was: slim tower - now: RS150 in .25 cuft box

[jeff_free69]

the 46" samsung A650 will be here this week, so its time to get serious about slimming down.

- Initially I will keep my 2" dome + ribbon tweeter and replace only the woofer (8" in a 12 x 12 x 24 cab). crossover is about 800 - 1K

-This will have to be a "wife-friendly" project with some unfortunate compromises that have to be made: The max width and depth is 6-8" .
Height is up to 36"

There is already a 12" subwoofer, so deep bass extension is not an issue.

Whats a good 5-6" woofer for this?

-------------------------------------------------------------
EDIT - PLEASE SEE MY CHANGE OF PLANS BELOW
------------------------------------------------------------

[Operandi]

So your going to be rebuilding a current set of speakers?

The Seas L16 would be a good option I would think.

[Dennis H]

Jed's Lineup thread should give you some ideas.

https://www.htguide.com/forum/showthread.php?t=27584

[jeff_free69]

change of plans...

Ok the Samsung L46A650 is up and running in full 1080P glory. Now I can see how it needs to integrate into its spot. unfortunately the tower speaker plan must be put on the back burner due to aesthetic and structural limitations.

The new plan will be to still use my existing dome mid + ribbon tweeter (one thing I didn't explain is that these are together in their own mini stand/enclosure, and sit on top of the woofer box). But to expedite this, the woofer box will be: Dayton TW-0.25BK .25 ft³ 2-Way Cabinet Pair Gloss Black.

The 12" subwoofer gets to stay, so I just need to find a good 5"-6" woofer to put in the .25 cuft box..

Once this is in place I can start evalutaing (dreaming) about what else needs to be done down the line.

[augerpro]

I don't know of any 5" drivers that will work with a .25cf box. The Peerless 832873 will work with the .38cf box, both sealed and ported. The Dayton RS150 and TangBand W5-704 will work in a .5cf box, sealed and ported. The Hivi D5.8 looks like it will fit in a small box, but I haven't modeled it to confirm that.

How do plan on crossing the mid/tweet to the woofer?

[Deward Hastings]

The RS150 works in .25 sealed with a conveniently-close to 80 Hz. rolloff (sealed) to work with the receiver's crossover. Max SPL comes in at about 102 dB, probably sufficient at optimum viewing distance for a 46". And it reaches up high enough to cross to any dome mid.

[augerpro]

Deward you are right-been a long time since I modeled these. In a .25cf box the RS150-8 will have a Qtb of .75 and F3 of 80Hz, which as Deward mentions would work well with usual bass management settings in an Ht reciever.

I also modeled the W5-704 and got Qtb of .73 and F3 of 94Hz. Not as much xmax but it would be easy to cross to your mid/tweet. 2.5dB more sensitive too. Although at 1khz the RS150 is probably far enough from it's breakup that it would work too.

I think the Hivi D5.8 combines the best aspects of the other two, but you pay for it. Small box, good motor/suspension, decent xmax, and it should cross easy at just about frequency you would use.

[Deward Hastings]

> at 1khz the RS150 is probably far enough from it's breakup that it would work

Yes. The breakup is 40 dB down up to 1600-1700 Hz. LR4 even without any additional notch filter. With a dome mid, which would probably cross well *below* 1kHz, it's not even a consideration. Plus, the dome mid will sound much better with a LR4 high pass. The RS150 is a great driver and easy to work with *if you give it a fourth order low pass*.

The W5-704 gives slightly less maximum SPL (which probably doesn't matter) and significantly higher distortion below 1kHz (which may). I haven't seen distortion curves for the D5.8. They'd have to be real low to justify the price . . .

[jeff_free69]

Thanks - The XO design was done by a knowledgeable friend a few years ago , but to the best of my recollection the Dome mid is second order and kicks in at around 800 hz. I just opened up the box and unfortunately the values of several components are unreadable (unless I start pulling things apart).

