IB makeover..........

[Rodny Alvarez]

Well....... some of you guys are going to love this and some are going to ...
One of the reasons I'm doing this is because the bass in the attic, it travels all over the house and my wife , so by doing the SLLT's it will stop must of the bass traveling so much in the attic , plus I'm feeling a little
So here is the deal.........Steve Callas is the brain behind this design.
I'm going to build a box behind the IB, 35cft with two ports 6" by 26" long on each side of the room, and adding another EP 2500
Here is on paper.......






What you guys think????????????

[SteveCallas]

I'm obviously all for the idea In that first picture, it's comparing max output and extension without exceeding excursion limits above tuning between this design and a traditional dual RLp-15 LLT. In the picture I am attaching, it's the proposed design vs Rodny's current IB, both at max output before exceeding excursion limits at 11hz.



As you can see, BIG difference. I've been suggesting adjacent room LLTs like this to those wanting to build IBs for a while now, but most have been scared to try it - I can only guess because it's not common....hasn't been done yet as far as I know. What people don't realize is that there is always room for improvement with a port. No matter how low the natural response of an IB goes in room, if you port it and tune it to the -6db point, you gain more clean low end as well as more usable headroom throughout the sub's range while maintaining a flat FR. You also avoid having to use lots of EQ to boost the low end.

I design LLTs with no port resurgence to better match with room gain, but in this case, the tuning is so low that I felt a 1.5db resurgence would actually be good to help counter some of the effects of electronics rolloff. I can't wait to see how it turns out.

[Sonnie]

Oh boy... I am going to have to make a trip to Millbrook....

This is gonna be really interesting.

[Ilkka]

SteveCallas wrote:

I design LLTs with no port resurgence to better match with room gain, but in this case, the tuning is so low that I felt a 1.5db resurgence would actually be good to help counter some of the effects of electronics rolloff. I can't wait to see how it turns out.

Also notice that the peak port velocity at tuning will be around 48 m/s (input power=2800W). According to Collo's research, core limit for 6" port at low frequencies is around 26 m/s, so there will be quite a lot of port compression. Adjust Qp value to 5 in order to get the peak port speed to 26 m/s. That will show the real max output around tuning frequency.

You should also add a some sort of HP filter, unless you want to exceed the Xmax below ~11 Hz. I know that's not the Xmech but that's where you limited the IB sub.

This is more close to real world difference when Xmax is not exceeded at any frequency.

SLLT: 3500W, Qp=10 (filtering also decreases output, so Qp=5 not needed), BW 1st order HP @ 10 Hz (electronics roll-off), BW 2nd order HP @ 9 Hz (to prevent bottoming below tuning).

IB: 2300W, BW 1st order HP @ 10 Hz (electronics roll-off).



Also notice the difference in group delay, around tuning the SLLT will have around 90 ms more group delay. Whether it's audible can be dabated. Just to inform that the performance gain doesn't come without any expenses.



If you want to squeeze more performance out of the IB, just add a 1st order HP @ 10 Hz and up the power to 3500W (equal with SLLT).



Just to keep things fair...

[SteveCallas]

My modeling for the ported version was done with 2600 watts, creating a max port velocity of 33m/s just below 11hz (are you using 70 cubic feet?). If you knock power down to 1500 watts, speed stays under 26m/s at all frequencies and max output is still 120db+. Including room gain, I highly doubt he will ever come close to using much more than 1500 watts down that low. That said, I did advise to use fewer larger ports, - two 10" looks great - but Rodny wanted to stick with 6".

I gotta disagree with you on the highpass Ilkka. Much higher tuned LLTs with significantly less output capability have yet to bottom with the lowest bass scenes played at near reference levels, so I see no reason to change course. Excursion demands beyond xmech should tie in pretty closely to his ultimate electronics rolloff, keeping him plenty safe.....and again, I don't really foresee too much 120db prior to room gain output coming into play that low

[Ilkka]

SteveCallas wrote:

My modeling for the ported version was done with 2600 watts, creating a max port velocity of 33m/s just below 11hz (are you using 70 cubic feet?).

Yes, 70 cu ft, four 6" x 26" ports. With 2600 watts, peak port velocity is ~46.7 m/s at ~10.7 Hz. Which of us has his simulation wrong?

If you knock power down to 1500 watts, speed stays under 26m/s at all frequencies and max output is still 120db+. Including room gain, I highly doubt he will ever come close to using much more than 1500 watts down that low. That said, I did advise to use fewer larger ports, - two 10" looks great - but Rodny wanted to stick with 6".

