Any thoughts on side mounted woofer (CSS SDX10) in floorstander designs?

[Efalegalo]

(Just more pondering with parts on hand)

I currently have a pair of:

1) SB Acoustics SB26ADC Tweeters
2) SB Acoustics SB17MFC35-8 Mid-woofers
3) CSS SDX 10 Woofers

I was contemplating using the trio in a slim (relatively speaking) floor-stander with the SDX10 mounted on the side. Kind of like the Aerial Acoustics Model 8

The crossover point would be in the 200Hz-250Hz range <-- mainly based on Erin's (of ErinsAudioCorner) measurements of the ceramic version of the woofer. He found distortion below 200Hz starts to climb rather quickly.

I wanted to know whether fellow forum members have any experience with such designs, or if they see any concerns with proceeding with such a design.

Also - I was thinking of utilizing the CSS in a 45liter sealed enclosure, which would result in a .707 alignment. F3 would be at approximately 43Hz, and F10 at 24Hz. With boundary reinforcement, I'm hoping the F3/F10 will drop by an additional few Hz. Another option would be a long rectangular port running up the back wall of the speaker.

Thoughts? (See picture below - obviously, I'll be including more bracing).

[Steve Manning]

You could also use a Dayton 10" PR on the opposite side of the woofer as well.

You might want to think about having the woofer a little higher off the floor as well. I remember Jon did some experiments with woofer height and found the bass was a lot tighter with the woofer ~12" off the floor.

Should be a cool project though.

[Efalegalo]

Thanks for responding, Steve. Will keep the 12" height recommendation in mind.

In regards to going passive, I'm assuming the ideal location would be on the opposite wall, but:
1) I don't think there will be any space as the motor on the SDX getts pretty close (just off by 1.75") from the opposite wall.
2) Given the excursion capabilities of the SDX, I imagine I would need a pretty large passive radiator.

I'm hoping the sealed output + boundary reinforcement will result in sufficient performance for general 2-channel music listening. However, if you think I'm wasting output/performance, please do let me know :-).

[Bear]

Are you looking to use a passive or active crossover? If you are looking at passive, a "textbook" crossover in that frequency range may require >500uF of capacitance and may have some unplanned interaction with the driver with respect to realized Qts (there was a post a while back on one of the senior members here testing such things). If you are going with an active or active/passive design, then you may have a LOT of freedom to play with realized Q (e.g., Linkwitz Transform), depending upon your target SPL. Piston volume still matters, but there are a lot of tools to use to hit your target F3 and slope.

[tktran]

The side mounted woofer will be okay.

Of the handful of measurements of 10” woofers in enclosures that I’ve seen; they seem omnidirectional until about 100Hz. The +0, -3dB at 200Hz extends to about 90 degrees, give or take. So side firing should be ok, from an SPL point of view.

The late Jeff Bagby used the CSS SDX-10 in his Continuum 3 way. The same 48 L box works in sealed mode; and ported (3” x 10”). He preferred the ported version, but it’s to just as easy build it ported; then plug the port to see what you prefer in your room with your music and tastes. He managed to get by with a 200uF cap on the woofer low pass using a twisted zister; a resistor to “damp the rise in response that occurs from the interaction between the crossover inductor and the capacitive part of the driver’s in-box resonance. And he did try the woofer side mounted and reported no discernible effect.

Have a look for it a meniscus.com for more inspiration.

[Dave Bullet]

On paper I used a Seas L26RFX/P sealed ~ 43Hz F3. Still needed a subwoofer. as you will have power/stroke to spare (the SDX10 won't be the weak link in this trio I'm guessing) - some EQ on the CSS SDX10 could easily offset bass extension. the old Linkwitz transform... or go ported.

It depends partly on corner loading / how close these will be. I play my speakers a way from the side walls (closer to front wall) so lose some boundary reinforcement.

I also can't recall if 200-250Hz is above the rule of thumb for localisation with a side firing woofer. I recall 100Hz or lower ideal - but that may be my poor recollection and "no science", esp. if the SDX10 is no where near beaming.

[tktran]

I've got some 90 degree off axis measurements of 6.5", 10" and 15" woofers somewhere here. I can post that.

The only thing is phase. I mean if a woofer is facing backwards it makes a difference, so at 90 degrees you would think so to...

[Juhazi]

I've got some 90 degree off axis measurements of 6.5", 10" and 15" woofers somewhere here. I can post that.

The only thing is phase. I mean if a woofer is facing backwards it makes a difference, so at 90 degrees you would think so to...

