Woofer: side firing pair vs front firing?

[tcpip]

I'm new to speaker design, and I'm learning how to use Speaker Workshop and a mic to design good optimised xo's, etc. I have a question (may be a very basic one).

I was thinking of what it would be like to build a 3-way system (a mental exercise, not an actual project I've started on), with a separate woofer enclosure. I was thinking of using two 10" woofers per box. I'd like to xo them using LR2 or LR4 to the mids at a not-very-low frequency... maybe 200Hz.

Now, my question. I'd like to keep these drivers on the sides of the enclosure, in a push-push config like Lynn Olson describes in his subwoofers for the ME2. Is this a good idea? Twin-opposed drivers cancel out some harmonic distortion --- that's why I want to use them that way --- but is it okay to have sound coming at the listener 90-deg off-axis at full volume till 200Hz and then at reducing volume till perhaps 500-800Hz?

If I keep the woofers front-firing, I lose the HD-cancellation. If I xo to the mids at 100Hz or below, everyone seems to agree that side-firing is as good as front-firing. But if I xo at a higher frequency, then what problems do I face if I have side-firing drivers?

Please note that my question has nothing to do with the ear's ability to localise the position of a sound source --- I'm not talking about having just one woofer enclosure for both channels. I'm actually talking about proper 3-way speakers where I want to align all drivers carefully and have an xo which takes delays and phase errors into account.

Thanks for the help.

 

[Scottmoose]

I'm new to speaker design, and I'm learning how to use Speaker Workshop and a mic to design good optimised xo's, etc. I have a question (may be a very basic one).

I was thinking of what it would be like to build a 3-way system (a mental exercise, not an actual project I've started on), with a separate woofer enclosure. I was thinking of using two 10" woofers per box. I'd like to xo them using LR2 or LR4 to the mids at a not-very-low frequency... maybe 200Hz.

Now, my question. I'd like to keep these drivers on the sides of the enclosure, in a push-push config like Lynn Olson describes in his subwoofers for the ME2. Is this a good idea? Twin-opposed drivers cancel out some harmonic distortion --- that's why I want to use them that way --- but is it okay to have sound coming at the listener 90-deg off-axis at full volume till 200Hz and then at reducing volume till perhaps 500-800Hz?

If I keep the woofers front-firing, I lose the HD-cancellation. If I xo to the mids at 100Hz or below, everyone seems to agree that side-firing is as good as front-firing. But if I xo at a higher frequency, then what problems do I face if I have side-firing drivers?

Please note that my question has nothing to do with the ear's ability to localise the position of a sound source --- I'm not talking about having just one woofer enclosure for both channels. I'm actually talking about proper 3-way speakers where I want to align all drivers carefully and have an xo which takes delays and phase errors into account.

Thanks for the help.

Actually it does have a lot to do with the ear's ability to localise the position of a sound source. Below 80Hz, sound is not directional -above that threshold it is, so if you cross-over your bass drivers to the mids at 200Hz you're going to run into some problems regarding image location between these two frequencies. As you're running two in push-push (good configuration) it will probably be less severe than attempting to use a single side-firing woofer, but it's not ideal, unless you're very firmly in the far-field -further than you'd really want to be when it comes to listening I suspect. That's where JM Labs of all people tripped over themselves with their new[ish] Utopia at around £6000 -they crossed the woofers over too high (same layout you're thinking of) at around 215Hz if memory serves, with poor results except in very large rooms -bass is huge, but it sounds diconnected from the rest of the frequency range, which is not what you'd expect from a pair of speakers at that huge price. What their designer (and he normally knows what he's doing) was thinking of even at the design stage I can't imagine.
That these things were released for sale in that state beggers belief.

The major advantage of push-push is not so much a cancellation of harmonic distortion as that it tends to minimise cabinet streeses, and hense, vibrations. As you state, you would loose this if you mounted both pairs on the front baffle, but, depending on how close to the wall you're intending to have the cabinets, you could always mount one on the front and one on the rear in bipolar push-push config. That would give you almost all of the benefits, and at a stroke minimise most of the drawbacks. The speakers would need to be about 20" or more away from the rear wall. A bit more would be better, but that's something you could experiment with, and such designs often don't need to be as far out into the room as people often think.

Hope some of this helps
Best
Scott

 

[Rademakers]

but is it okay to have sound coming at the listener 90-deg off-axis at full volume till 200Hz and then at reducing volume till perhaps 500-800Hz?

That would be quite high. The higher you cross, the shorter the difference can be between sub and tops, with the sound beïng experienced as coming from the tops.
Crossing as low as possible, is the rule of thumb (1/4 wavelenght or lower).

I use a side firing sub crossed over at 80 Hz with about 60 -70 cm pathlenght between the sub and the first top. All the sound appears to come from the tops unless of course you put the sub relatively too hard or if I raise the crossover beyond about 120 Hz.
I use the sub in reversed phase for better sound.

Mvg Johan

 

[Bill F.]

If your drivers can maintain full-space loading through their passband, and a bit beyond, side-firing shouldn't be a problem. I'd say a side panel 16" wide or less should perform OK. For better time alignment, just tip the tower backward.

