Woops, here's the midrange questions!

[jessegabriel]

Dear Sirs, and Jeff Stahl,
I am finally going to build my enclosures this weekend, however, I am getting conflicting information about my midrange enclosure. First, here are the speakers I am using: Woofer= Dayton 295-120 Midrange= Davis 297-560 Tweeter= Dayton 275-130. Previously, Jeff Stahl sent me a passpro6 (?) calculation suggesting a .0898 cu. ft. enclosure (vented) with a 1" diameter and 1.789" length. I used Win Isd and it came up with .057 cu. ft. vented enclosure! Which am I to use? And, how important is this considering it is only a midrange? I personally prefer sealed enclosures than vented, but if I can get better sound form a vented, then that is what I'll go with. Can you give me a suggestion on the alternative sealed enclosure for this Davis speaker? Final question is this; I have a choice of either using this Davis 297-560 for the mid, or an eminence 290-542. Between the two, which will give me more quality of sound? (I'm building two sets of speakers, and want to keep the best set for me!) Thank you techs, you have been alot of help to me so far, Rock On!

Sincerely, Jesse
PS I just have received the LDC7 software for speaker design, but cannot get it to run on my computer,(computer illiterate) Can you suggest where I can get some help to be able to use this program? Thanks

[wg_ski]

> Dear Sirs, and Jeff Stahl,
> I am finally going to build my enclosures
> this weekend, however, I am getting
> conflicting information about my midrange
> enclosure. First, here are the speakers I am
> using: Woofer= Dayton 295-120 Midrange=
> Davis 297-560 Tweeter= Dayton 275-130.
> Previously, Jeff Stahl sent me a passpro6
> (?) calculation suggesting a .0898 cu. ft.
> enclosure (vented) with a 1" diameter
> and 1.789" length. I used Win Isd and
> it came up with .057 cu. ft. vented
> enclosure! Which am I to use? And, how
> important is this considering it is only a
> midrange? I personally prefer sealed
> enclosures than vented, but if I can get
> better sound form a vented, then that is
> what I'll go with. Can you give me a
> suggestion on the alternative sealed
> enclosure for this Davis speaker? Final
> question is this; I have a choice of either
> using this Davis 297-560 for the mid, or an
> eminence 290-542. Between the two, which
> will give me more quality of sound? (I'm
> building two sets of speakers, and want to
> keep the best set for me!) Thank you techs,
> you have been alot of help to me so far,
> Rock On!

Use sealed sub-enclosure for the midrange, and stuff the crap out of it to deaden it. The volume isn't critical as long as your crossover is above 300 Hz.

The Davis would be better sound quality wise, but it may be tricky to get the crossover to work right. The sensitivity on the 4" is a bit low - I would be looking for something in the 89 dB range or higher - like maybe the 5" version. It's *much* easier to pad down if you need to than to rely on band pass gain - which causes the impedance to fall. There are a lot of good 5" to consider. The Eminence is closed-back and you do *not* want that. That pushes fs too high for hi-fi applications and it's wayyy too high on sensitivity. The open back version with a small sealed and *damped* subenclosure would be more appropriate.

[curt_c]

> Use sealed sub-enclosure for the midrange,
> and stuff the crap out of it to deaden it.

Be careful here, as too much stuffing truely does 'deaden' it. -To the point it sounds lifeless. Just enough stuffing to kill the standing waves is the point to shoot for.

Another possibility is to make the mid enclosure a transmission line. Consider something like a 5" tube running the entire depth of the enclosure and open at the back. Lightly stuff with hollowfill to taste. This will not have the standing wave issues as a sealed enclosure can, and can provide a more spacious sound.

C

[Pete Schumacher ®]

> Be careful here, as too much stuffing truely
> does 'deaden' it. -To the point it sounds
> lifeless. Just enough stuffing to kill the
> standing waves is the point to shoot for.

> Another possibility is to make the mid
> enclosure a transmission line. Consider
> something like a 5" tube running the
> entire depth of the enclosure and open at
> the back. Lightly stuff with hollowfill to
> taste. This will not have the standing wave
> issues as a sealed enclosure can, and can
> provide a more spacious sound.

