Fun in a theatre - 3x Faital 18XL1800

Thought I'd share this project I've been working on.

Background: As part of my PA system, I've been running some 15" ported subwoofers for a while. Beyma 15P1200Nd, compact boxes tuned to 40Hz. They've generally been good to me, but there are some problems in the construction (drivers hitting the grilles, stuff like that - mostly my own fault) that meant I was never 100% happy with them.
Drivers have moved on since those Beymas - more Xmax, more power handling. As an example, I happened to have accumulated 3x Faital 18XL1800 drivers, which feature an Xmax of about 20mm, and long-term thermal power handling of 1600W. Short-term, they'll take in excess of 3KW.

I have a good relationship with a local theatre director, and he'd hired me for a show called Into The Woods, which features (among other things) a giant. When I suggested that I build some subwoofers capable of strong very-low-frequency output for the giant's footsteps to be felt as well as heard, he was enthusiastic about the idea.

After some messing around with Hornresp, playing with cabinets of varying sizes and complexity, I decided on sealed boxes for these drivers. The reasoning is this: at 40Hz, they'll keep up with the previous ported subwoofers, which were usually loud enough. If I need these new subwoofers to go louder, I can raise the low-frequency cutoff, gain efficiency, and put out more SPL. Similarly, if there's plenty of Rig For The Gig, I can throw more power at them and get more LF extension. The cabinets would also be very compact, which is a Good Thing in the live sound world.

Side-note: it may be of interest to see what sort of performance an off-the-shelf product gives. Here's one: https://products.electrovoice.com/binary/EKX-18S_F01U308106.PDF


The venue in question is small (100 seats), so I decided to push for maximum LF extension:
Frequency Response.png


I set off with a 12dB/octave lowpass filter at 18Hz (the red curve), which gave me a response that's pretty flat across the subwoofer range. Then, I decided to increase the fun-factor, and made it 18dB/octave (green). Afterwards, I increased the subwoofer level accordingly, so that the crossover to the main speakers (approx 70Hz) lined up well.

Power and processing was done by the rack of Powersoft T-series amps:

With the T602 bridged into two of the subwoofers, and the first two channels of the T604 bridged into the third subwoofer. Result: around 3KW per driver - enough to make use of the driver's generous Xmax for short-term peaks.


The giant's footsteps were played by timpani, as part of the show's pre-recorded music. Running that section through an RTA, the timpani had strong content down to 30Hz. While that's pretty useful (and would've been fine, I'm sure), these subwoofers can go lower, and I wanted to make use of that. Within the mixing desk, I took a digital split of the music input, applied a low-pass filter at about 80Hz, put it through an octave down, and mixed that into the subwoofer feed. The results were a lot of fun - even the lighting engineer was impressed.

This venue hadn't ever been subjected to a PA system like this, so all sorts of interesting rattles were apparent. By far the most distracting was a light fixture, so that one was removed entirely. With that gone, the quieter and more distant rattles contributed to the "there's something big coming this way" effect, so we decided to stop chasing perfection and have some food instead.

I found time to play a few bits of music (without the octave down), and the high output combined with serious LF extension was an excellent experience. Violence by Andy Stott was a good demo.


I think that about covers the story for the subwoofers.

A few other notes, if anyone's still interested:
  • The main PA speakers use a Faital 10HX230 per side. Ported boxes, tuned to 70Hz. I rather like those.
  • The subwoofers have a port at the back, where a tunnel can be placed to vent the rear of the driver's motor. I plan on doing some thermal testing to see if that's necessary, but for now the vent is blocked and heat remains in the cabinet. Not a problem for this show - the duty cycle was very low.
  • While the mixing desk and amp rack was placed side-of-stage, I mixed the show on a set of iPads from the control room. That kept the cable runs relatively short, and allowed me a good mixing position. During the more dramatic moments, the cupboards at the back of the control room were rattling.


In all, I've had a lot of fun here. It's good to be back.

Chris
 

Attachments

  • 20220527_223623.jpg
    20220527_223623.jpg
    420 KB · Views: 749
Nice report the subs are very compact. I get max SPL (one sub/1m):
20Hz: 107dB
30Hz: 114dB
40Hz: 119dB
using the sensitivity (95dB/1w) and rated power handling (3200W) I get max SPL I get 130dB/1m. In the sealed boxes this is achivable at a minimum frequency of 70Hz before it becomes excursion limited. The subs are very small though!

I have run sealed subs at high duty cycle for PA before (also with lots of low end EQ). Over a few hours the box itself gets quite hot and eventually I lost a driver due to adhesive failure in magnets so i would recommend venting them. The vent can be tuned below the pass band of the speaker using a long narrow diameter pipe pair which will effectivly exchange the air in the box.
 
  • Like
Reactions: chris661 and Flaesh
Looks like fun. Care to share some pics of the building process?