 

[tcpip]

Re: Re: Woofer: side firing pair vs front firing?

... As you're running two in push-push (good configuration) it will probably be less severe than attempting to use a single side-firing woofer, but it's not ideal, unless you're very firmly in the far-field -further than you'd really want to be when it comes to listening I suspect...

Thanks for a detailed answer.

I can understand what you described, but can't understand why. Sorry if I'm being a pain, but can you (or any one of you) explain? Is it something to do with the fact that with side-firing, I'll get to hear the wall-bounced reflections more than the direct radiated sound, thus screwing up the localisation?

... you could always mount one on the front and one on the rear in bipolar push-push config. That would give you almost all of the benefits, and at a stroke minimise most of the drawbacks. The speakers would need to be about 20" or more away from the rear wall.

Yes, I could try this, but I really wanted to avoid the need for room at the back. A group of friends and I work on these speaker projects, and some of us just don't have any place for rear space in our living rooms. That's why we're going very slow on OB, for instance. Also, going front-firing will force us to make the cabinet at least 13-15" wide, which has aesthetic implications...

But I'll keep the rear-firing option in mind as a last resort.

Thanks, Scott.

 

[tcpip]

That would be quite high. The higher you cross, the shorter the difference can be between sub and tops, with the sound beïng experienced as coming from the tops.
Crossing as low as possible, is the rule of thumb (1/4 wavelenght or lower).

I can cross over at 200Hz, do side-firing woofers and still have the mid less than quarter-WL from the woofers (just about). You see, I am not contemplating a sat-sub configuration. I am thinking of a woofer box with big flat side panels where I can mount the woofers on the top front corners on either side. And I'll keep the mids+tweeters in a separate box which I'll place on top of the woofer box, just above the woofers. That way, though the mids and woofers will be pointing 90-deg apart in direction, they'll be pretty close to each other. But thanks for the warning about quarter-WL, I'm trying to keep the distance in control.

 

[tcpip]

If your drivers can maintain full-space loading through their passband, and a bit beyond, side-firing shouldn't be a problem. I'd say a side panel 16" wide or less should perform OK. For better time alignment, just tip the tower backward.

I am not sure I understood this. What is full-space? Is this something related to the BSC, where the radiation pattern transitions from 4-pi to 2-pi? Can you please explain in a bit more detail how this 16" comes into the picture? And why should tipping the tower backward help? If you're talking about vertically aligning the mid's acoustic centre with the woofers' centres, then I was thinking of achieving this by a slanting front "baffle", ie. a box shape which would be wider at the bottom than at the top, thus allowing my woofers to be vertically directly below the mids.

Am I understanding you right? If yes, then that 16" wide side panel may be a dampener, because I was hoping to have large side panels to be able to build a slim box big enough to get the space needed for the box tuning. Is it okay if the distance from woofer centres to front baffle is small, but distance from centres to rear baffle is large? That way, will the woofers do 4-pi towards the front and 2-pi towards the rear, and thus make the 2-pi behaviour inaudible for the listener who is in front?

Am I getting what you're trying to explain at all? Really sorry if I'm running off in a tangent....

 

[Scottmoose]

Re: Re: Re: Woofer: side firing pair vs front firing?

I can understand what you described, but can't understand why. Sorry if I'm being a pain, but can you (or any one of you) explain? Is it something to do with the fact that with side-firing, I'll get to hear the wall-bounced reflections more than the direct radiated sound, thus screwing up the localisation?


...I'll keep the rear-firing option in mind as a last resort.

Thanks, Scott.

You're on the right lines. Basically, it's all down to wavelengths. Below approximately 80Hz, the ear cannot discern the location of a sound. This actually varies from person to person, a bit like some people can hear fractionally higher than others, but not by much -only by a couple of Hz at the most.

Now, above that average figure of 80Hz, sound becomes directional. What does that mean in practical terms? Well, take a 3 way speaker as an example. Let's assume that in this 3 way the woofer handles frequencies up to 300Hz, when it hands over to the midrange, which goes up to 4Khz, and passes on to a tweeter, which covers the rest of the audio band. With all drivers mounted on the front panel, you will get a seamless integration of sound from all 3 drive-units, bass, mid and tweeter, particularly if you mount them all quite close together, so that everything seems to emanate from the same place: what we're approaching here is one of the holy grails of Hifi: Point Source. That's why the Tannoy Dual Concentric and KEF UniQ drivers, which have a tweeter that radiates from the centre of a mid/bass cone are so highly regarded.

Now; keeping everything else the same, suppose that we decide to move the woofer away from the other drivers -away from the front panel, and mount it on the side of the enclosure for example. Now, we know that below 80Hz, sound is not directional, so from however low your speaker can go up to that frequency, it doesn't make any difference where the driver is positioned. All well and good so far, but now we come to the sticking point. Above 80Hz, sound is directional. And those sideways firing woofers haven't yet handed over to the forward-pointing midrange driver. This is where the problem lies -from 80Hz, until the bass-drivers cross over to the midrange, which in this hypothetical speaker is at 300Hz, sound is being fired sideways, while everything above that point is firing forward. So for a frequency band some 220Hz wide, sound localisation is not going to happen. You will hear bass coming from one point, and the mids and highs from another; a disconnected rather than unified sound in other words.

