Rybaudio

I wouldn't even bother with a subsonic. If the box is tuned to 18 Hz, excursion probably isn't a problem until you're below 15 Hz and there's hardly anything down there. If the sub can't hard bottom I would just let the suspension do the work those few times that you might have a peak at 10 Hz... hell, the amp itself my roll off down there and do the job for you. Secondly, I wouldn't use the Behringer "low-cut" filter under any circumstances. In the 2 amps that I tested, turning the low cut filter on not only cut out the very lows, but it put a 12 dB peak in around the cutoff!

Oh yeah... FYI, - the enclosures are covered in spray-on bedliner - the foam panels are placed to absorb (at least from 1k up) the first few early reflections - the sub in the back of the room plays about as loud as the main speakers from 80 Hz down, but it rolls of gently before it is cut sharply at 150 Hz, so it blends into the mains. It's also has a little delay on it to drive a major room resonance out of phase with the mains and to simulate reverb above that. - the subs on the sides of the room play at a significantly lower level than the mains and the rear sub and have quite a bit of delay on them (30 ms I think) to simulate LF reverb. I'm still playing with this.

My setup: 50" Plasma Media PC Behringer DCX-2496 American Audio VLP300 (2) Behringer EP2500 L&R: 2-way, B&C DE200 on a 15" OS waveguide and an an Eminence Kappalite 3015 (3.5 cubes ported tuned to 28 Hz) crossed (LR8) at 900 Hz, mids play all the way down subs: 3 JBL GT1214D (2.5 ported tuned to 28 Hz) playing 150 Hz-down. Overview of the listening room Media PC: Left Speaker Right Speaker A sub another Third sub in rear of room with amp rack

LOL... no beating around the bush there.

This rule of thumb doesn't apply to ported boxes vs sealed. Above tuning the ported enclosure "looks" sealed and hence the HF behavior is the same as a sealed box. On the low end the bandwidth is extended by the additional resonance of the port/box. Thus the total bandwidth is greater, even though you've added an additional resonance.

Thanks... nice to see some numbers.

I was under the impression that the output impedance of most car audio amps was very small. What kind of efficiency difference are we talking here (for say, 1 ohm vs 4 ohm)?

4 15s and 3 18s will have roughly the same cone area. If the motors are fairly similar you should be looking at similar performance. The issue IMO is cost- 18s are harder to make/get cheap but you need more 15s. I'd do whatever is cheapest in total. If they are close to the same, I'd go 3 18s just because they're 18s.

Efficiency of the speaker or of the amp? I see no reason why it would lower the efficiency of a speaker... for example, the efficiency of a dual coil sub is the same with the coils in series and in parallel. With the amp I could see the efficiency dropping a little because the output impedance of the amp is a larger fraction of the total load impedance but that's pretty minor. Are there other effects I'm not thinking of?

Do you mean BP1200.1? I used to own four of those... they're cheap and do a little over the rated power. I did have one burn up under strange circumstances though... I'm not entirely sure what went wrong. I've heard of others having similar problems. Maybe yours is not prone to the same problems being it's survived this long.

You should space the midrange and tweeter as tightly as possible. The reason is that this minimizes the path length difference when you are off-axis to the speakers. You may not sit off axis, but sound radiates in that direction and bounces around the room, eventually arriving at you. I would worry less about spacing to the edges and rounding the edges, though for the former you might try to put the tweeter a different distance from each edge. For an amp, I'd say go as cheap as possible... a Behringer a500 would work well.

Here's a picture showing what has been desribed above. If the disc in the picture is the speaker's cone, the O-looking letter (the greek letter theta) is the off-axis angle to the listener at point P.

A while back I picked up 4 MJ18s for use in my home system and I was recently finally able to get them on a Klippel analyzer to see how they perform at small and large signals. I'm posting the information here for anyone else who cares. First off, the unit to unit variation was small. For the linear parameters I'll list the average and for the nonlinear curves I'll show one of them; the rest are similar. The linear parameters are determined in three circumstances: the driver at very small signal, the driver at large signal but cold, and the driver at large signal and warm (after playing some time). Which of these to use really depends on the application; in my case I use the small and large + cold because in my home setup they don't see much power. If you're main usage is street beating then large+warm is probably better. In any case, here are the 7 electromechanical linear parameters for those cases. Thiele/Small parameters can be calculated from these. Small signal + cold Mm: 292 g Rm: 6 Ns/m Cm: 110 um/N Bl: 20.15 Tm Re: 3.45 ohms Le: 2.3 mH Sd: 1030 cm^2 Large signal + cold (same except) Cm: 150 um/N Le: 2.92 mH Large signal + warm (same as Large/cold except) Re: 4.2 ohms For the nonlinear performance there is a lot of data available but it can be confusing and/or misleading if you don't know how to interpret it. The basic issues are the variation of parameters over cone displacement and current through the coil (because these things cause nonlinear distortion) and the variation of parameters depending on the power dissipated in the device as this causes thermal compression and performance variation with time. I'm not going to go into variation with current through the voice coil but I'll tell you that it's a minor factor in this case. For variation versus excursion I'm going to display the Bl, Cms, and Le versus displacement as they give a rough idea of usable limits. These are the three curves below. Bl is symmetric so the coil is centered pretty well. It's down to ~70% at 10 mm and ~50% at 15 mm, which are roughly the "clean" and "usable" limits. Cms is very assymmetric which reflects the spider being dished by about 10 mm, but it doesn't really matter because the variation over 15 mm each way is fairly small. The Le variation is not very good, but if you're using this as a sub it doesn't matter. There is various information available for thermal performance but the clearest is the set of two curves depicting input power and power compression. These are shown below: What I read out of this is the following approximate thermal compression: .5 dB @ 30w 1 dB @ 75w 2 dB @ 100w 3 dB @ 125-150w This is OK for a 3" coil sub, but it's not great. Keep in mind this is actual power being dissipated in the driver, which is a very different thing than the size of amp hooked up to it which is what most people mean when they talk about power handling. Hope this has been helpful. If something is confusing I can try explain it better.