Thanks! They're a lot of fun. I'm afraid I didn't take any pictures of the build process, but it really wasn't exciting: 18mm MDF, wood glue, and a lot of clamps. I had the drivers installed in the baffles during the build, so that I could accurately space them off the back wall. Bracing the driver against the back wall (while making sure it can still exchange heat) is a neat way of bracing the two largest panels against each other.

Nice report the subs are very compact. I get max SPL (one sub/1m):
20Hz: 107dB
30Hz: 114dB
40Hz: 119dB
using the sensitivity (95dB/1w) and rated power handling (3200W) I get max SPL I get 130dB/1m. In the sealed boxes this is achivable at a minimum frequency of 70Hz before it becomes excursion limited. The subs are very small though!

I have run sealed subs at high duty cycle for PA before (also with lots of low end EQ). Over a few hours the box itself gets quite hot and eventually I lost a driver due to adhesive failure in magnets so i would recommend venting them. The vent can be tuned below the pass band of the speaker using a long narrow diameter pipe pair which will effectivly exchange the air in the box.

IIRC, I've got it so they can exceed Xmax by a few mm when hit with 3KW. These drivers don't produce any nasty sounds (apart from the motor exchanging a LOT of air) until around 50mm p/p, so there's a little bit of wiggle room there.

I have plans to connect the pole vent to the outside world via a short tunnel. Should keep things nice and cool, albeit at the expense of wind noises at the back.

The red curve is a lovely shape for home cinema use, boxes look very compact too! What's the volume of them ?
Cheers,
Rob.
About 70L internal, once you've taken away the driver's displacement etc.
I did try one of these in my living room, and it was pretty silly. Three would've been dangerous.


Thanks, all.

Chris
 
Yep, a bigger sealed box would mean less power needed to reach Xmax. However, for portable PA use, I had to balance power input vs size. The amplifiers I'm using are very efficient, and have enough output to ensure the driver's maximum potential is reached in these compact cabinets.

For home HiFi use, I think these drivers in large sealed boxes would be excellent.

Chris
 
Nice to see someone else going the sealed route for PA. Whilst hideously inefficient compared with more conventional ported and horn loadings, the sound quality is unsurpassed, and I now find it difficult to listen to the latter. (My system: 4 x sealed 21", 120l; 6 x 18" sealed, 50l, each speaker with its own amp.).
 
  • Like
Reactions: abradolf and metzb1
I have run sealed subs at high duty cycle for PA before (also with lots of low end EQ). Over a few hours the box itself gets quite hot and eventually I lost a driver due to adhesive failure in magnets so i would recommend venting them. The vent can be tuned below the pass band of the speaker using a long narrow diameter pipe pair which will effectivly exchange the air in the box.
I found the best way is to mount the speaker outside the enclosure, car audio style. Great cooling and any motor noise is swamped by the output.
 
I have run sealed subs at high duty cycle for PA before (also with lots of low end EQ). Over a few hours the box itself gets quite hot and eventually I lost a driver due to adhesive failure in magnets so i would recommend venting them. The vent can be tuned below the pass band of the speaker using a long narrow diameter pipe pair which will effectivly exchange the air in the box.

That's a great idea! I've been running four sealed 18's for years, as my mobile PA rig. Had to beat the snot out of them and keep them clustered in the center, but they "just" manage to cover most smallish venues. I've never lost a coil even when melting amps and tripping breakers, but this still sounds like a very good idea. I'm considering building new versions, with higher-powered drivers, and will definitely try this.
 
Chris,

I have two thoughts for you.

1: Regarding possibly making the enclosure larger. I assume you already know that there's a "sweet spot" for sealed subs, where the QTC is considered ideal. I've also read on some forums that a larger box will be less punchy. (though I can't personally confirm that.) When I was considering going larger than my current 3 cubic ft sealed boxes, the numbers in Winsd didn't improve much, so I decided that the small gain in output wasn't worth a sonic tradeoff. That might be different with those Faital drivers, but probably not.
IMO, cost aside, it's better to keep them small and use more of them. (Then you also get less distortion.)

But you went even smaller than this, correct? I can't imagine why. 3 cubic feet is pretty common for getting QTC of 1.0 with an 18", and that's still incredibly portable. You obviously know what you're doing, and that is a different driver, but I'd be very surprised if 70L is ideal, sonically.
If it is, I'd love to know, and also your thoughts on this.
----------------

2: Regarding a separate EQ setting for the tympani FX: Bag End's processor has a clever circuit that limits only the very low end (I assume below 40 Hz) when it exceeds a certain level. The threshold is user-adjustable. This of course lets you keep the HPF really low, all the time.

If your mixer has the capability, you could do the same thing with a frequency-dependent (or side-chain, or multi-band) compressor. I retired my Bag End processor a while back, as I now use the EAW version of a Drive Rack and put the same LF compensation into that.. Since I use a DAW as my mixer, I'm able to run a multi-band compressor plugin on my 2-mix, and it works even better than the analog version in the old Bag End electronics. Can that QSC load VST plugins?