> C

Be careful here too. Just because it's an open line does not mean it won't have an impact on the response of the mid. T-lines DO have standing waves and can impart passband ripple since the acoustic impedance the mid will see varies with frequency.

[damkor]

> Be careful here, as too much stuffing truely
> does 'deaden' it. -To the point it sounds
> lifeless.

The only negative effect that overstuffing could have on a midrange enclosure is if you pack the stuffing so tightly that the volume of the enclosure is actually reduced, rather than virtually increased. And that could increase internal reflections having an effect on the cone from the backside and more likely increase the Q of the system, causing possible peaks around 100 to 200Hz.

[curt_c]

> Be careful here too. Just because it's an
> open line does not mean it won't have an
> impact on the response of the mid. T-lines
> DO have standing waves and can impart
> passband ripple since the acoustic impedance
> the mid will see varies with frequency.

I didn't state that clearly, did I... Ths stanading wave issues will still exist. The potential ripple will be ameliorated with the relatively large cross sectional area of the line compared to the driver Sd, and of course with stuffing. IME, the ripple will be a non-audible issue.

C

[curt_c]

Have you ever built a 3 way? -Just curious...

C

[damkor]

Yep. Best results: Large enclosure, 80% "stuffed". But not so tight that the volume goes down.

I just take issue with your use of the word "deaden". I think you are describing the result of internal reflections affecting the mid cone. The sound stends to close in and become less "open"? Keep the mid enclosure large, even open back if the xover point is high enough and stuff it.

[waynew]

[waynew]

80% seems like an awful lot of fill to me Madrok. That definitetly will "deaden" the sound of the midrange. It will also "damp" cone movement because of the compliance variation. I have experimented with this and find that in a midrange situation the right amount is usually a lot less than you think...e.g. 20 to 40% range

[damkor]

I think you have to keep the sound from reflecting off those hard walls.

[waynew]

> I think you have to keep the sound from
> reflecting off those hard walls.

I agree with that, but with 80% stuffing, it is very likely that it will limit cone movement, causing the sound to sound "deadened". I should ask what you consider 80% stuffed. How loose or how tightly packed the fill is. I would urge you to try an experiment. Set up two speakers of the same volume with the same driver and fill one to your 80% and fill the other with about 30%. This may give you an idea of what Curt means when he says that the sound can be "deadened"

[paul_n_oneal]

I've built 3 3 ways, 1 3.5 ways and many two ways. I never stuff 80%. My current 3 way has about 20-30% stuffing and the walls lined with 1" open cell foam. Jee at 80% your probably stuffing the woofer spider and maybe even touching the woofer cone. Definitely not a good idea to do that. So if you're using 80% stuffing then you built the box to small to begin with! This comes from many years of experience. Anyway, that's my two cents--TAKE IT OR LEAVE IT AS I DON'T REALLY CARE :)

Paul

[wg_ski]

> I agree with that, but with 80% stuffing, it
> is very likely that it will limit cone
> movement, causing the sound to sound
> "deadened".

It won't limit cone movement in the mass-controlled region! The mass of the cone and air load does the bulk of that. It may sound "deader" if internal reflections and box/frame resonaces are removed, but you want that, don't you?

What does happen is you get a lower Qtc, and that may cause low end roll off if you're not crossing over 2 octaves above the in-box fc. Large sub-enclosures or T-lines do the same thing. Deal with that in the crossover - you just need a 2nd order network with Q > .707 . Heavier stuffing also reduces the magnitude of the impedance peak at resonance (it's like lowering Qms) so the crossover is easier to implement.

[waynew]

> It won't limit cone movement in the
> mass-controlled region!

If you remove some of the compliance with stuffing, the mass-controlled movement will be damped due to increased air pressure. I liken this to a sheet of plywood going down the road at 30 mph vs. the same sheet of plywood with 50lbs attached to it. Air pressure is still going to slow it down even though there is added mass element!