IMO, for most applications the most useful and important information is the thermal compression. The variation of parameters with excursion leads to nonlinear distortion, but what a lot of people don't realize is that (1) music spends a hell of a lot more time around 0 than it does at high amplitude and (2), that nonlinear distortion, if it's low order, can be quite high in a sub before it's a problem. On the other hand, even a 1 dB shelf from 80 Hz down is audible and 3 dB will significantly affect the tonal balance of the system. If you put one of these in a ported box and play music through it, the dominant issue by far is the thermal compression. On the other hand, in some situations the variation with excursion may be an issue. For example, if you're using them IB and for very low frequency, short-duration effects in movies, then they're not seeing a whole lot of continuous power but could be seeing some high excursion.

This reminds me of the good ole MTU AES days. Is this your first year at MTU? If you think this is fun wait till winter carnival!

Nice post. Not saying you didn't say this, but I think the bottom line is that a company will sponsor people when there is an incentive for them to do so, and no, feeling good about being nice to other people doesn't count. Advertising that will lead to sales corresponding to more money in profit than has been invested does, and in order for your vehicle to give them that, you have to possess a number of the above characteristics.

Amen to that one. I don't do that much any more, but I still have a pile of equipment that I'm trying to get rid of. As far as the main question goes, I'd have to say taking my DD 9515s out of the car and selling them about 4 years ago is what I regret most.

Wouldn't ported have a flatter FR over the targeted bandwidth of the driver? In 3 cubes, ported tuned to 30-35 Hz is definately more flat than sealed from 30-35 Hz up. If you're going for very low frequency extension (<15 Hz) or are in a car, I can see going sealed or IB, but if the target is music, and you're already burning up 3+ cubes, why not go with higher senstivity, flatter FR, and lower cone excursion? Especially if you are not using EQ!

I hear ya- I run a pair of the DCX's myself. I don't think you can find a comparable unit for twice the price.

I have to wonder, if you're going to invest 4 or more cubes, why not port? It seems to me that sealed is only worth it if you are too small to port it.

So much value it is grotesque?

What about the MJ18? ~.3 cubes?

Damn! If I was going to be around home for a couple more weeks I could take a swing up for it.

Here is my input on the EP2500 I have used about a dozen of them in various home/pro applications under some pretty rough conditions (cold, hot, low voltage, low impedance (by accident)) and I haven't seen (or heard of for that matter) one of them die yet, so they seem to be pretty reliable. On the matter of output power, I have somewhat crudely tested a couple of them and they appear to do as advertised. When measured with the amp bridged and a DMM at the terminals I got about 150 volts RMS driving no load and 100 volts RMS driving a 4 ohm nominal load. I also used an oscilloscope was used to make sure the onset of clipping was indeed when the clip light flashes. When measuring the output on the 4 ohm nominal load, I measured over a wide range to make sure I wasn't square on the impedance bump at resonance, but it is still possible that I may have only been driving a minimum of 5-6 ohms when you throw everything into the mix. If presented with an exact 4 ohm load, the rails might sag a little more...I really don't know. Either way, who cares for playing music (much higher crest factor) anyways- the amp will do what amps rated at 2500w do. There are a couple downfalls with this amp. The first one is its behavior at clipping- when pushed to the limits it does not hit them smoothly. There are a lot of clicks and pops that make driving it slightly into clipping unbearable. If I am going to be doing this I just run limiters upstream anyway, so it is not a big deal for me. It may be for other people. Another downfall is the low cut filter...what a joke. In the one amp I tested, the 30 Hz highpass was such a high Q that I got a 12 dB boost at 40 Hz. I haven't measured it on any other EP2500s, but by ear I can tell there is a sizable boost. I don't know how this varies from unit to unit (maybe I got a bad one) but I don't want it. This isn't a problem if you don't use it anyways (I don't). All around, for the power there is no competition price wise. This amp along with the a500 are about as good of deals as you are going to find out there for 2 channel amps. I wish Behringer would make a 6 or 8 channel amp w/ about 200 per channel. BTW, here is one somewhat extreme application I have used 9 of the ep2500s in: http://audio.students.mtu.edu/pictures/win...rnival2007.html