No, phase is not a problem. In a 3-way speaker and woofer crossed below Schröder limit, and in normal domestic conditions, phase shift effect will be minimal. With dsp-xo mid+tweeter delay setting is useful, to compensate this. Floor/side/rear surface reflections will boost bass spl with some ms delay and complicate things anyway. This makes a difference to what can be simulated or measured in a huge anechoic chamber or outdoor/ground plane. This effect is also the main reason why bass sound changes so much with different positioning (distance to wall).

In a rather small room modes will also complicate things even more, but the lowest mode will also give extra boost which will help closed box bass. Closed has tighter bass and I recommend that warmly in this case. Naturally the port can be blocked...

An example of my 10" closed box bass nearfield vs. room response.

[JonMarsh]

Build test enclosures and measure and reconfigure in room- also, look at both 1m and expected far field listening position.

Don’t forget about floor bounce interaction in far field- this guides the positioning of your midwoofer, as well as your woofer. Suggest building a woofer module with driver off set to one edge, so you can flip it around and get measurements near floor boundary and at greater length.

The situation Steve was talking about referred to a dual woofer array, and having them stacked near the floor didn’t produce as good results in the operating range (which was higher than 250Hz) as compared with raising the whole array by about 14” or so.

Use room boundary simulator, such as in VituixCAD, or simply prototype quick and dirty cabinets and experiment in your own room.

Regarding LF alignment for SDX10, sealed is good, it will respond well to slight EQ (Linkwitz transform, passive implementation in preamp to power amp signal path), see what you’re getting in room (selective boundary mode reinforcement for lower frequencies means pulling them out from the wall more). And if you want to “rock and roll”, then model an RSS315PR in Unibox or VituixCAD, but expect to use a larger enclosure. Hemholtz alignments require that for extended LF.

[Efalegalo]

Thanks for the feedback all.

This will be an all passive design (and thus will have its limitations). I like the idea of building ported, but stuff if preferred.

I guess it's all a matter of compromises, as with any speaker build. The more I think about it, the sole benefit of the side-mounted woofer is skinner baffle. I was going to try this because I came across few posts/treads by forum members (here and other sites) indicating a narrower baffle producing a more "3D-sound"

But perhaps I can still go with a front-mounted woofer, but tapper the cabinet as I move towards the tweeter+mid.

Also - since there is a lot more talk about polars today (than there was say 4-5 years ago) - I'm concerned about the balance of the sound in one directionion versus the other (in a side-mounted configuration). I understand at a lower frequency the sound is mostly omnidirectional, but I feel my target crossover of 200-250hz is going to result in a less symetrical sound field.

Nonetheless and i'll experiment on this a little more.

[JonMarsh]

It's very useful to build quick and dirty cabinets for the purpose and evaluate live with test signals- even using a simple three way active crossover to experiment with music may be helpful.

For example, setup your woofer test cabinet with front and side firing configurations, and measure on axis and off axis, out to at least 45 degrees (I usually use 60 degrees). Try to determine both the "native" response of the cabinet with the different firing positions and evaluate the impact of reflections from boundaries.

Frankly, with the SDX10, I think it would be much more interesting to try developing a system patterned after the Dutch & Dutch 8C- use it as a rear firing woofer, the system can be out a ways from the wall and still have boundary loading over a moderate range. Then work on a cardioid Midwoofer setup, to minimize side boundary reflections.

[tktran]

I agree with Jon, before you throw away the idea.

And if you don’t want to build it; there’s a whole heap of pre-fabricated boxes designed for car audio. They are usually very basic and won’t win any beauty contests.
But for about $30-50 you could get a 1 cu ft sealed cabinet built from 3/4” MDF covered in carpet with a precut hole. Surface mounting a woofer makes no difference at low frequencies; and you could do some sweeps from 20Hz to 1Khz and see when they start to make a difference.

At what frequency does it start to diverge at 90 degrees?

In the near field, again at 1m and the listening position eg. 2-4m.

Im actually surprised no one has done this…

[Juhazi]

For final boxes, put woofers on different sides! Then you have the possibility to find out which works better - in or outwards! And still there will be difference between L/R at listening spot! Room modes will rule...

DSP enables easy testing different xo frequencies and topologies as well. Theoretical best in simulation may not work best in your room. I don't have personal experience with sidefiring woofers, this is just tinkering based on what I've read. But I have two pairs of speakers with downfiring 10" (sub)woofers and had several with normal frontside bass 3-ways. Crossed around 3-400Hz LR2 downfire speakers sound fine in stereo, nothing strange. Eg. double bass plucking can be located, because the transient is actually harmonic overtone of the basic note! And sealed has better GD which means better match of transient for the pluck/kick and steady tone, better slam! Kick drum "tone " is around 60Hz, but some recordings carry also subharmonic tones and room modes from the recording room. And some mixers add subsonics to bass, actually that's pretty common for rock/disco/pop/trance.

https://www.waves.com/sub-bass-mixing-tips-from-the-pros

[Efalegalo]

Thanks All - I'll continue investigating this design further.

