What's the hot new 8" midrange these day?

[sfdoddsy]

I'm pondering upgrading the midrange drivers in my surround speakers. They are 2 way active dipoles in the vein of an Orion top panel but using the Seas L21 crossing over to a Seas 27TDFC at 1.44kHz at the top and 100Hz at the bottom to a sealed DPL12 sub.

There's no particular reason for the change except that I feel doing something.

Any potential replacement would need to cover the same range and have good xmax because they are actively EQed.

So, any thoughts?

[Piotr]

Steve,

wouldn't the obvious be to look at your main speakers and see what driver are being used there?


/Peter

[sfdoddsy]

Yes, that would be obvious. But there might be something funkier than the Excel W22s out there.

Something along the lines of the Adire Extremis.

[JonMarsh]

Fads come and go (try to buy an Extremis new at this time), but the classics are around for a while.

You could consider the Seas Nextel version of the W22, but I think it's more a bit of a step back, a Scan speak cone with a Seas motor....

The W22 still seems to be the reigning midrange distortion champ, so if that's your main criteria (you did say 8" midrange, not 8" woofer), there's your answer. :W

~Jon

[sfdoddsy]

Strictly speaking it's a midrange since I'm crossing at 100Hz, although as I said the more p-to-p the better.

Anyways, it sounds like I may as well as just grab another couple of W22s. How tiresome.



Next you'll be telling me there's no obviously better bass driver to replace my DPL12s in the main speakers.

[JonMarsh]

Strictly speaking it's a midrange since I'm crossing at 100Hz, although as I said the more p-to-p the better.

Anyways, it sounds like I may as well as just grab another couple of W22s. How tiresome.



Next you'll be telling me there's no obviously better bass driver to replace my DPL12s in the main speakers.

Now that may be another matter.... the DPL has a pretty linear motor with regards to BL, but no copper in the gap... can you spell "inductivity modulation"? That means that it is not state of the art above 60-70 Hz or so. There have been a lot of woofer introductions since the DPL that do pay attention to that- generally just look for the drivers with Le of 1 mH or less, comparable or higher Xmax, and BL linearity. There's always the RS315HF, with a nice blend of woofer and subwoofer capabilities. A newcomer I haven't looked at yet is the XJ-12 available at DIYCABLE - I have the XJ-15, but haven't had time to test it yet, either. Both look very promising from specs, but the 15 is no longer available- the Tempest-X is the "real" version of that driver, as I understand it.

I'm sure others will suggest even more interesting 12's, but those are the first that come to mind with well controlled Le.

I'm a fan of the Aurasound NS12-513, but it's a little closer to a woofer than a sub woofer; lower Xmax, and even more extended top end and distortion that steadily drops as you go up towards 1 kHz.



~Jon

[cjd]

Strictly speaking it's a midrange since I'm crossing at 100Hz, although as I said the more p-to-p the better.

Thing is, the W22 really doesn't have a match on the top end of its usable frequency range. If you were going woofer, the Dayton RS225 is a different option, plus others I believe. But for upper end, the W22 is where it's at.

The only other alternative that comes to mind would be a big paradigm shift perhaps, but to look into some of the prosound 8" midrange drivers. You'd have to cross higher on the bottom though as they have no excursion to speak of on average. However, in the high(er) excursion (sub)woofer world there are now some superb options that can cover quite a bit higher in frequency range making this a distinct possibility.

You could also push that lower end up to 150ish and consider a pair of 180mm drivers.

C

[Paul Ebert]

Steve, if you're patient, you could see how the SB Acoustics 8" driver turns out.

Is there an Accuton driver worth considering?

Chris, I'm curious about your recommendation of pro drivers. Are there ones that can match the W22 with respect to distortion? How about linearity?

[AJINFLA]

Hi Steve,

IIRC, you fear not DSP XO's and use not wimpy tubes, so lower impedance no problemo.
Here ya go rw220
How's your Norwegian? rw220

cheers,

AJ

[joecarrow]

I'd be curious to see some of those pro 8" drivers measured. I once built a two-way with an Eminence Beta 8 and a Morel MDT20, and the midrange clarity was quite a step up from anything I'd yet heard in college.

I had no crossover skills at the time, so the frequency response was garbage (until I corrected it with parametric EQ), and the dispersion was worse- beaming like crazy until the tweeter kicked in. I ran the darn things up to around 5 khz with just a first order crossover. I know it's not very subjective to sing any praises of such a mess, but it sure seemed like the midrange was nice and clean and clear.