The mass of the cone
> and air load does the bulk of that. It may
> sound "deader" if internal
> reflections and box/frame resonaces are
> removed, but you want that, don't you?

Only to a point. I urge you to try the experiment that I have tried and draw your conclusion from that.

> What does happen is you get a lower Qtc, and
> that may cause low end roll off if you're
> not crossing over 2 octaves above the in-box
> fc. Large sub-enclosures or T-lines do the
> same thing. Deal with that in the crossover
> - you just need a 2nd order network with Q
> > .707 . Heavier stuffing also reduces
> the magnitude of the impedance peak at
> resonance (it's like lowering Qms) so the
> crossover is easier to implement.

[wg_ski]

> If you remove some of the compliance with
> stuffing, the mass-controlled movement will
> be damped due to increased air pressure. I
> liken this to a sheet of plywood going down
> the road at 30 mph vs. the same sheet of
> plywood with 50lbs attached to it. Air
> pressure is still going to slow it down even
> though there is added mass element!

The air pressure changes are really really tiny at 500 Hz. It's just not moving very many millimeters before it has to put on the brakes and go back the other way. Even less at 2500.

[waynew]

> The air pressure changes are really really
> tiny at 500 Hz. It's just not moving very
> many millimeters before it has to put on the
> brakes and go back the other way. Even less
> at 2500.

I still urge you to try the experiment. It does make a significant difference.

[Deward Hastings]

It really all comes down to why one stuffs the box in the first place. The enclosure is just a reduced size and portable infinite baffle . . . its function is to ensure that no part of the back wave (energy) from the driver's cone escapes and contaminates (distorts) the radiation off the front of the cone. So one puts enough stuffing in the box to completely absorbe the back radiation . . . enough to insure that there is no energy radiating through the walls of the enclosure, and expecially no energy from the back wave bouncing around and then back out through the cone. In a sealed box that means absorbing pretty much everything . . . and really calls not so much for *more* stuffing but for *more efficient* stuffing, something like the fiberglass or rock wool bats used in bass traps and flat panel sound absorbers rather than the much less effective acoustic foam and polyfill now in common use. Such bats do not need to completely fill the enclosure . . . in fact they work better if there is some air gap around them.

Resonant enclosures, ported or transmission line, pose a special problem . . . one wants to absorb sound energy above the box "tuning", but leave enough at the box or TL resonance for it to "do its thing". As a result the amount of stuffing, its absorption curve and its placement becomes a part of the "tuning" of the enclosure, and too much stuffing can defeat its purpose. If adding stuffing makes a sealed box seem "dull and lifeless" it's probably because the unabsorbed back wave (distortion) was compensating for a frequency response problem (which should be fixed in some other way than adding box coloration to the sound). Ideally the driver/crossover should sound the same in a sealed-and-stuffed box as it would on an infinite baffle (not counting baffle step, which is another issue).

Of course not everyone agrees . . . if they did speakers in boxes would not so commonly sound like . . . speakers in boxes.

[dthomas]

I agree with Warren that cone movement is not the concern hear. But as it as already been stated, absorption of the back wave is the big issue. Insuring that as much as possible is dissipated in the absorption material will result in a more natural sounding mid. Not sure how you calculate 20-80% stuffing but I agree it needs to be fairly substantial. I would venture that an enclosure with only 20% of the area stuffed will suffer fron coloration. But this is too some degree dependant on the size of the enclosure. I also think the enclosure should be oversized and the early roll-off dealt with in the crossover. I typically loosely fill the entire enclosure. By loosely I mean I attempt to seperate and fluff the material whether it is polyfil or fiberglass.

Otherwise you wind up with a boxy sounding midrange that will suffer from coloration due to the backwave escaping through the driver cone.

Dave

[waynew]

Again. I would urge you to try the experiment and draw a concrete conclusion. Every one's taste is different, some may prefer a more lively or dead sounding midrange. However, it would be beneficial to base your assumptions in fact.