Just to be clear - I should be taking all these measurements outside (ground plane), correct? Max reflection-free window inside the house is 200hz+

And I'm really supposed to take measurements from 0-degrees to 180degrees (correct?), which in a left-side mounted woofer design would represent:

0 degrees (directly facing tweeter/mid) would represent 90degrees from the perspective of facing the woofer
-90 degrees (directly facing the woofer) would represent 0degrees from the left side mounted woofer perspective
+90 degrees (opposide of woofer) would represent 180 degrees from the left-side mounted woofer perspective.

Is my understanding above correct?

[Efalegalo]

Wanted to circle back on this. I took some outside ground plane measurements of a passive subwoofer (Dayton Titanic MK4) I had on hand. The woofer is 12", but its baffle is as wide as the depth contemplated for the design above.

I then simulated the (late) Mr. Bagby "Universal Crossover" (i.e., 5mh in series, followed by a 200uf + 3ohm in parallel). The simulated response is below.

The dotted red line represents the 0-design axis (i.e., woofer facing right, as the mic sits facing the cabinet)
The dotted purple line represents the -90 degrees (i.e., woofer facing away, with the mic directly facing the side opposite of the side the woofer is mounted)
The dotted turquoise line represents the +90degrees (i.e. woofer now facing the mic).
(please note the rotational axis was the side of the bass cabinet, as it would be in a side-mounted woofer floor-stander).

Based on my review, there is quite a bit of variance between +90 and -90 (about 8db at the crossover frequency, approximately 200hz - 250hz), and even between +90 and 0 (about 5db in the 200Hz - 300hz range)

I also did this comparison between +40 and -40, and the variation again was 5db at between 200hz - 300hz.

Isn't that quite a bit of variance? I mean, I love the idea of a slim tower, but will I be compromising too much when it comes to sound?

[tktran]

What's the distance of your GP measurement?

[Efalegalo]

Approximately 2 meters.


Sent from my iPhone using Tapatalk

[tktran]

I thought that you were measuring too close, resulting in differences in SPL due to relative distance between the mic and the driver. But I think 2 metres is enough- suppose your cabinet dimension is 50cm deep. Turning your cabinet backwards moves the mic from 2m to 2.5m and would account to 2dB loss, not 3dB or more loss that you’ve measured. (8dB at 200Hz)

So there’s 2 outcomes I can infer-
A) bass is not truly omnidirectional - in the bass you lose at least 3dB between 20Hz and 100Hz by rotating a cabinet 90 degrees
B) we’re not measuring what we want to measure.

Do you have a turntable or lazy susan type device to place the subwoofer; to rotate the sub and measure in the nearfield? (eg.20mm from cone or cabinet wall). (Sanity check)- should have same outcome as ground plane.

Now to throw a spanner in the works, what you hear in your listening room is going to swamped by room modes.
So does a woofer blend a midrange well when the crossover is below the Schroder frequency?
Does that have anything to do with the common practice of people prefer to cross the W/M crossover around 300-400Hz instead of ~200Hz?

More questions than answers… ��

[Efalegalo]

Thanks for your thoughts Tktran.

I think part of the difference observed between 20Hz and 100Hz can be partially explained by the increased physical difference between mic and actual woofer locations.

When the woofer is directly facing the mic (i.e.,+90) vs when the woofer is facing away (i.e., -90), that actual woofer (sound source) is now approximately 18"-20" inches away. At least that was the case with the base cabinet I was testing with. Maybe with the floorstander contemplated above, this difference wouldn't be as severe because the actual baffle width would be 8"-9".

[tktran]

Yes I agree with you.

So it looks to me that you could certainly do a side firing woofer; as long as you cross about 100Hz; at 200Hz you will encounter some challenges unless you design for it in your crossover/DSP, or accept the 2-3dB loss and adjust for it with in-room EQ. That’s what’s all my professional musician friends do- on location EQ for live sound.

The only documented designed I’ve seen of a side firing 3 way is from a fellow by the name of Dan Wesnor; he designed a 3 way with side-firing 10” woofer. I can’t recall the name of the design; but it used the original Peerless 830452 XLS woofer paired with a standard (Prestige range) SEAS mid woofer and tweeter. IIRC he crossed over around 120Hz or so.