Regardless- they would go LOUD with 250 watts behind them. A higher-end pro 8" with a proper crossover and a waveguide tweeter ought to be a stunning combo. Isn't the Accuton used in the Isiris fairly efficient? If so, it might go fairly loud as well. I haven't been properly following that thread.

[JonP]

Thing is, the W22 really doesn't have a match on the top end of its usable frequency range. If you were going woofer, the Dayton RS225 is a different option, plus others I believe. But for upper end, the W22 is where it's at.

C

Thinking about Jon's comments on the RSS315HF... I started wondering how well the new RS 8" sub might do in the not so sub ranges... Not sure if it has a better bag of good features than a RS225, probably goes lower and cleaner, may not go higher, more excursion but costs twice as much, etc. I'm waiting to see some of our measuring gurus check them out...

[Jed]

Thinking about Jon's comments on the RSS315HF... I started wondering how well the new RS 8" sub might do in the not so sub ranges... Not sure if it has a better bag of good features than a RS225, probably goes lower and cleaner, may not go higher, more excursion but costs twice as much, etc. I'm waiting to see some of our measuring gurus check them out...

What are they like 83db/1watt? I guess you can't have everything- but they do look nice.

[JonMarsh]

Thinking about Jon's comments on the RSS315HF... I started wondering how well the new RS 8" sub might do in the not so sub ranges... Not sure if it has a better bag of good features than a RS225, probably goes lower and cleaner, may not go higher, more excursion but costs twice as much, etc. I'm waiting to see some of our measuring gurus check them out...

That's the only reason I bought a sample of the RS210HF to test, is to see how they might fare as a driver in the 60 Hz to 1,000 Hz range- I'm not doing any 8" subs!

Sensitivity at 86 dB is a little low for an 8", but close to what I measure on many 7's. Linear Xmax is not rated all that high, at 9 mm, but the suspension is capable of more, and it uses tinsel leads woven into the spider. I'm hoping to NOT see the typical notch suckout near field that curvilinear cones with phase plugs show, whether they're from Seas or Dayton.

Just haven't had time to make a test baffle and do the deed....

[capslock]

[QUOTE=JonMarsh] I'm hoping to NOT see the typical notch suckout near field that curvilinear cones with phase plugs show, whether they're from Seas or Dayton.[QUOTE]

What frequency? Could that be an acoustic short circuit through the vent between plug and former? Or, since you mention curvilinear cones, do you think this is a cone mode?

The only plugged curvilinear Seas metal cone that I am aware of is the H1224, all other W and L drivers have a straight profile, as did the H1224's predecessor, the H1142.

[JonMarsh]

you've corrected me, it's mainly phase plug drivers- the RS225 has that problem around 1500 Hz, the W22 around 1800; I believe it's a cone mode (measured nearfield) because it correlates with linear distortion measured using shaped sine burts as per SL.

Concave cone diaphragms like the Hi-Vi M8a or the Aurasound NS 10 don't exhibit this behavior. If it weren't for rear masking by the d
frame, I'd try the RS210HF in the 3way upgrade I'm planning for the M8ta.

[AJINFLA]

The only plugged curvilinear Seas metal cone that I am aware of is the H1224, all other W and L drivers have a straight profile, as did the H1224's predecessor, the H1142.

The one I just linked appears to have a curvilinear cone. The motor design might be a preview of the W23

cheers,

AJ

[tktran]

I'm probably dreaming, but what would be lovely is a hard-coned 8" with a breakup node far past 5Khz. Say 7-10KHz.

Even the new SEAS W16NX, if compared to the W15CY/CH series, has an improved throw, but main breakup node has dropped down to 5Khz. What a disappointment, and a step backwards IMHO.

The W22EX is often declared as one of the best drivers from 100Hz to 1.5Khz, and I know that Linkwitz declared the W22EX a winner for his Orion (after taking a a few isolated measurements) but I wonder how the old kevlar 21W/8554 would fare in rigourous measurements like those conducted by John Krutke or Mark K.

I would not be suprised if the non-musically related odd-order distortion is lower, even if the even order is slightly higher.

[JonMarsh]

Having built systems with the old 21W/8554 (Linkwitz did too, at Audio Artistry), I have to say the motor linearity is good, but the cone is not so nice, with energy storage issues (linear distortion) that start at 800 Hz.

Though the W22 starts getting into linear distortion problems above 1200 Hz, the even and odd order HD up to about 1 kHz are pretty much SOTA, though the original version of the Peerless 830883 has lower third order above 500 Hz; both hit the -60 dB range from 100 Hz to 500 Hz at ~90 dB output. The W22 slowly climbs from -60 to about -52 dB. The 883 had HD near -60 for both HD2 and HD3 up to nearly 2 kHz.

