Unibox and general sub design question.

[CJ Paul]

I was posting on another forum about porting my current sub just to play around and learn because the sub box will be tossed in the next six months when I move and the driver either get sold, given away to my brother (if I'm really nice) or used in the IB. Any way, I was introduced to Unibox and spent about a week figuring out how it works. I've been playing with it ever since. So Is started a new thread at the other forum with some questions and trying to confirm my understanding but got no response. No beef about it, just the plain fact that I got no response and I'm still curious. Any way, here, esentially, was my question.

OK, I posted about porting my 121L sealed Tempest and eventially got directed to checking out Unibox, which I did. I've been running all kinds of models through it since then. The first thing I noticed is that if I port my current box to a reasonable tuning point I get almost no increase in output accross any significant bandwidth. Basically I could gain 1-2 dB from 30hz to the tuning point. It was then pointed out that if I model that port in my current box then I should look at the woofer excursion charts for sealed vs. ported. Holy Smokes! From that 30hz point to the tuning point, the excursion of the woofer drops, leaving me tons of headroom (of course it rises quickly again below the tuning point, but so does the sealed version) I've been introduced to a concept that I never even considered before. So, after fiddling around a bunch more, I present this concept that I have arrived at for anyone to validate if they choose.

This is all in theory as I've not decided whether to port this box or build a new one or leave it alone. But after the suggestion about looking at woofer excursion I decided to play around with wattage, as the woffer excursion charts showed that not ALL things are as simple as looking at the frequency response chart.

So I decided what I would do is determine that IF I ported the same size box (even though the SPL v. Frequency graphs were almost the same, given the same amount of power) how could I use the extra available excursion provided in the 35-15hz range to my advantage. Well, I could add more power. See, with a simple 200 watts, my sealed design would reach its excursion limit at 15hz. A reasonable number. The excursion graph of this sealed sub ramps up exponentially as frequency drops, so as I add more power, it causes the sub to bottom out at exponentially higher frequencies. With a ported sub, there is a big dip in excursion around the tuning point, which means I can add more power. I could hit 400 watts and still have the excursion limit be reached at 15hz. That means an even higher maximum SPL. So any way, I saved off the excursion graphs and then I made a comparison of my current sub with 210 watts where the excursion limit is reached at 15 hz and a ported sub in the same box where the excursion limit is reached at 15 hz so that you can see the net effect of porting the box AND gaining power handling.







Does this make sense, or am I making incorrect assumptions?

Since that original post, I've been modeling different drivers for IBs and have run accross the same concept. Quite a few driver combinations will give the exact same output given the same total power (i.e. four shivas vs. two tempests in an IB with 250 watts of power) BUT if you look at the excursion charts, one of those two might have more excursion left so if you adjust the watts until they run out of steam at the same frequency, one of the setups gains an advantage in output. Is what I am doing by setting a point where I want to see excursion limitations and basing my maximum wattage and thus SPL on this point valid at all?

Thanks for reading my long post.

CJ

[JonMarsh]

CJ,

You're on the right track here; as I think your're discovering, sub design is all about traeoffs, including box size, power handling, and especially outright power handling at very low frequencies.

The simplest and easist thing to do is a nice sealed box or IB - they have excellent time domain characteristics without doing hardly any work on the designers part (as long as you don't pick a driver with screwy T/S parameters), but they're both limited in maximum acoustical output by the swept volume issues - Xmax time Sd.

Bass reflex enclosures (originally known as Helmhotz resonantors, after their inventor) all use a reonant box system to augment the rear wave. Unfortuantely, that means energy storage, and some compromise of the group delay and transient response. An intelligently designed ported system seems to offer a reasonble tradeoff at times when you want to extend the low frequency reach and power handling. Passive radiators are another form of this augmentation- they seem to usually wind up with even worse energy storage and group delay issues, so I generally stay away from them.

Both the frequency response plot of the system, as well as the peak cone excursion for a given power input are fairly important parameters. Often, when you tune the box quite low, you may wind up with a saddle in the output capability cuve between the upper frequency range and the box tuning; you have to consider that, too, becuase it might not make sense to have a sub with 110 dB capabity at 20 Hz and only 100 dB at 35 Hz. Depends on your goals...

I'll usually play with the box size sealed and ported, and look at the alignments as you raise the nominal power level, until I'm crowding the Xmax limits.

Consider an example I was just working with today- the Extremis 6 midwoofer (long throw- 12 mm Xmax). Let's assume for evaluation a maximum continuous input power of 50 watts. In a sealed configuration, below 35 Hz it won't handle that much power- it's Xmax limited, and the output drops as a straightline function- at 25 Hz the maximum output capability is about 89 dB at Xmax limit. Righteous for a 7" driver, but not going to shake the house down. Box size doesn't affect his output- just how much power it takes to reach that point- smaller box, more power.