I’d like to read more about the history (and science) of how 120Hz became the de-facto standard for the .1 or the LFE crossover for home theatre. It might have to do with all this…

[5th element]

A side mounted woofer, I would have thought, is going to have the off axis response of the side mounted woofer combined with the diffraction of the cabinet it's in when at angles greater than 90 degrees and then just the woofers off axis response when at angles less than 90 degrees. So the deeper the cabinet the worse the greater than 90 degrees will be.

All speakers are optimised around one listening axis, usually with the speakers at a specific angle vs the listener, ie toed in or not etc. I suppose the side mounted woofer is no different it's just an additional thing to optimise around with you then accepting the additional driver integration deviations when you sit away from the optimum axis. Crossing very low, just as with tweeter to woofer crossovers, mitigates these issues. So cross higher, if needed, but accept the penalties.

Interestingly if you have two side firing woofers, one on each side of the cabinet, that should also mitigate the problem as you'll get identical responses whichever way you go off axis, to the left or the right. It's not particularly helpful in this case but just an observation. I can see the main issue with this being side mounting two opposed woofers, magnet to magnet, would require a cabinet of decent width, which kind of defeats the point. Shallow mount subs? Or place one woofer above the other.

[tktran]

I thought I'd followup this post with some more information about side firing woofer tower speakers.

I've been looking into dispersion and spinorama and CTA2034A etc and noticed that one of the best measuring speakers, thus far, is the Infinity IL60.

https://pierreaubert.github.io/spinorama/Infinity%20IL60/Infinity/index_vendor.html

This is a rather old design (1999) but after looking at the service manual, I think it was well designed speaker from a company that had significant engineering talent and resources.

Anyway, the relevance of this topic is that Infinity IL60 is a 4 way tower speaker with a side firing woofer. For the low end they use a 12" crossed at 150Hz LR4 to a 6.5". Then between the 6.5" to 4" and 1" drivers the crossover filters at 500Hz, 2800Hz, all LR4. After looking at dozens if not hundreds of measurements over the years of individual drivers, these crossover frequencies are very reasonable, allow for good dispersion and comfortable operating ranges for the drivers.

So I think you can proceed with confidence with a side firing woofer speaker.
In fact, I feel tempted to convert my 2 ways into a 3 way, or my 3 ways into 4 ways with side firing woofers.

[tktran]

OK I have some empirical evidence that might serve useful.

Erin Hardison @ Erins Audio Corner recently popped on my speakers on his Klippel Near Field Scanner.
In short, this robotic machine gives anechoic level measurements with a frequency resolution of under 2Hz from 20Hz to past 20KHz, and take measurements in a full 360 degree arc, and uses sophisticated maths to give the readings at 2 metres.

Here is a 6.5" mid-woofer on a typical 2 way cabinet box, 21cm W x 40cm H x 35cm D cabinet.

Here's the polar measurement:






Zoomed in:



Now changing the axis to focus on what happens if the woofer was side mounted.
That means + 90 degrees would be the front, -90 degrees would be the rear.
Viewing up to 300Hz.



Not viewable to you, but using my mouse- here is the 90 degree
-1dB point: 147Hz
-2dB point: 172Hz
-3dB point: 290Hz

So my 1 line conclusion now:
Completely omnidirectional up to 120Hz (confirms THX / Dolby standard for LFE channels)
OK to crossover a side mounted woofer around 2-300Hz. Closer to 150Hz is marginally better than 300Hz, but 300Hz is totally acceptable. Consistent with real world findings of designers who made speaker with woofer around 200Hz who, when the woofers were turned to the side, felt there wasn't any appreciable difference.

[JonMarsh]

Very good points. Purely as an observation of one company's efforts, PS Audio was doing development for a long time with side mounted woofers, near the floor, but their best efforts required a 6-1/2" "mid bass filler" driver above the floor located woofers before crossing over at 400Hz to a large, long, dynamic planar midrange.

In the 70's, working on minimum phase designs, imagine a relatively wide Ardent style cabinet with the bottom in a shallow vertical V and a 10" woofer on each side, with an Addax Pro 6-1/2" woofer well up near ear level, with a dome mid and MB Quart tweeter above that. It could reproduce good 500Hz and 1kHz pulse waveforms. Very thick front baffles, though of MDF. That's where the baffle stuff started for me, and I gave those speakers to Charlie Hansen because my new wife didn't like the size and look of them- granted, just flat black. A developmental setup. He definitely picked up on the benefits of thick dead baffles.

So, IMO, floor woofers is OK, if you plan on a upper bass/lower midrange driver and midrange and tweeter on top of that. Just my $0.02.