The SD1 motor has an advantage in distortion below 100 Hz. Doesn't seem to above, though. The linear distortion is the deal killer for the 8554. It's OK in a 3 way, not suited IMO to a two way.

~Jon

[tktran]

Hi Jon,

I'm not sure whether it's the SD-1 magnet system, or whether it's really the suspension system that keeps the HD under 100Hz low in the SS drivers.

Have you seen any full range HD sweeps of the 8554? You know the type that John Krutke does with SoundEasy? I find it very helpful in comparing the HD spectrum of similar sized drivers.

Both the W22EX and W18EX have loud resonances at 5Khz, and I wonder if the W22EX shows a similar rising HD pattern to the W18EX, with 3rd and 5th order harmonics increasing from 500Hz to 2KHz? (gray and blue lines).

[Nemophyle]

i don't know of anybody having tested it , but this could be a good midrange driver

Scanspeak Alu 8" Woofer

9mm one way excursion, break up seems up far, i just wonder about the suck up in FR at 500hz, unlikely to be breakup, but it could be a cone / surround problem. Anyway you'll need to test it to see if it's not a measurements artifacts and if it generates energy storage

a shame scanspeak doesn't do smaller alu drivers i'd get my hands on a 6" curvilinear alu driver from them

[TacoD]

i don't know of anybody having tested it , but this could be a good midrange driver

Scanspeak Alu 8" Woofer

9mm one way excursion, break up seems up far, i just wonder about the suck up in FR at 500hz, unlikely to be breakup, but it could be a cone / surround problem. Anyway you'll need to test it to see if it's not a measurements artifacts and if it generates energy storage

a shame scanspeak doesn't do smaller alu drivers i'd get my hands on a 6" curvilinear alu driver from them

I use the 10" as bass in a 3-way. In the past I had plans to do a 2-way with this 8" in combination with one of Jon's cauer filters . The sound from these woofers is clean and bass is tight. Measurements of these Scanspeak alu cones are very good (measured in HobbyHifi). Distortion is as low as the Aura's but the impulse response is cleaner. The cone breakup is also better behaved.

Soundwise I would prefer Audiotechnology. I really like the 15H and 18H, but also the 23I is very nice sounding.

[JonMarsh]

Hi Jon,

I'm not sure whether it's the SD-1 magnet system, or whether it's really the suspension system that keeps the HD under 100Hz low in the SS drivers.

Have you seen any full range HD sweeps of the 8554? You know the type that John Krutke does with SoundEasy? I find it very helpful in comparing the HD spectrum of similar sized drivers.

Both the W22EX and W18EX have loud resonances at 5Khz, and I wonder if the W22EX shows a similar rising HD pattern to the W18EX, with 3rd and 5th order harmonics increasing from 500Hz to 2KHz? (gray and blue lines).

Haven't seen any HD sweeps of the 8554- my issues was cone modes starting at relatively low frequencies. Not particularly high amplitude, but audible, like in the Eton 8" midwoofer. While it's peak was at 3.2 kHz, it had a lower notch type mode at 1400 Hz.


Here are nearfield sweeps of the 8554 at center, midpoint, and near the edge of the cone.





Here's the HiVi M8a measured under the exact same conditions; it uses an aluminum/magnesium cone that's an inverted concave shape; very rigid until the primary breakup mode. Doesn't have nearly as nice a motor as the SS, but the cone is far better behaved in a classic sense. This is why for me the upper frequency for the 8554 is 400-600 Hz.

[Fred333]

Thanks for the sweeps. Always good to learn new info.

[tktran]

It would be interesting to see a harmonic distortion sweep of the HiVi M8a, compared to the 8554.

The rising SPL starting around ~1.7Khz, and peaking just around 2.3Khz may manifests also as rising 3rd order harmonic distortion, peaking at around ~ 2.3 KHz / 3 ~= 750-850Hz or even higher. Jon as you, or other musicians, may know, 3rd or other odd order harmonics have nothing to do with musically related content. Even if it had an advanced motor, it's optimal application is under 1KHz.

I'd take the SS8554 with the advanced, higher HD2 but lower and flatter HD3, even if it does have a dip at 800Hz to due cone/edge deflection.

[Dennis H]

As you or other musicians may know, 3rd or other odd order harmonics have nothing to do with musically related content.

That's not really true. Here's the spectrum of a Fender bass playing a low E with a fundamental note at about 41Hz. The 2nd harmonic (82) is the strongest note, much stronger than the fundamental, but the 3rd harmonic (123) is still a bit stronger than the fundamental and the 4th (164) is about the same as the fundamental.