[Deward Hastings]

It might be helpful if we had a more descriptive terminology. In my experience properly equalized midrange drivers in infinite baffle or dipole do not sound "dead", and neither do they with a fully absorptive box behind them. I'd call the sound "articulate" or "precise" or "clear". With an underdamped box behind them they sound (to me) muddy, distorted, decidedly unclear . . . I'd call it UNlifelike. Speakers in boxes don't have to sound like that, but in my experience they commonly do . . . so commonly that people are used to the "box" sound, and may even come to prefer it, finding the kind of speaker I like too "cold" or "clinical". Box resonances amount to a kind of artificial reverberation, and sometimes people do like that . . .

[dthomas]

> It might be helpful if we had a more
> descriptive terminology. In my experience
> properly equalized midrange drivers in
> infinite baffle or dipole do not sound
> "dead", and neither do they with a
> fully absorptive box behind them. I'd call
> the sound "articulate" or
> "precise" or "clear".
> With an underdamped box behind them they
> sound (to me) muddy, distorted, decidedly
> unclear . . . I'd call it UNlifelike.
> Speakers in boxes don't have to sound like
> that, but in my experience they commonly do
> . . . so commonly that people are used to
> the "box" sound, and may even come
> to prefer it, finding the kind of speaker I
> like too "cold" or
> "clinical". Box resonances amount
> to a kind of artificial reverberation, and
> sometimes people do like that . . .

Deward... I think you have said it best. Dissipating the backwave is the goal. It eliminates resonances and reflections that hit the rear of the cone and cause that muddy, distorted sound.

Fact is if you model a box with heavy fill vs. light fill you can easily see the effects. And since I have experimented on dozens of three way systems I can tell you if you underdamp the midrange enclosure the results can be just as you described.

I think one also needs to understand that in the midrange there is little need for large excursions.

Dave

[curt_c]

> It might be helpful if we had a more
> descriptive terminology.

Agreed.

> In my experience
> properly equalized midrange drivers in
> infinite baffle or dipole do not sound
> "dead", and neither do they with a
> fully absorptive box behind them. I'd call
> the sound "articulate" or
> "precise" or "clear".
> With an underdamped box behind them they
> sound (to me) muddy, distorted, decidedly
> unclear . . . I'd call it UNlifelike.

Soooo... How can a low Q driver work in a (underdamped) dipole, but not work in an oversized enclosure?

No, I get your point: An sealed enclosure must have some isothermal mechanism in place to convert the energy of the back wave. -What doesn't get converted to heat, must come back out the cone.

> Speakers in boxes don't have to sound like
> that, but in my experience they commonly do
> . . . so commonly that people are used to
> the "box" sound, and may even come
> to prefer it, finding the kind of speaker I
> like too "cold" or
> "clinical". Box resonances amount
> to a kind of artificial reverberation, and
> sometimes people do like that . . .

I agree this is a very subjective topic, both from what amounts to ‘heavy’, light, 80%, stuffing, and from the differing driver response due to changes in the stuffing, enclosure size, etc.

There is a certain ‘boxy’ sound that is evident in many monopole designs. For the purposes of this discussion, the exact reasons why may not be necessary to elucidate, but I think everyone will agree they certainly exist, and that sound is not evident in dipole designs. The dipole design of course doesn’t magically eliminate the back wave, it just wraps around the baffle, or bounces off the wall instead. The time delay and losses due to distance may be significant here.

If we assume (and this may be an incorrect assumption to some) that the dipole sound is more accurate, then I suppose we could call this ‘sound’ some kind of standard, or reference.

IME, a correctly constucted and sized (larger) midrange cab, lightly stuffed, sounds to me more like the good dipoles I’ve heard. That is, excellent detail without coloration. Stuff that same cab, and to my ears, the sound becomes dull and the detail goes away. One could say the added detail is just distortion from the back wave, but then why does the lightly stuffed cab sound more to me like the detail of a dipole? This is significantly different than the sound of standing waves, which occur only at a few frequencies, and are very obvious as response aberrations, -but I would not describe those aberrations as sounding anything like added detail.