Now, let's drop it into a vented 34 liter box tuned to 23-24 Hz. Boy, do thigns change. Besides the response now being down only about 3 dB at 24 Hz from 100 Hz, peak output is up from 89 dB to over 100 dB. That's a lot of gain in headroom. Of course, it falls off below that, but it doesn't drop to the sealed box level until about 17 Hz. Worthwhile tradeoff? I'd say so.

For big 15" subs, I usually look for tunings in the 15-18 Hz area. This is in the bottom end range of "things that go boom" in movies, and above 30 Hz or so, the impact on music signals is minimal- which is good.

ThomasW built a design around that concept which we call the Aerial Stryke 15, using an HE-15 woofer (yeah, pre XBL2 tech!); once you trim the EQ on it to flatten the response a skosh at both extremes, it's pretty awesome what it does in the low end with 25 mm of Xmax and a 6" flared port... shakes a big home pretty well.








~Jon

[CJ Paul]

Thanks Jon. I've seen that subwoofer construction project on the web before. I guess my biggest question was the part about evaluating drivers for output in a large room (or to have headroom and potentially decrease distortion). If I pick a reasonable frequency (20 or 15 hz) and then adjust the wattage until each woofer (or set of woofers) reaches xmax at that frequency and THEN check the frequency response v. SPL graph, does that put the woofers on a level playing field?

I suppose the key is, one woofer might have more headroom at 40hz and another has more down low so with my method, it might end up making the one with more room to breath down low look like the woofer with better output, but frankly, most of my models have shown that similar woofers have similar curves in FR and excursion so, as in my original post, if 2 tempests vs, 4 shivas in an IB have the same output with 100 total watts, but the tempests can handle 150 and the shivas cannot (say for example xmax is reached at 50hz or something) then the tempests SHOULD have more output, right?

[Scott Simonian]

Well, it should have more, clean output.

[JonMarsh]

Thanks Jon. I've seen that subwoofer construction project on the web before. I guess my biggest question was the part about evaluating drivers for output in a large room (or to have headroom and potentially decrease distortion). If I pick a reasonable frequency (20 or 15 hz) and then adjust the wattage until each woofer (or set of woofers) reaches xmax at that frequency and THEN check the frequency response v. SPL graph, does that put the woofers on a level playing field?

I suppose the key is, one woofer might have more headroom at 40hz and another has more down low so with my method, it might end up making the one with more room to breath down low look like the woofer with better output, but frankly, most of my models have shown that similar woofers have similar curves in FR and excursion so, as in my original post, if 2 tempests vs, 4 shivas in an IB have the same output with 100 total watts, but the tempests can handle 150 and the shivas cannot (say for example xmax is reached at 50hz or something) then the tempests SHOULD have more output, right?

Oh yes, they definitely will have more output. And for drivers like the TEMPEST's and SHIVA's there is a pretty reasonable correlation between the power level handled and maximum output level.

What a project like the one I've shown again does, in simple terms, is provide up to 6 dB or more in maximum LF headroom compared with a sealed box. More complex to build, more complex to design, but sometimes worth it.

OTOH, if you have the room for a nice IB setup or several sealed boxes with a Q of 0.5 and a good long throw driver, other than the extra power and room needed, that's a nice way to go. And easy to design.

Now, imagine what I go through trying to figure out the driver count and type needed for reasonable playback level with dipole midrange, woofers, and low bass- takes a lot more drivers, and a fair amount more head scratching, do the dipole cancellation with falling frequency, dependent of course on the baffle dimension.

[cjd]

CJ:

Part of the point of my original suggestions with regard to excursion vs. tuning is that you decided you wanted to explore porting your existing sealed box. Given the volume and the fact that you already have EQ, plus the fact that an IB is in your future, I went on the assumption that you felt it was lacking low end impact. And the best way to get this with what you have is to increase power handling (or really, to increase your ability to dial in EQ).

It's still a game of trade-offs, but you may choose to augment output level by doing something like, say, using some higher excursion drivers, using more drivers, etc. Understanding all your options is always a good thing though!

C

[CJ Paul]

Thanks guys, and thanks Chris. I dont know if you saw my other post. Now I'm just learning. I have no clue what my next project will be, could be the IB in six months, could be a totally different enclosure in 2 weeks. Point being simply that I want to learn so I can answer my own questions and help others. I've been astounded buy the information pull from thin air about this driver or that design or do this in that situation etc. Even in the last week of playing with the software, researching parameters, etc. I've learned how these people do it. I'm no expert for sure, but I'm discovering that it's not magic either.