PS, the circled spike is 60Hz hum.

[tktran]

Dennis,

Thanks for your response. I agree that a musical tone of any instrument consists of a mixture of harmonics of various levels. This is what separates an instrument playing 'A' to a sine tone at 440Hz.

So when we play the low E of this Fender through a loudspeaker, the loudspeaker has to reproduce the low E, and all those harmonics too.

In the ideal world your loudspeaker would produce the exact tone of that low E, which includes the fundamental, all those harmonics, at exactly the right levels, and nothing more.

But a real loudspeaker will also produce some harmonics of it's own. So it will play the 41, 82, 123 etc. But also all of:

ie. 41 x 2, 41 x 3, 41 x 4, 41 x 5 etc.
82 x 3, 82 x 4, 82 x 5 etc.
123 x 2, 123 x 3, 123 x 4, 123 x5 etc.

AWe have a hash of higher level harmonics.

Introducing more 2nd order harmonic distortion to all those harmonics, like how a valve pre-amp or digital-audio-workstation does, is gives a pleasing sound, "rich", or "musical" sound. In theory, this like hearing another E, except 1 octave higher.

It would certainly be interesting to see the harmonic spectrum of the Fender through the pickup, then what it looks like when passed through the pre-amp stage, and through the power amp stage, then finally through the bass-guitar speaker.

[Piotr]

tktran

This is what separates an instrument playing 'A' to a sine tone at 440Hz.

That and also the start/transient and stop/decay of a note differs a lot between instruments.

But a real loudspeaker will also produce some harmonics of it's own. So it will play the 41, 82, 123 etc. But also all of: ie. 41 x 2, 41 x 3, 41 x 4, 41 x 5 etc.
82 x 3, 82 x 4, 82 x 5 etc.
123 x 2, 123 x 3, 123 x 4, 123 x5 etc.

AWe have a hash of higher level harmonics.

And more important, there will always be intermodulation distortion (sum and difference tones) when more than one tone is passed thru a non-linear device/system. Tiny small harmonic distortion products will affect the sonic little if they coincide with harmonics in the original note. Depending on phase harmonic distortion will add or subtract a little from the originally occuring harmonics. Now, an electric apparatus will have matematical perfect harmonic distortion, in contrast to some musical instruments that actually shift the harmonic slightly making them non-perfect and actually slightly non-harmonic. This means that instead of a 100Hz note giving a 200Hz 2nd harmonic it may be 202Hz.

Non-harmonic overtones is natural and is also what separates the timbre of different instruments. Problem is that the overtones of a "non-harmonic" instrument will not match the overtones created in the non-linear process in the apparatus and possibly will create problems.

Introducing more 2nd order harmonic distortion to all those harmonics, like how a valve pre-amp or digital-audio-workstation does, is gives a pleasing sound, "rich", or "musical" sound. In theory, this like hearing another E, except 1 octave higher.

Non necessarily so. Wether a tube circuits non linearity results in dominant even or odd order distortion depends on the topology. Also a lot of non-musical IM products will be created even if a circuit has pure 2nd harmonic on a single sinewave. Distortion from a tube amp with a high degree of non linearity gives (to my ears) a very non pleasing result.

Did a blind listening test recently on a couple of amps. The procedure was to listen at the signal directly from the soundcard and then level adjusted after the amp driving a complex load. A 2A3 SE amp of about 0.1-1% distortion and a LM3886 amp was the contenders and the LM3886 kicked the tube hard in the butt. The tube amp sounded compressed, grainy, flat and non musical. The LM3886 on the other hand had something like 100 or 1000 times less distortion, both symmetric 3rd, 5th... and asymmetric 2nd, 4rth.. and sounded very good. Actually I could not tell the LM3886 from the direct non-altered signal meaning that in this particular set up with that particular material the LM3886 was transparent and totally neutral.

A couple of years ago me and a friend recorded some acoustic music in a big room with natural reverberation. We used an A/B set up with a pair of Earthworks QTC1 and compared different preamps. One of the amps used was the DW Fearn VT-2 tube mic pre. This can be adjusted for the level of "cleanlyness" and "tubyness" and the clean low distortion settings produced the best sound to our ears on voice and acoustic steel strings. The "rich" setting with lots of tubedistortion sounded kind of tired and thick in comparison to the more clean setting.

