XLS10s in a box

[sploo]

Hi All,

I have a set of Peerless XLS10 drivers, and I'd like to make a couple of enclosures - preferably with two in each box.

Long term, I plan to build a pair of speakers - probably two way, with a 7" midbass and tweeter, and use the subs to make them a pair of three way speakers.

In the short term, I'd probably just use them to augment my existing floor standing two way speakers, likely as a pair of mono subs.

BTW My aim is sound quality, rather than ultimate output.

I've been playing around with bassbox pro, and just can't seem to find a setup that gets a decent looking flat response. I was looking at this app note from Peerless (http://www.tymphany.com/papers/appxls2.htm). If I start a project with a single driver in a 20 litre box, the normalised response is -15dB at 25Hz, not -9.3dB, as shown on the first graph on the app note - so I'm guessing I'm doing something dumb.

I've also found this (http://www.bkelec.com/HiFi/Sub_Woofers/XLS200.htm). How on earth are they getting -6dB at 17Hz in an 18 litre sealed enclosure? EQ?

I have no strong feelings about going for sealed, sealed + EQ, ported etc., and would welcome any advice.

[rabbitz]

Have a long think about using the XLS10 in a 3 way speaker as they are a subwoofer driver. If you are after sound quality for music, you are better off with a woofer designed for 3-way use as the requirements are completely different.

There are several local builders who independently came to the same conclusion so I'm not alone. You may lose some SPL, power handling, and a higher F3, but the sound will be better IMO.

I'm sure that there are builders who have had success with the XLS for music but I imagine crossed over very low or added some other tricks. BTW, I do have a 10" XLS in use, but not for music.

Just a thought from one designer and builder.

[sek]

The XLS series woofers are best suited to be used with their passive counterparts. The reason is that their parameters don't suit a good closed tuning, and that a vent would have to be bigger than the box volume.

I use an XLS10 with the 10" passive radiator in 25l. Works like a charm since years. But not for music, as Rabbitz mentions. They are not designed to play precisely, they are designed to go down low within a small box enclosure. Group delay is the main problem, it's just too large. They are "slow" compared to conventional woofers (in their matching enclosures).

XLS drivers are good for infrasonic extension, home cinema LFE channels and car applications.
I partially use mine for music (2.1), though, because it can be fun.

Cheers,
Sebastian.

[richie00boy]

Gordon can you post up what Thiele-Small parameters you are using and I'll run a check with my software on the response in a 20 litre box.

As for BK, I suspect they are quoting a typical in-room response. Most sub makers do this.

[sploo]

Originally posted by rabbitz
Have a long think about using the XLS10 in a 3 way speaker as they are a subwoofer driver....

...I'm sure that there are builders who have had success with the XLS for music but I imagine crossed over very low or added some other tricks. BTW, I do have a 10" XLS in use, but not for music.

Thanks for the info rabbitz. My plan would be to cross them over somewhere between 100-150Hz, so they would be used for only the lowest frequencies.

Siegfried Linkwitz is using a pair of them (open baffle) in his Orion speakers - and I certainly wouldn't argue with him if he thinks they're good enough for music. He is using a fair bit of EQ BTW.

Originally posted by sek
The XLS series woofers are best suited to be used with their passive counterparts...

...I use an XLS10 with the 10" passive radiator in 25l. Works like a charm since years....

...Group delay is the main problem, it's just too large. They are "slow" compared to conventional woofers (in their matching enclosures).

Thanks also Sebastian.

Many (most) of the XLS10 box designs I've seen do use a passive radiator. Perhaps I will have to take a look into that. Would you recommend just using one XLS10 with one passive (I'm not after monster bass)? If I were to use two, would you match that with two passives?

I'm not convinced about group delay problems. I've run a few ABX blind tests using sound clips that have been modified to mimic pretty severe group delay problems (30ms). I've not found anyone that can tell the difference between a 'perfect' and group delay file, and that's with high quality headphones and no room issues.

Originally posted by richie00boy
Gordon can you post up what Thiele-Small parameters you are using and I'll run a check with my software on the response in a 20 litre box.

As for BK, I suspect they are quoting a typical in-room response. Most sub makers do this.

Thanks Rich - parameters follow. Good point about the in-room figure. A bit cheeky if you ask me, as different rooms can have wildly different affects on a speaker!

