800W, ferrite) which shows a clear 4 to 5 dB higher output below 100 Hz than many of the newer drivers. It has a much lighter cone at about 158 grams: compare to the B&C 18SW115 (18", 1700W, neo) cone at about 275 grams.

So I'm wondering, what are the major advantages of the multi 1000 watt drivers? A clear disadvantage is the need for much higher power, and higher voltage (for 4 and 8 ohm) amplifiers.

The biggest advantage is in linear displacement, the Xmax or Xvar.
Just like in engines, there is no replacement for displacement, for low frequencies we need to "move some air".
The JBL 2242H has only 7.87 mm Xmax, the B&C 18SW115 has 14mm Xmax, 16mm Xvar.
Doubling the displacement nets a 6 dB increase in output, but also requires 6 dB more power, hence the greater power handling.
Given the same Xmax, a 21" has about 28% more displacement than an 18", so can put out almost 2 dB more SPL.
The increased displacement and power handling require heavier cones and voice coils, which reduces efficiency a bit, so given limited power many of the old cones will outperform the new crop.
With amplifiers exceeding 90% efficiency, and power available for about 25 cents a watt, loosing a few dB in speaker efficiency to gain 6 dB output from the same enclosure size is a compromise I'm certainly willing to pay.

The B&C 18SW115 gets rid of heat better than many older designs, which reduces power compression.

The heavy, stiff cones also can handle the high compression ratios used in horn loaded designs with far less distortion than lighter cones.
Horn enclosures can increase the output of a capable driver by 6 dB. As an example, my single B&C 18SW115 Keystone tapped horn design, a bit smaller than a dual 18 JBL SRX728, has slightly higher sensitivity, but due to the increased excursion capability, has more output potential.

Art

 

[Bennett Prescott]

Re: Comparing LF drivers

So I'm wondering, what are the major advantages of the multi 1000 watt drivers? A clear disadvantage is the need for much higher power, and higher voltage (for 4 and 8 ohm) amplifiers.

Michael,

There are a couple of problems with scaling listed 1w sensitivity by rated power, but in this context we'll ignore power compression and sound quality and focus on what makes subwoofers different.

The listed sensitivity is usually measured over a frequency range like Fs -> 10Fs, so maybe 35 - 350Hz. In practice this number is dominated by the midband sensitivity, which in fact doesn't matter to us very much at all. This kind of statistic is very useful for midrange and HF drivers, which are often used for several decades and so a free air measurement of their midband sensitivity corresponds well to real world performance. Subwoofers are somewhat unique in that operation for even two octaves is relatively rare, so it is better to look elsewhere for performance comparisons.

It is very important to look at subwoofers as more than a discrete driver, since every application marries a subwoofer with a box, and the two form an integral system. Modern sub woofers are over-damped, with a Qts in the range of 0.2 - 0.4, which ends up working well with under-damped box and port designs - not to mention the atmosphere. This makes life a lot easier, actually, since given a woofer designed with this in mind you can pretty much tune the box to the woofer's Fs and get a good magnitude response. That said, apparent output in the subwoofer range is mostly a function of Qe, as more damping equates to reduced LF. Therefore, you can build a woofer with a light cone (which will give good sensitivity), and put a loose suspension on it (which will get the Fs back to the expected range), and put a weaker motor on it (which reduces Bl and improves apparent LF), and have a woofer that looks great on paper with significantly increased LF sensitivity but reduced cone control. The real tradeoff is in Vas, or how big a box the woofer needs. If you look at the JBL 2242H you reference, its Vas is almost twice our 18SW115. What we've done is put on a massive motor, a very stiff suspension, and a heavier cone. In our experience this results in better reliability and sonics, not to mention the ability to get similar performance in nearly half the box size. Certainly there are many applications where a high Vas is no big deal, but most of our customers build products that have to travel regularly by truck and so this advantage in output from a relatively small box is A Big Deal™. Look at our 21SW152, for example - significantly higher Bl (with a lower Re!) than the 18SW115, more cone area, more excursion, more power handling, while maintaining a Vas of about 7ft^3!

Many of these tradeoffs result in off-spec-sheet differences that can end up being really significant. Careful attention to Klippel data will reveal more of the magic here, all our woofers exhibit extremely low DC offset, excellent symmetry, and relatively flat Bl. The end result is a woofer that sounds better, especially when driven to rated power. Our voice coil, former, and motor cooling technologies are a big part of the jump in power levels you have observed as well, but don't think of them as spurious. We really can build woofers now that have more than twice the power handling of the best of the previous era's, while having better usable LF sensitivity, and that behave much better at max output - letting you actually utilize these hard won advantages.

 

[Josh Ricci]

Re: Comparing LF drivers

In addition to what Bennett has said...

Do not put too much faith into what you see in speaker simulation programs or with using calculations to scale up output. These are "perfect" results with zero compression or "drift". This is not what happens in the real world. They are useful but always remember that the data put out is only as accurate as the data put in. Driver specifications are typically captured at the small signal level and many of them will change drastically once stronger signals are sent into it. (This is why Klippel type testing is very useful) The higher power drivers with more excursion capability and better cooling will better cope with and respond to increased output demands.

I have taken measurements at high power on 2242HPL drivers and many other high power 18" drivers and while the 2242HPL looks very good on paper just as you said it does, it exhibits much lower power handling both thermal and mechanical, less excursion limited output, less tolerance to abuse and a cone that is rather fragile when compared to current high power driver designs like the B&C SW series. The 2242HPL is still a nice driver that can sound very good and it is very sensitive so it can perform well but it is a much older, less powerful design so it reaches its limits sooner with less power than many of the top drivers today and is far less tolerant of abuse. A lot of this stuff is not going to jump out from the typical spec sheet. There is a reason that JBL has drivers like the 2269H now.

 

[Chris Van Duker]

Re: Comparing LF drivers

As someone who has pointed the business end of an IT8000 at a set of 18SW115-based cabinets and used it in anger many times, I can attest that the heavy cone and large voice coil do indeed work as claimed. Not only have I never blown one, but the suspension feels like it's new. Even compared to the earlier generation TBX100 woofers, they're beasts.

-Chris

 

[Michael John]

Re: Comparing LF drivers

Thanks Bennett et. al. This info is very, very helpful.

Cheers,
Michael