Also, you probably already know this, but those Powersoft amps probably have some kind of " intelligent" limiting. That will add extra protection for the drivers, though of course is not the same as frequency-specific limiting.
 
Last edited:
The giant's footsteps were played by timpani, as part of the show's pre-recorded music. Running that section through an RTA, the timpani had strong content down to 30Hz. While that's pretty useful (and would've been fine, I'm sure), these subwoofers can go lower, and I wanted to make use of that. Within the mixing desk, I took a digital split of the music input, applied a low-pass filter at about 80Hz, put it through an octave down, and mixed that into the subwoofer feed. The results were a lot of fun - even the lighting engineer was impressed.


It might be fun to try using a subharmonic generator on the entire mix. Very carefully, of course!

Either that, or wire every seat in the venue with a mechanical thumper. 🙂
 
CableAddict,

There are many differing opinions on "ideal". Some like Qtc = 0.5, others Qtc = 0.7. My version is to size the cabinet where Xmax (or slightly above Xmax, if the drivers overload gracefully) is achieved at the expected maximum power input. ie, for these I was aiming for a little over Xmax (I chose about 25mm IIRC (rated Xmax = 20mm), since they're just about starting to sound bad at 50mm p/p) with 3KW input.

I don't know or care what Qtc came out as. I've minimised box size while making full use of the mechanical output of these drivers, which is perfect for portable PA systems.


An excursion limiter is on my to-do list. Unfortunately, the Powersoft amps can't do it. The mixing desk might be able to with some EQ/limiting on a subgroup etc, but it doesn't have a multiband comp. I could route through a DAW, I suppose.
So far, though, I've done it manually and that's been fine. I EQ the subs to be flat down to 20Hz, and then set a HPF if the sub amp is hitting the limiters frequently. Smaller shows are often left flat down to 20Hz. Medium-sized indoor or smallish outdoor shows might be 30Hz, and the louder ones (outdoor rave at an army base, for example) was 45Hz.

The subs have been a lot of fun, and I'm looking forward to using them for some shows this year.

Chris

PS - I do have a subharmonic generator set up on my desk, for music playback. Digital split => LPF one at 120Hz => octave down => Main Mix. Some of the old records have very little going on below 80Hz, and a touch of VLF can help with the boogie factor.
 
FWIW, I love that we're (partially) discussing sealed subs for PA use.

Despite the one obvious downside, there are so many upsides, but also so many variables. And there is very little discussion or experimentation on this to be found online, I guess because most guys just rule them out due to the efficiency issue.

Even driver selection is a near-mystery, since most designs are focused on HT use. For a PA, where you don't need anything below maybe 38 Hz, driver and box design decisions can be different.

MrKlinky now has me thinking about possibly trying some 21" drivers in compact boxes. Or maybe a dual 15", sealed, mounted in push-pull configuration. Either of those should give more dB per watt than a single 18". Still not approaching BR levels, but more dB than I have now. As long as I keep them center-clustered.

I think I can go to about 4.5 cubic feet, and still get four of them into my truck. Hmmm .......
 
Last edited:
  • Like
Reactions: Hydrogen Alex
An excursion limiter is on my to-do list. Unfortunately, the Powersoft amps can't do it. The mixing desk might be able to with some EQ/limiting on a subgroup etc, but it doesn't have a multiband comp. I could route through a DAW, I suppose.
So far, though, I've done it manually and that's been fine........

For long-term use with sealed subs, in a PA / DJ environment, I strongly recommend that you keep this near the top of your "to do" list !
Especially with $800 drivers. 😀

Maybe something like a Driverack would have this?

FWIW, I'm using Lab Gruppen C48:4's, and they have three different types of voltage and OL protection, which all run at the same time.
Great sounding amps, too. (Nothing wrong with the sonics of Powersoft, though!)
 
Man, I'm now totally into thinking about new sealed subs. The rabbit hole has got a firm hold of my foot.


I have a question about cabinet size vs processing, vs phase issues. (Or group delay, or whatever the correct term would be.)

Obviously, sealed subs need quite a bit of boost using a low shelf filter. And filters have phase issues. (Something I don't have a good grasp on, but I know this is a thing.) Plus we want the subs to be as phase-aligned with the tops as possible.

So would it be correct to assume that one benefit of a larger enclosure is that there is more acoustical LF extension? In other words, you need less gain on the low-shelf boost, and hence less phase anomalies?

Would it also mean that a processor with FIR filters would be preferable?
 
  • Like
Reactions: Hydrogen Alex
For live performance, FIR filters usually have too much latency at low frequencies so you're almost certainly staying IIR for this.

That being the case, the phase/GD effects of a driver+filter combination are what matter, and if you have 2 different boxes each filtered to end up with the same final frequency response, they will end up with the same final phase/GD as well.
 
  • Like
Reactions: KaffiMann