Why does stuffing affect the sound of the driver when its in its mass controlled region?
I don’t have a good answer for that. Perhaps as the stuffing density is increased, it becomes a progressively more non-compliant mass, and acts more like an adiabatic boundary, rather than an energy sink. This begs other questions, such as how near the cone the stuffing should be allowed, and the effects of smaller enclosures on midrange performance.

I just know that IME, one can reach a point in stuffing that is detrimental to the sound, and it is significantly before 'heavily stuffed'.

C

[Deward Hastings]

> The dipole design of course doesn’t magically eliminate the back wave, it just wraps around the baffle, or bounces off the wall instead

Indeed. At the lower frequencies where it "wraps around the baffle" one gets cancellation, and the "signature" dipole radiation pattern (and reduced room reflection from the side walls and ceiling). In any case (even after the driver starts beaming, unlike monopoles) the back wave does reflect off the wall behind the speaker, perhaps contributing to the "soundstage" that dipoles (or other speakers with rear-firing drivers) are often noted for.

As it happens I much prefer that wall to be (relatively) absorptive and diffusing . . . in my listening room it's a wall-to-wall (excepting the TV and built-in subs) bookcase with sound board behind the books. Only the upper wall and cove (lath and plaster) is significantly reflective. The room is not notable (g) for its "broad and diffuse sound stage", but rather for "precise imaging" and the easy ability to discern separate instruments (when the recording permits). That, obviously, is how I like it . . . years ago when I had Maggies I even hung fiberglass bat behind the ribbon to keep the tweeter from radiating to the rear. The first speakers I had in the room were Dyna A-25, actually sitting in the bookcase. No reflections there (g).

My work-in-progress HT center channel is itself a dipole with highly absorbant rock wool panels mounted a yet-to-be-determined distance behind the drivers, to produce a true unidirectional speaker with a half dipole radiation pattern. It remains to be seen how well it will work, but you can see the direction of my thinking. I want to hear the front wave only. Period. No reflections. Period. At least not from the wall (or box) behind the speakers . . .

The "sound" I'm looking for might be described as "the clarity of headphones, over a wide listening area, and with good image localization where the recording permits". I'll leave it to the recording engineer to determine how much of the acoustic signature of the recording venue ("soundstage") should be preserved and projected into my listening environment . . . if I need to "create" one I'll do it electronically when needed, not permanently through the (fixed) reflection/reverberation pattern of my listening room. That perspective sort of colors (g) how I react to boxes big and little, for several reasons (though I do use them for HT surround sources, mounted flush to the wall).

[dlr]

is that the target highpass of the midrange is not being addressed. There are several ways to do that. Use a box alone. Use a box stuffed lightly. Use a box stuffed densely. Use a box with some version of the above, but with a crossover highpass network.

In every case above, the target has to be met. If not, the arguments above are all moot. The only way to assess the impact of a box/stuffing/dipole is to design the midrange such that the highpass meets a predetermined target.

The case of comparing a single driver box with the changes alone being stuffed/unstuffed is useless. This is comparing two differing highpasses. If the matching woofer lowpass is not adjusted to compensate PERFECTLY for the altered midrange highpass, the comparison is moot.

The single issue that has been discussed that is valid as a separate issue from the above is the point about internal reflections. These can be and are easily measured since they are manifested as resonances in the midrange's passband, regardless of the highpass used.

In this regard, I'm fully in agreement with Madrok (er, Damkor).

dlr

[Deward Hastings]

True enough, at least to the extent that "stuffing the box" does change the highpass. I'm not sure how much that is, though, and it can be easily compensated for by changing the lowpass on the next driver down. In any case the box is typically small (relative to wavelength) at the highpass corner. If "stuffing" does change the apparent size of the it's not clear to me how that would effect the overall "tonality" (live v. dead, for example) apart, perhaps, from changing Q (which might itself be beneficial).