I also compared the GamuT D200 mosfet amp to the Audio Research VT100 tube amp and D200 was more natural and pleasant sounding than the VT100. The D200 actually has the dominant 2nd order harmonic on a sine wave while the VT100 has (at least on some levels) dominant 3rd harmonic. As mentioned, wether a circuit gives dominant even or odd order distortion can depend greatly on the level of the signal being amplified.

For what it's worth!


/Peter

[capslock]

Did a blind listening test recently on a couple of amps. The procedure was to listen at the signal directly from the soundcard and then level adjusted after the amp driving a complex load. A 2A3 SE amp of about 0.1-1% distortion and a LM3886 amp was the contenders and the LM3886 kicked the tube hard in the butt. The tube amp sounded compressed, grainy, flat and non musical. The LM3886 on the other hand had something like 100 or 1000 times less distortion, both symmetric 3rd, 5th... and asymmetric 2nd, 4rth.. and sounded very good. Actually I could not tell the LM3886 from the direct non-altered signal meaning that in this particular set up with that particular material the LM3886 was transparent and totally neutral.
/Peter

Most soundcards, even high-end cards like E-Mu 1212m or ESI Quatafire 610 leave a lot to be desired when it comes to clock generation and routing, supply decoupling and analog output stages (i.e. op-amps and coupling caps). I suspect that your soundcard generated a distortion signature that was very similar to the LM3886's but on a higher level, so whatever the M3886 added was simply swamped by the soundcards's signature.

[capslock]

Just to bring the thread back to the topic, the Usher 8137K seems a strong contender. Motor is better than Scan Speak, cone pretty rigid (Kevlar / Carbon), first and only top end break up is at 3.7 kHz, i.e. a tad lower than W22, but also not as strong. The only other thing that don't like (except for the looks) is the 5 dB step at 1.1 kHz. It can be dealt with in the XO easily, but not having any idea what kind of mechanical mode this is, I suspect it will add a signature of its own.

[Piotr]

Capslock,

the Audiofire4 have performance on the level with the Lynx cards.

Chances that the distortion in the DAC would swamp those in the amplifier is close to non existant. The THD of Audiofire4 is 0.0008%. 2nd and 3rd at
-107dB.

Please also notice that I mention this results being specifik for this particular test. It's possible that something with higher resolution than my HD600 cans would highlight some problems not detected in this test. Also other more careful choice of program material may have given other results.

Listening blind and preferably in a "before/after" set up is enlighting. Try it!

Sorry for OT mr thread starter! ops:


/Peter

[capslock]

Peter,

the Echo Audiofire4 seems to be based on the same BridgeCo chipset / IO board that all the other Firewire boxes (ESI, Terratec, M-Audio) use.

I had to take my Quatafire apart because there was a hardware problem, and I can tell you that the implementation of the PLL stinks, that clock routing is far from optimal, that the convertors are by no means the flagship products, that the OP-amps are poor, that the analog power supply layout and decoupling is poor and that there are plenty of 'lytics in the signal path. Still, by reducing the signal level to - 7 dB fs, I can get all harmonics to below - 130 dB, surprisingly good given manufacturer's specs, converter and op amp specs and the overall design, but also do I think that I could do better.

Also, the clock issue can be easily seen with a spectrum analyser such as ARTA. There is only a 24.576 crystal on the I7O board, and 44.1 / 88.2 kHz clocks are generated by a PLL. Those frequencies will have a higher noise floor and broader base of the harmonic. By switching back between 44.1 and 48 kHz a few times, I can make the 48 kHz spectrum look worse than the 44.1 kHz spectrum. I suspect the PLL is not always properly shut down and gets to contaminate the 48 kHz clock.

[sfdoddsy]

Thanks for all the, as usual, extremely intelligent and cogent responses.

But it does appear that SL knew what he was doing with the W22 and that the L21 (or current L22) is indeed the best alernative, especially in my usage.

I have two W22s in my mains, but the extra Xmax of the L22 allows me to use one of them in the surrounds.

None of the drivers mentioned seems to get close to SEAS claimed Xmax for the L22 (12mm P-P linear, 21mm Max). Whilst I don't believe those figures for a second and would guess about half that, I have never bottomed the durrounds, whereas I have the dual W22s.

The L22 is gasping for breath at the high crossover I use, but since I can EQ it to tip up at 1.4 I do. And the breakup mode isn't an issue.

Intially I started the thread because I just felt like changing something, so I'm a little diappointed that five years alater there is nothing worth chaning to.

Although, of course, dipole friendly 8 inch midranges that can play up to 1.5K aren't your normal cup of tea.

The one suggestion I am curious about is the Dayton 8 inch sub and whether it could be pushed upwards, although again the Xmax on that seems a little feeble for a 'sub' driver.