So i would just need a woofer to cover up to 1 K at the most. On the low end , 80 hz would be fine , so RS150 4th order in .25 sealed to 80 Hz rolloff sounds viable.. (and of course sealed makes it that much simpler)

Wouldn't mind paying extra for Hivi D5.8 . PE customer reviews has a couple of positive reviews for small cab applications, although the sensitivity seems low.

[Deward Hastings]

At 800 Hz the dome is nearing its own rolloff, so second order electric probably gives close to a fourth order acoustic slope. Depends on the dome, of course, but that's how a RS52 behaves. I've used the RS150/RS52 combination . . . it's easy to work with. I'd guess that other 2" domes wouldn't be all that different.

You'll want to get the woofer and the dome as close as possible, so put the woofer at the edge of the box (long ways) if you're going to use it vertical. There's no choice but to center it on the short axis. Since they're in separate enclosures you can adjust phase (offset) physically, but you'll still want to look at the crossover regarding phase tracking (similar slope to the phase curves) either side of the crossover frequency.

[Amphiprion]

Seas L15RLY/P works perfect in the PE 0.25CF boxes. I have the impedance charts to prove it

[jeff_free69]

Seas L15RLY/P works perfect in the PE 0.25CF boxes. I have the impedance charts to prove it

Was that ported ? (if so, what was the length /diam?).

Since its got that aluminum peak thing going on I assume 4th order is also called for, yes? I see Zaph used this in one of his designs with a relatively simple XO , and included a notch filter http://www.zaphaudio.com/audio-speaker12.html

[rc white]

You can run into excursion limiting problems using small woofers down to 80Hz.
I your amp has a small speaker setting at 80Hz., you can use this filters as a part of a filter assisted reflex alignment.
Details of these can be found in this article.

http://sound.westhost.com/qb5align.htm

[jeff_free69]

crossover

It seems from your informed responses in this and another thread, that the RS150 in .25 sealed box is a very viable solution and the way to go.

So then if a use the the Madisound crossover calculator (http://www.mhsoft.nl/CrossoverNetwor...udspeakers.asp)
for -6 DB @1k, 4th order, it presents values for several options:
"Critical" , "Bessel", "Butterworth", "Chebyscheff"

I assume butterworth is the way to go. Is that correct?

BTW these will be up against the wall...

Just for fun in the meantime , I've temporarily moved my Zaph Sr-71s into this spot, too bad that cabinet is just too deep for this application ! (the cab i built is actually a little deeper than the prebuilt)

[rc white]

Using the rs150 down to 80Hz. in a sealed box runs into excursion limiting before the output reaches the standard 110db. peak out put for hi fi.
Using the reference of 112db. for one acoustic Watt this is around 103db.., this is optimistic in a practical sense and a design figure of 108db. is better ,giving a peak output of 99db.
If your amplifier has a 80Hz. Butterworth filter at the output then a 7.1 liter box tuned to 60Hz.will run out of the rated 65Watt voice coil power handling and reach an excursion of just over 3mm. for a peak output of 109db.

[Deward Hastings]

Where does this "standard 110db. peak out put for hi fi" figure come from? By that "standard" not one of the speakers I've owned over the years (including two generations of Magnepans and my present ORION) qualify as "hi fi". Not to mention that few of the thousand plus recordings I own would be listenable for long if 0dB fs produced a 110dB output.

[jeff_free69]

If your amp has a small speaker setting at 80Hz., you can use this fi...

Yeah its got one of those,
but I didn't think I'd ever have to use it
Oh well, life goes on (didn't think I'd ever give up my 8 cuft 2x12 Dynaudio subs either)

As far as volume goes, 110 db sounds pretty unhealthy to me. I've been measuring levels in my studio where I try to keep it down to 85 DB for extended periods. 90 is more impressive of course. 95 starts to kick butt. But anything more than that and you'll be looking around for the whistling tea kettle someday soon...

[rc white]

The ability to reproduce peaks of 110db. with a stereo pair of speakers in an average room is a standard that has been long accepted by the hi fi industry.
To put this into perspective this represents an average sound pressure level of 80db. for 30db. dynamic range orchestral music, or 90db. for what is usually described as the most often occurring dynamic range of 20db.
Since in a concert hall orchestras have been measured to produce 120db.peaks in crescendos this is in fact quite modest.