If you sit in the far-field (i.e. a good way from the speakers -over, say 20 feet from them), you would get a better integration of sound, because, as you hit on, you will be hearing a far greater proportion of reflected to direct sound than you would if you sat in the nearfield, where there is a far greater ratio of of directly radiated sound from the drivers to reflected sound from the room. Oh -time aligning drivers by the way is where you ensure that all the different drivers produce a sound that reaches the listener at the same time -it's usually done by angling the front baffle slightly backward.

There is nothing wrong with your basic design concept -in fact, I rather like it, but I would cross the woofers over to the midrange much, much sooner if you possibly can. Preferably around 80Hz or a little lower if possible. That way, you'll get a much better integration of sound, and ultimately far more enjoyment.

Best
Scott

 

[Bill F.]

I am not sure I understood this. What is full-space? Is this something related to the BSC, where the radiation pattern transitions from 4-pi to 2-pi?

Correct. Your side-firing woofers will have a baffle step that will correlate with the width of the side panel. At 200 Hz, a 16"-wide surface will not yet load the wave, so radiation will be 4pi--the same all around the speaker, so you should be able to hear it in front just fine.

And why should tipping the tower backward help? If you're talking about vertically aligning the mid's acoustic centre with the woofers' centres, then I was thinking of achieving this by a slanting front baffle.

Fine. That accomplishes the same thing.

 

[tcpip]

There is nothing wrong with your basic design concept -in fact, I rather like it, but I would cross the woofers over to the midrange much, much sooner if you possibly can. Preferably around 80Hz or a little lower if possible. That way, you'll get a much better integration of sound, and ultimately far more enjoyment.

Thanks for a very well-written and patient explanation.

I think I understand. Basically, this means that if I want the woofers side-firing the way I'd described, I'm limited to midrange drivers which can reach down to 100Hz or below at decent SPLs. This means Vifa P13 and their ilk, the Jordan JX92S, GR-Research M130, the CSS WR driver, some high-Xmax TangBand 4" drivers, and all the midbass drivers around. Am not sure the Fostex full-ranges will do, because of their very small Xmax. (I'd really wanted to try some Fostex's.) Don't know whether the HiVi B3S will work.

And you're right, I can't afford to sit 20 feet away from the woofers. I don't live in palaces.

One question about your rear-and-front firing woofers though. Suppose I take up your suggestion and do twin-opposed woofers firing front and rear, won't that cause problems due to the rear radiation in the 100-300Hz range adding to room echo and generally muddying up the soundstage? Upto 100Hz (or 80Hz, whatever), I can undertand that rear-firing is as good as any-other-firing. But what about above 100Hz?

So does this mean that if I want to allow my woofers to breathe till 300Hz, there is no good alternative, really, other than letting all drivers fire front?

Thanks once again.

 

[tcpip]

Correct. Your side-firing woofers will have a baffle step that will correlate with the width of the side panel. At 200 Hz, a 16"-wide surface will not yet load the wave, so radiation will be 4pi--the same all around the speaker, so you should be able to hear it in front just fine.

What about the direct versus reflected sound issue, which Scott has been trying to highlight? Will it muddy the sound (specially the imaging) at 200Hz?

 

[Scottmoose]

If the baffle's wide enough, you might get away with it, so long as you're not sitting too close. That's where bipolars have an advantage over mounting on a single baffle. But I still wouldn't cross over that high though. 1) You need a very deep cabinet, and 2) If you can avoid a problem, or the worst of it by circumventing it in the first place, then that is the route I'd go down! With the reflected sound, it going to rather depend on how close to the side walls your speakers are, how close to the rear walls they are, how far from the speakers you're sitting, and whether you're room is 'alive' or damped. It also depends on what you want from imaging -pin-point accuracy, or the more diffuse sound offered by wide-baffle designs like the Stradivari, and omni type speakers. Your call.

 

[Bill F.]

What about the direct versus reflected sound issue, which Scott has been trying to highlight? Will it muddy the sound (specially the imaging) at 200Hz?

By definition, 4pi radiation is the same all around the speaker, so sidewall reflections will be the same, whether the speaker is side, front, or rear-firing. That said, there will be some out-of-band leakage, depending on the steepness of your XO, which may begin to be loaded by the side baffle.

 

[tcpip]

By definition, 4pi radiation is the same all around the speaker, so sidewall reflections will be the same, whether the speaker is side, front, or rear-firing. That said, there will be some out-of-band leakage, depending on the steepness of your XO, which may begin to be loaded by the side baffle.

What you're saying makes sense to my very limited understanding. And yes, I too was wondering about the traces of output till 500Hz or more, specially if I use a 2nd order xo.

All things considered, do you feel it's best to try this side-firing twin-opposed setup only if I xo at 100 or lower? After getting Scott's and your feedbacks, it's almost beginning to appear to be too much trouble to do otherwise... Sad part is, if my mids have to reach to 100 and lower without strain, I'll lose the option of using all those pure-mid drivers or low-Xmax FR drivers (Fostex).