Driver Properties
Name: 830452 (SWR269) (2001)
Type: Standard one-way driver
Company: Peerless (Denmark)
Comment: 269 SWR 51 147 NX ALP 4L XLS
No. of Drivers = 1
Fs = 18.9 Hz
Qms = 2.63
Vas = 89.7 liters
Cms = 0.52 mm/N
Mms = 135.3 g
Rms = 6.12 kg/s
Xmax = 12.5 mm
P-Dia = 212 mm
Sd = 352 sq.cm
Qes = 0.18
Re = 3.4 ohms
Le = 4.3 mH
Z = 8 ohms
BL = 17.5 Tm
Pe = 300 watts
Qts = 0.17
2.83-V SPL = 88.4 dB

[richie00boy]

Didn't realise how darn low Qts is on that driver. Must be one of the lowest I've ever seen.

I get results nearer to you. My software is based on Bullock's work, which is accepted as sound.

[sploo]

Hmmm. Probably correct then. I recall checking those values (from bassbox) against a Peerless datasheet.

Perhaps some EQ would do the trick - with two in a box that would lessen the chance of running out of excursion at decent volumes with some low end dB boost applied.

[sek]

Hi,

If I were to use two, would you match that with two passives?

I would. One passive for each active.

But you don't neccessarily have to, because the passive radiators can be tuned with added mass (bolted on behind the dust cap from the reverse, a tapped hole is prepared). Peerless (Tymphany) has got some application notes to give advice. It depends on maximum SPL and lowest f3 you want to achieve, of course. There's no free lunch, though, as efficiency goes way down with added resonator mass.

Oh and BTW, there's also a 12" passive that might suit your needs with two active 10" drivers. Who knows...

Siegfried Linkwitz is using a pair of them (open baffle) in his Orion speakers

Don't get us wrong on this, the XLS10 isn't slow on it's own. It's slow in low-f3 resonant tunings (such as vented or with a passive) due to it's combination of high moving mass and strong motor. BTW, if I'm not mistaken (which I might be) Linkwitz is using the XLS10' s bigger 12" brothers XLS12.

And he's of course using extensive EQing on everything he builds. Who, if not he?
To me he's pretty much of an idol in that regard.

Speaking of which, the XLS10 with a 10" passive (and the right tuning) turns out to require no EQ at all. Tymphany's app note nevertheless holds valuable excursion vs. SPL data and EQ recommendation if need be. I tried boosting/shelving my XLS10 with a DCX2496 and didn't manage to bottom it out.

My plan would be to cross them over somewhere between 100-150Hz, so they would be used for only the lowest frequencies.

I honestly wouldn't consider 150Hz the lowest frequencies...

I resorted to filtering my XLS10 at 80-90Hz, 24dB/oct. LR, starting from an initial 120Hz cutoff when I built them.

Those heavy cones with their large excursion can yeld significant harmonic distortion and impose ugly modulation on the upper bass/lower mids.

Bottom line of my experience with the XLS series: if you want to have 'uncoloured' fun with them, you want to build 'em solid and drive 'em hard (efficient amping recommended, but anything from gainclone to UCD works). In order for them to sound good, you want to limit their power bandwith to about two octaves and filter 'em steep! The latter is mandatory for good sound at higher loudness levels.

Cheers,
Sebastian.

[Cal Weldon]

Try not to use less than a 1 to 1 ratio with your passives. Adding mass will help but you will still work the passive too hard if you try two woofers, one passive all the same size. If you are thinking two woofers even a single 12" passive can be a bit small. Consider two 10" or one 15" passive for two 10" woofers.

[sek]

That's actually right. Normally it is mandatory to use a passive radiator utilizing a significantly larger surface are than it's active driver. This is because at and below the resonance point the passive radiator can move out of phase. A passive of the same size would then exhibit a DC buildup, thus be driven out or in and collapse.

Peerless' solution to this problem was making the passive radiator's suspension a little stiffer than the active driver's one. As the passive has at least the same moving mass as the active driver, higher stiffness together with the same surface area warrants an excursion limit (and thus protection) of the passive at the expense of efficiency. This way they ensure that a 1:1 relation of active and passive radiators will work safely. They explicitly designed the passive to match the visual appearance of the active driver...

Though, you can overload the passives in a 1:1 setup by tuning the enclosure way too low (i.e. increasing box volume or passive radiator weight too much). That's right.

Some early adopters had reported about ripped glue joints separating the spider from the basket during excessive mechanical overload. But those were bad glues in early production series that could even be fixed by glueing back on with appropriate epoxy. Other than that it's very difficult to wreck a properly tuned and moderately amped XLS, you just have to live with low bandwith and low efficiency.

Thanks for the hint, Cal.

Cheers,
Sebastian.