[Deward Hastings]

What an orchestra can do in a concert hall and what's actually present on recordings are two quite different things. If uncompressed recordings were made that preserved the loudest (potential) peaks a good part of the typical performance would be lost in the noise. Since it's the bulk of the performance that people want to listen to those occasional transient peaks are always compressed and limited long before they get to your loudspeaker. There's no need to be able to reproduce something that isn't there in generally available recordings.

I'd love to see some references to the "long accepted" 110dB standard, as it's not something I recall having heard of before. I didn't find anything in my copy of "Handbook for Sound Engineers", though it could be in there somewhere, or in the "Master Handbook of Acoustics". The closest I found was in "Acoustics and Psychacoustics", which in a table on page 82 lists 110dB as "Peak levels on a night club dance floor", and "This Is Your Brain On Music", which lists 100-105dB as "Classical music or opera concert during loud passages" and 110db as "a jackhammer three feet away" (I presume they mean a pneumatic jackhammer . . . my electric one is, though unpleasant, not that loud).

All of which suggests more than a little divergence of opinion regarding what is required for "hi-fi", even after room size and musical taste is taken into account. My personal tastes do not require 110 dB in my 16x24 ft. living/listening room. Perhaps a discussion of what we really expect (or want) in our living rooms is in order . . .

[rc white]

A point I would make is that the "standard" 112db. at one meter for one acoustic Watt is not really relevant for most people living in western urban environments these days.
Recent studies have shown that the hearing threshold has moved upward by 4-5db. in recent years meaning that as a design figure my 108db. might be a bit optimistic.
Another thing that I find perplexing in speaker forums is that people are very concerned about whether a crossover point should be several hundred Hertz this way or that, in order to avoid small signal driver distortions, and yet are perfectly happy to use a system that is driven into quite severe excursion non linearity that can easily increase non linear distortions by an order of magnitude.
I am at a loss to understand your reluctance to use a high pass filter when tuning a quarter cubic foot box to 60Hz. and using the filter will reduce average cone excursions bellow 150Hz. by around 20 times and give distortion figures that are accurately reflected by the small signal ones.

[Deward Hastings]

> I am at a loss to understand your reluctance to use a high pass filter

Huh? The RS150 in a quarter cubic foot box has a natural second order rolloff down 3dB at about 80 Hz. Combined with the 2nd order filter in the typical HT receiver that gives a 4th order LR highpass that mates with the 4th order lowpass to a subwoofer that is also built into the typical HT receiver. Configured thus a single RS150 is *linear* excursion limited to about 102 dB at the low end of its range, although it will play louder (with increasing distortion) if overdriven. That's probably quite sufficient.

Tuning the box to 60 Hz. with a resonator requires a filter to remove the booom (although it will not remove the other problems the resonance brings) and would then require an additional LR4 highpass (which would probably have to be made up from a bi-quad and a 2B) to mate with the subwoofer. That's a lot of unnecessary filter complexity. It's also not clear that the required resonator would even fit in a PE .25 cu.ft. box.

My reluctance is to suggest something that does not address the original poster's needs and constraints . . .

[rc white]

I am not sure why you would need the plethora of filters you mentioned.
A 80Hz. Butterworth high pass in conjunction with the tuning I mentioned constitutes a compliance scaled QB 5 II alignment and it needs no other filter.
It is true that the total high pass is then sixth order but any effect from this is completely swamped by room effects and the proximity of the sub woofer, and as I mentioned it has at least 10db. more headroom than a sealed box and will sound a lot cleaner on signal peaks.
Being to a large degree of accuracy a minimum phase system, (the driver is in its piston range), any concatenation of filters is equivalent whether they are electrical electronic or acoustic. This being the case the extra second order roll off will introduce just over 1.5milliseconds more settling time than the total fourth order high pass will have, but more importantly the system will have considerably less distortion overall.
I would suggest that this does meet the original posters requirement, i.e. to use a smaller more compact set up in the interest of domestic bliss, this is the reason why I embarked upon the project of finding a way of making satellites smaller withought compromising spl. in the first place.