I gotta disagree with you on the highpass Ilkka. Much higher tuned LLTs with significantly less output capability have yet to bottom with the lowest bass scenes played at near reference levels, so I see no reason to change course. Excursion demands beyond xmech should tie in pretty closely to his ultimate electronics rolloff, keeping him plenty safe.....and again, I don't really foresee too much 120db prior to room gain output coming into play that low

I knew you were going to say this. But whether 120 dB etc. is enough for his needs isn't the point. This is about how much SPL can each system (SLLT/IB) output before exceeding some specific excursion (Xmax, Xmech - doesn't matter). You limited (input power) the IB system to the point where it exceeded the Xmax at 11 Hz. Then naturally you have to do this also to SLLT. Of course I agree with you regarding the electronics roll-off (my own simulations included it), but when comparing two systems, you have to use the same rules, if you know what I mean.

Of course the ported version will have more output, especially in the 10-30 Hz range, but when optimizing both systems, the final difference isn't as large as you originally showed. And the extra output doesn't come without any expenses (added group delay and ringing around the tuning frequency, and port noises near maximum output level).

[SteveCallas]

Ilkka wrote:

Yes, 70 cu ft, four 6" x 26" ports. With 2600 watts, peak port velocity is ~46.7 m/s at ~10.7 Hz. Which of us has its simulation wrong?

I use RMS as opposed to peak, that would be the cause of the difference.

This is about how much SPL can each system (SLLT/IB) output before exceeding some specific excursion (Xmax, Xmech - doesn't matter). You limited (input power) the IB system to the point where it exceeded the Xmax at 11 Hz. Then naturally you have to do this also to SLLT.

That's exactly what I did. The comparison I attached has both systems reaching xmax at 11hz.

Of course I agree with you regarding the electronics roll-off (my own simulations included it), but when comparing two systems, you have to use the same rules, if you know what I mean.

I'm not sure I follow - I didn't add a highpass or anything to the sealed design. What I was showing was the likely max output before room gain of both systems if they both reached overexcursion at the same frequency.

but when optimizing both systems, the final difference isn't as large as you originally showed.

Replace the four 6" ports with two 10" ports that are each 34" long. Now, even if peak port speed is used, it's well within the limits of a 10" at all frequencies, so there shouldn't be any port compression. What else are you taking into effect to lessen the gap?

And the extra output doesn't come without any expenses (added group delay and ringing around the tuning frequency, and port noises near maximum output level).

I'd call both of those non issues at 12hz, even if using the 6" ports. 10"s would make it that much more of a non issue

[Ilkka]

SteveCallas wrote:

I use RMS as opposed to peak, that would be the cause of the difference.

Why?

That's exactly what I did. The comparison I attached has both systems reaching xmax at 11hz.

Well yeah, but look at the shape of the excursion plots: the SLLT shoots sky high and will reach the Xmech (I believe it's close to 40mm) almost instantly below 11 Hz, while the IB won't even exceed 30mm with 1200 watts. The SLLT exceeds the Xmax below its tuning already with 700 watts.

I'm not sure I follow - I didn't add a highpass or anything to the sealed design. What I was showing was the likely max output before room gain of both systems if they both reached overexcursion at the same frequency.

You seem to be neglecting the fact that any ported sub without a proper HP filter WILL bottom out below its tuning frequency. And don't tell me that he doesn't need that kind of levels or that there will be some electronics roll-off to help - it doesn't matter when doing a comparison like this.

The main thing is that you allowed a ~49mm excursion below 11 Hz for the SLLT, while the IB stayed below ~30mm. That isn't a fair comparison in my book. If you want them to be comparable, you'll have to use some HP filtering with the SLLT.

I'm sure both systems will be more than enough for his needs, it's just that your simulations of the proposed difference weren't fair to both systems. The IB can take much more power than what you showed before reaching the same excursion below the tuning frequency of the SLLT, especially when noticing the electronics roll-off.

This is the real world difference when both systems are optimized for max SPL at all frequencies and Xmax is not exceeded at any frequency.

Replace the four 6" ports with two 10" ports that are each 34" long. Now, even if peak port speed is used, it's well within the limits of a 10" at all frequencies, so there shouldn't be any port compression. What else are you taking into effect to lessen the gap?

Yes, I would definitely use two 10" ports in order to get the maximum performance around the tuning frequency. There's no reason to knock it down with too small ports that will start to chuff when approaching max output.

I'd call both of those non issues at 12hz, even if using the 6" ports. 10"s would make it that much more of a non issue

Even if port noise isn't an issue for you, you will still have to check whether the port velocity exceeds the max core speed. Otherwise your simulation is flawed since the port can not move that much air i.e. the output will be cut down around the tuning frequency.

[Ilkka]

The previous plot was with four 6" ports, this one is with two 10" ports.

[Sonnie]

That looks pretty significant to me.... at least from 11Hz up to about 25Hz.

Rodny... you gettin' all this. You need to go with the 10" ports... No reason to do it halfway...

That Bradley Plumbing, IIRC, had a bunch of leftover pipe that they will probably let you have. Go to the back, not the front office and talk with the guys back there. They had all kinds of pipe in the yard. Or you might find some sonotube type stuff around Montgomery somewhere.