At the upper end of the midrange passband, however, the box can be multiple wavelengths in size, and one sees the obvious problems of standing waves and re-radiation of reflected waves. It's true that the standing waves should show in the frequency response . . . what's not clear is that the overall reflected/delayed/re-radiated sound will. While it's obviously correlated in some way to the input signal that correlation is quite complex, and would probably be lost in the resolution of the measurements. Perhaps it would show as a general rising response of the driver in-box compared to infinite baffle, but that would still miss the reality that the delayed re-radiation is now different from the applied signal in more than just amplitude, and that when summed with the original it is more than just "louder".

As with most "distortions" I'd be content with a metric that could let us say, for example, that the back radiation was "99% absorbed", or "40dB down", or . . . and would argue that "stuffing" beyond that point was overkill (g).

[waynew]

I see the logic in the assertions of both you and dlr. I would still like to see one of you guys perform this simple expirement and quite possible extract the metrics that can enable all of us to make better use of stuffing. There is plenty of information on stuffing types, e.g. fiberglass vs. dacron vs. wool. There is also some info available for changing the Q of an enclosure. There is not a consensus as to how much is enough. That being said, my main point is, was and will continue to be that too much stuffing can be detrimental to the sound quality of the midrange. Or, as CurtC so succinctly put it, it (when using too much stuffing) deaden's the midrange to a point of lifelessness.

[dlr]

The box needs sufficient stuffing to adequately damp the back wave. Less than that and the internal standing wave resonances will become a factor. If the box alone is used as a highpass, then it must be properly sized and stuffed. The goal is to achieve a specific highpass target.

If and electrical highpass is used in conjuction with the box and a specific highpass is achieved, stuffing it additionally will again alter it, but the electrical (whether active or passive) network must, MUST be modified to again achieve the target. If not, then the comparison is not valid. Stuffing in a box is part of the design.

If it's overstuffed and not compensated, it's not designed correctly.

If the lowpass to the woofer is changed to compensate, it's not the same design.

To say that adding too much stuffing is the reason for a dead midrange does not hold up under analysis IMO. If the overstuffed box is compensated through a crossover to return the combined box/crossover back to the original target, there should be no discernable difference, unless the previous amount of stuffing was less than adequate for damping the rear wave.

As long as the two passband examples for the midrange are identical, then there really is no difference. There are a number of ways of achieving a specific target, box only, box/XO, various stuffings, etc. But if the passband result is the same, in will sound the same with one exception.

That exception is that when only a box is used, stuffed in whatever manner, then there is only a lowpass network limiting the signal to the midrange driver. This means that the entire signal below the passband is going to the coil, which will increase thermal conditions adversely. But this is a separate issue from the discussion of stuffing too much or not.

dlr

dlr

[damkor]

[damkor]

...can certainly hear the effect of an open back to the midrange! And can hear the increased box noise from a lack of stuffing. And I can hear the effect of cramming the box so full, that Q is INCREASED and you get a warm boomy sound above the xpoint.

But I have honestly not been able to compare the sound of 30, 40, 50, 60 70 percent stuffing with wool, acousta-Stuf, Polyfill, eggcrate foam, fiberglass, on rear walls, side walls. Life is not long enough.

But I do consider best practice to be to densely pack the walls with stuff, and loosely fill almost the entire space behind the midrange driver with looser fill, leaving a little empty space (no added air resistance?) behind the driver.

[dlr]

No. The arguments were that the problem alone was in overstuffing. The transfer function no longer matching the desired target was totally ignored.

The evidence is statements that the difference is is simply due to over-stuffing, not in the altered transfer function that can then be corrected for the additional stuffing. Stuffing is an integral part of a design. If the box is stuffed or stuffed additionally AFTER the design is complete, then the design is altered, due to alteration of the transfer function.

It's emphatically no different than taking a design and altering a component (cap, inductor, resistor) in the highpass AFTER constructing it. The highpass transfer function is altered, thus the design is altered. It sounds different, sure, it should. The transfer function is changed.

Essentially, I agree with your original statements about the need for and amount of stuffing. Over-stuffing can be compensated, but it may require larger highpass components than would be necessary for a larger, more optimally stuffed box. But it's part and parcel of the design.

dlr