[Deward Hastings]

Say again please what the port diameter and length must be to get this tuning with a RS150 in a .25 cubic foot box, and how you get the high pass slope and phase to match the LR4 low pass of the subwoofer?

[rc white]

The port is 47mm. in diameter and 169mm. long, it has a peak air velocity of 13.5m/s at a frequency of around 68Hz.

As you might expect the frequency response is not quite flat through the crossover region, Bode plots of the theoretical response are here..



Red is of course the sealed box.

The sub woofer is modeled as having a Butterworth roll off three octaves above the crossover frequency, the sealed satellite an exact fourth order L-R at the crossover frequency, and the vented system with high pass filter a q=.7 at the crossover frequency.

This simulation also assumes that the acoustic centers are aligned and that the distance between the drivers is small.
In practice the frequency response anomalies are completely swamped by positioning and room modes.
The the comparative impulse response is here, red again is the sealed box.

[Deward Hastings]

To say “not quite flat through the crossover region” is a bit of an understatement, eh?

Perhaps I’m reading the table “2.4 Group III Alignments” in your paper incorrectly . . . doesn’t it say that for a driver “Qts” of .51 Vas/Vb should be .533? Since the Vas of the RS150 is listed as .5 cu. ft. doesn’t that mean that the box should be almost 1 cu. ft., not .25 cu. ft.?

The RS150 is rated 40 watts continuous (which gives 102 dB over its useful range), 65 watts max. How much additional power is required to give the higher output in theory made possible by your alignment? Just how much greater output is that, in this particular example? Is the second order electrical filter enough to protect the driver from overexcursion at that power *below* the box resonance? Is there an excursion hump *above* the box resonance, as there is with many vented alignments?

Too many questions, perhaps, but the suggestion that nothing more than 8” of 1.5” pipe will increase the acoustic output of a RS150 (in the same .25 cu.ft. box) by a factor of four (or more) with no detrimental effect is a bit . . . extraordinary.

[jeff_free69]

Just to follow up on the "how loud does it have to be?" question -

in my 12 X 20 X 9 living room I just measured music playing impressively loud to be 90-92 DB (C weighted) at ~ 1M in front of one speaker, and about 85 at my listening location (6-8 ft away). Anything speaker putting out 100DB is perfectly adequate here.

As far as 120 db Orchestras - I ask where that was measured - in between the brass section and the tympani?? I've been to my share of concerts at world class halls, and let me tell you how suprisingly low the volume is out in the seats. Of course I've been to more rock concerts and I don't even want to know how loud that was

[rc white]

In my previous post I was using data for the Dayton rs150, the rs150-8 has different specs.



The blue trace in the above is the sealed box with 80Hz. filter, the red a 5litre box tuned to 50Hz. with the same filter, the yellow a 80Hz. filter with a q=.5
All of these are adjusted to get close as possible to the same -6db. point.
The q=.5 filter is practically identical to the sealed box, and the vent plus Butterworth filter deviates from it by a couple of db. max.
The q=.5 filtered box will have practically the same frequency response as the sealed box through the crossover region, the Butterworth plus vent will have less deviation than the previously described box and should be no more than 2db.



The cone excursion plot is for a 200Watt peak.
The sealed box just exceeds the linear cone excursion limit by about half a millimeter, of the other two plots the q=.5 filter is the best but the other one still gives useful excursion reduction, other drivers exist that give better results.
It is amazing what you can do with a tube isn't it?

[Deward Hastings]

So if your HT receiver can deliver 200 Watts (after crossover losses) you can get 6 dB over what 40 Watts deliver in a sealed box, at the expense of reflex ringing, a 3 dB “response deviation”, some phase change that may result in a more than 3 dB error in summed response, and a potentially blown driver. It’s not that far from what one would expect from any reflex alignment, with most of the same tradeoffs.