[SteveCallas]

Ok, I gotcha, so you want the excursion use below 11hz matched more closely to represent a more fair comparison. How's this for a nice compromise?

Ilkka wrote:

Why?

Well I don't think he'll be listening to 12hz sine waves - I'd imagine RMS would be a better representation of actual material, no?

[Ilkka]

SteveCallas wrote:

Ok, I gotcha, so you want the excursion use below 11hz matched more closely to represent a more fair comparison. How's this for a nice compromise?

You can play with the numbers, but the difference isn't that large when both systems are optimized. Something that most of your LLT vs. anything comparisons are missing. And I don't want to bash any alignment, I just want to keep things fair.

Well I don't think he'll be listening to 12hz sine waves - I'd imagine RMS would be a better representation of actual material, no?

All the signals are some sort of variations of basic sine waves. I wouldn't use the RMS value for accurate simulations.

[SteveCallas]

Ilkka wrote:

You can play with the numbers, but the difference isn't that large when both systems are optimized.

Aren't that large? It's up to ~10db at some points. Though different drivers are in play, porting is the main reason why a 12hz tuned SVS Plus can provide as much distortion limited output in the low teens as Velo DD18.

[Ilkka]

SteveCallas wrote:

Aren't that large? It's up to ~10db at some points.

Sorry about the bad wording, I meant that the difference in SPL at different frequencies isn't as large as your simulations show when both systems are properly optimized. For example your simulation shows around 7.5 dB difference at 20 Hz when the correct value is 3.5 dB. At 50 Hz you show 5 dB when the correct value is 1 dB. So in other words you tried to inflate the difference in favor of SLLT. As I have said many times, I just want to keep things fair.

Though different drivers are in play, porting is the main reason why a 12hz tuned SVS Plus can provide as much distortion limited output in the low teens as Velo DD18.

I've never claimed that porting wouldn't help to boost the output around the tuning frequency. Are you forgetting who you are talking to?

And although pretty much OT, but since I have heard both subs in question (even owned the other), I can tell that there is no question which sub sounds subjectively cleaner and more powerful at any frequency, at any output level. SVS' port noise becomes very audible at those output levels (THD test doesn't pick that up). Plus if you look at AV Talk's measurements, you'll notice that when pushed any higher than 95 dB average (90 dB absolute SPL) below ~17 Hz, SVS' THD will shoot sky high, while Velo maintains its composure much better. Also AV Talk's measurements don't show what happens below 15 Hz, but based on my own measurements, I'd predict that SVS' "blue line/90 dB" THD will grow extremely rapidly below 15 Hz (already before the small signal level tuning frequency at ~11 Hz due the fact that the port will start to compress heavily).

[Rodny Alvarez]

If I use the 10" port it will have to real close to the floor,is this a problem?? at the bottom of the box depth is 48" and from 8" up is less, because of the angle.
The port will be about 14" from the back wall, is that enough??
The only thing with the 10" port is...... my kids might hide inside the box

[SteveCallas]

Ilkka wrote:

For example your simulation shows around 7.5 dB difference at 20 Hz when the correct value is 3.5 dB. At 50 Hz you show 5 dB when the correct value is 1 dB. So in other words you tried to inflate the difference in favor of SLLT.

I'm not sure what other changes you are making. I believe the new power and high pass filter adjustments I made make it about as fair as it can get The IB isn't using 70 cubic feet. Are you including some EQ boosting down low?

Ilkka wrote:

And although pretty much OT, but since I have heard both subs in question (even owned the other), I can tell that there is no question which sub sounds subjectively cleaner and more powerful at any frequency, at any output level. SVS' port noise becomes very audible at those output levels (THD test doesn't pick that up). Plus if you look at AV Talk's test, you'll notice that when pushed any higher than 95 dB average (90 dB absolute SPL) below ~17 Hz, SVS' THD will shoot sky high, while Velo maintains its composure much better.

I've got no problem with any of that, one is using a significantly larger driver with more power. I just highlighted the comparison to show how big an effect porting down low can have.


And please don't take any of this back and forth as disrespectful or anything like that - you have contributed more to the overall knowledge and understanding of subwoofer performance on these forums than anyone else I know of, by a pretty long shot Until we are seeing the same data though, I see the differences as being quite significant.

[SteveCallas]

Rodny wrote:

If I use the 10" port it will have to real close to the floor,is this a problem?? at the bottom of the box depth is 48" and from 8" up is less, because of the angle.
The port will be about 14" from the back wall, is that enough??

If you can manage 10" of clearance in all directions from the openings, you should be fine

The only thing with the 10" port is...... my kids might hide inside the box

Would that be kind of like a "time out" punishment? Ok Jimmy, you know what you did wrong, get inside the sub. Lol, sorry, I know that's heartless, but picturing a little kid sadly marching to and climbing in a 10" port is kinda funny, in a sick sort of way