There may be some situations where it makes sense, as in extending bass response in stand-alone loudspeakers. Most of us, were we not faced with a rigid size constraint, would probably just get the “extra” dB without the negatives by switching to a RS180 (or be content with a clean and safe 102 dB) . . .

Thanks for the revised explanation. I’m still curious as to how this conforms to your published table showing a considerably different Vas/Vb ratio.

Ps. In my previous post I should probably have said 2” pipe . . .

[rc white]

A thing that I always keep in mind Deward is the principal fallacy that audio people fall into, as stated by Floyd Toole, is that, "what offends the eye offends the ear in like measure"
It is true that the reflex alignment has a bit more ringing but it is an amount that double blind testing indicates you can't hear. The same can be said of the frequency response which again put against the background of real room responses means that small movements of subwoofers or satellites have more effect, and as a result you cant hear small anomalies in the crossover at those frequencies.
All testing I have ever seen however finds that the ear is very sensitive to non linear distortions of all sorts and this being the case the method I outlined addresses this by greatly reducing such distortions, overall taking care of what matters to the ear and not what might offend the eye.
rcw

[Dennis H]

All testing I have ever seen however finds that the ear is very sensitive to non linear distortions of all sorts and this being the case the method I outlined addresses this by greatly reducing such distortions, overall taking care of what matters to the ear and not what might offend the eye.

"Testing I have seen" and "what might offend the eye" -- careful you aren't falling into your own trap. Have you done an A/B with a sealed box and one of your ported alignments, both crossed to a sub with a THX-style receiver crossover, and heard an audible difference?

Edit: it would be really easy to test with this proposed alignment -- .25 cu.ft. box, RS150-8 (pretty close to Q=.7 at 80 Hz), and an 80 Hz receiver crossover. Build it ported, play some music and have somebody insert/remove a cork while you listen. My bet is on the better XO blend of the sealed alignment, if you are able to hear any difference at all (which is a big question.)

[Deward Hastings]

I'm not seeing where this alignment reduces driver distortion *when the driver is operated in its linear excursion range* (in this case below 102 dB in the sealed alignment) by any significant amount at all. In other words, below 40 Watts input you don't gain much of anything at all in reduced distortion, so *any* resonance or fr effects (which occur regardless of level) are incurred to no particular benefit at all.

In my own experience I find that I can and do hear 3 dB response deviations in the bass range (try it with a graphic equalizer some time), and the difference between ported (resonant) boxes and my relatively non-resonant dipole woofers is even more striking. I tend to look for "theories" that explain what I know I can hear, rather than look for theories which "explain" why I can't possibly hear what I most certainly do.

If I added up all the things I'm not supposed to be able to hear I'd save a lot of money, and be listening to a $49 boom box . . . the response and distortion curves only *look* bad, don't ya know . . .

[rc white]

I do not as a general rule trust things that are not double blind tested.
All of the evidence available suggests that what people "hear" is in fact as much what they see as what they hear, and of course every audio person thinks that whilst that may be true of everybody else it does not apply to them.
I can report that I modified a system some years ago along these lines and another party who didn't know this and used it commented that it sounded very clean and "open', especially when playing moderate to loud orchestral music, this had been my experience also, but then again that is to be expected.
Another thing that should be noted is that the rs150-8 in the 7.1 liter sealed box will also generate extra distortion.
This is because as Small showed the volume of the box has to be vas/3 as a minimum so that the air spring of the box dominates the restoring force, the rs150-8 used in such a box would have f =96 Hz. and q= .81.
This is not a problem with the reflex box because the high acoustic impedance supplied by the port dominates the restoring force, the nonlinear nature of cone suspensions is one of the great weaknesses of moving coil drivers.
As I have pointed out previously driving the room one quarter the distance across its smaller dimension one quarter of the ceiling height from the floor has the same practical effect as dipole woofers and is cheaper smaller and uses a lot less power.
rcw

[rc white]

I do not as a general rule trust things that are not double blind tested.
All of the evidence available suggests that what people "hear" is in fact as much what they see as what they hear, and of course every audio person thinks that whilst that may be true of everybody else it does not apply to them.
I can report that I modified a system some years ago along these lines and another party who didn't know this and used it commented that it sounded very clean and "open', especially when playing moderate to loud orchestral music, this had been my experience also, but then again that is to be expected.
Another thing that should be noted is that the rs150-8 in the 7.1 liter sealed box will also generate extra distortion.
This is because as Small showed to minimize distortion, the volume of the box has to be vas/3 as a minimum, so that the air spring of the box dominates the restoring force.
The rs150-8 used in such a box would have f =96 Hz. and q= .81, a more suitable driver for a sealed box then has a qt=.35 and an fs=40Hz.
This is not a problem with the reflex box because the high acoustic impedance supplied by the port dominates the restoring force, this being another distortion reduction mechanism, the nonlinear nature of cone suspensions being one of the great weaknesses of moving coil drivers.
As I have pointed out previously driving the room one quarter the distance across its smaller dimension one quarter of the ceiling height from the floor has the same practical effect as dipole woofers and is cheaper smaller and uses a lot less power.
A small frequency response aberration in the 80Hz. crossover region is just as likely to smooth out the overall room response as it is to make it worse, and I will reiterate my point, non linear distortion matters a lot and a small aberration of that sort hardly at all.
rcw

[Deward Hastings]

> Small showed to minimize distortion, the volume of the box has to be vas/3 as a minimum

In theory, theory and practice are the same; in practice . . .

Many poor fools use dynamic drivers (much like the RS150 in design) in dipoles and infinite baffles where there is no “air spring” at all. Complaints of increased distortion are few. Most loudspeaker distortion tests are, in fact, conducted on infinite baffles . . . I’m all eyes (even though you don’t think that eyes matter) for distortion curves showing significant reduction of woofer distortion when said woofer is mounted in a box.

> As I have pointed out previously driving the room one quarter the distance across its smaller dimension one quarter of the ceiling height from the floor has the same practical effect as dipole woofers and is cheaper smaller and uses a lot less power

That’s *your* theory, so it’s understandable that you cling to it, and repeat it.

In my listening room I have had the rather well thought of Thiel 3.6 (reflex with passive radiator) and Linkwitz ORION (dipole) in the same location (about one quarter the distance across the smaller dimension). The Thiel woofer was, as I recall, close to one quarter the ceiling height above the floor (the passive radiator was much closer to the floor). Cost difference was negligible. The overall box dimensions are within inches. The two ORION woofers use two 60 Watt amplifiers (120 Watt total per side). The Thiel sounded constricted with anything less than the 400+ Watt output of Adcom GFA-565 monoblocks. The ORION bass is notably more “clean” and “open” to my ear, and I have it on personal report that others agree.

That’s been my “practice”, so it’s probably understandable that I cling to it.

I agree that non-linear distortion matters . . . you don’t need to keep repeating that. What you do need to explain is what difference that makes when the driver remains in its linear range (for practical purpose less than 40 Watt drive on a RS150, sealed). Such distortion reduction as the reflex alignment might provide only occurs if you beat the driver with a 200 Watt hammer and try to make smoke come out of the box. I don’t make a habit of doing that (although I might have, once or twice, 30 years ago).

This is getting boring . . . no minds are going to change because of it. Have the last word and let's be done with it.

[jeff_free69]

Now, now, you boys go to your room and don't come out till you agree to play nice!

Obviously there's many ways to skin a cat ( do people still do that?), and the discussion has helped me get up to speed with the state of the art.

For my immediate needs I think I'm sticking with the RS150 .25 cuft sealed.
This will give me time to plot how I might carve up the living room to accommodate a more substantial design.

Now, which is a better computer - PC or MAC?
(Don't answer that!!)

[rc white]

I must say Deward it is gratifying to learn that it is possible to have robust discussion with people in this forum withought matters degenerating into a slanging match, for that much thanks.
rcw