Three Way Evil Design Study

[TacoD]

I had the same experience about finding the best trade-off for 'a' and the centerfrequency. I see 'a' as 'the amount of fill' of the mid driver. From experience I can tell that finding the sweetspot is as difficult (or even more when you consider the constraints on the drivers) as a normal xo. It isn't as plug and play as when you think when first reading the Duelund page.

[Evil Twin]

I had the same experience about finding the best trade-off for 'a' and the centerfrequency. I see 'a' as 'the amount of fill' of the mid driver. From experience I can tell that finding the sweetspot is as difficult (or even more when you consider the constraints on the drivers) as a normal xo. It isn't as plug and play as when you think when first reading the Duelund page.

Agreed- NOT plug and play- further complicated in that "a" or aleph also tunes the Q of the filter transfer functions-

but with fascinating possibilities that are releatively easy to calculate and consider, once a worksheet is setup. Realizing the described acoustic transfer functions will not be any simpler that for a conventional network, but it also should be no more difficult as regards matters like BSC, and I most heartily approve the potential reduction in number of filter elements for the midrange. Consider the disaster one winds up with for a LR4 or evan all pass 3rd order bandpass.

Three projects are immediate possiblities for this approach.

There is much work to be done to prove this out in a useful form, but the essential qualities are intriguing, and less daunting in execution than a massive battlestar.

Let us hope there will not be some similar weakness in the final realization...
my master is not known for his forgiving nature.

[TacoD]

You are correct that the q is also altered and indeed the small amount of components in the midbas branch is very appealing, but, and you already applying this in your other designs, you need compensation for everything, zobel, lcr on resonance peaks (driver and enclosure) to name a few. Ending up with >30 components in a filter (not counting parallel C's) is standard. My attempt at a 4way Duelund was even more daunting (50 components, while omitting the 800 uF series cap for the tweeter )

[Evil Twin]

You undoubtedly recall these words...

A passive network needs spacing, so use bottom and backside of the
loudspeaker box for it — the bigger the better. Also do not expect this
important part to be cheap. I personally use more money on the dividing
network than on the rest of the loudspeaker.

-Steen Duelund

Yes, the supporting components do add up, but with fewer components in the bandpass, there are less interactions to fight.

A four way Duelund? Most impressive... you have been concealing your skills, young Jedi.

[TacoD]

You undoubtedly recall these words...



Yes, the supporting components do add up, but with fewer components in the bandpass, there are less interactions to fight.

A four way Duelund? Most impressive... you have been concealing your skills, young Jedi.

Less electrical interactions that is, acoustical overlap demands for clever spacing in addition to demanding boundary conditions for the used drivers.

You are correct about the young part, but the skills you are referring to are still very limited when compared to yours.

[Evil Twin]

You are becoming the Noah's ark of Accuton.

sigh.... apparently so.

As we speak, a pair of C90-6-79 are in transit by Imperial shuttle to the Seinar Speaker labs.

Perhaps I should just give in to the Dark side fully, and order a pair of C90-T6-89 also...

Where's my ordering droid?

[CraigJ]

Where's my ordering droid?

I'd walk over and get your drivers, but would that really help?

Cj

[Jonasz]

sigh.... apparently so.

That didn't sound like a quote from the evil dark lord...

[jdybnis]

With some experimentation with the damping coefficient aleph and the center frequency, a proposed 3 way system transfer function can be developed, which should then be checked for intended playback levels using tools such as SL's SPLMax which calculate the maximum "linear" SPL possible for a given driver Sd and Xmax and operating frequency.

It would probably be worthwhile incorporating part B of SPLMax into your worksheet. SL kindly gives his formulas at the bottom of SPLMax for determining the displacement necessary to produce a given SPL level at a given frequency. It would make a nice addition to your transfer function plots. It's hard to tell what frequency the max excursion is going to be at just by eyeballing it.

[TacoD]

When you look at the pricing of the diamond offerings, the ceramic drivers immediately look more reasonably priced ...

[Evil Twin]

It would probably be worthwhile incorporating part B of SPLMax into your worksheet. SL kindly gives his formulas at the bottom of SPLMax for determining the displacement necessary to produce a given SPL level at a given frequency. It would make a nice addition to your transfer function plots. It's hard to tell what frequency the max excursion is going to be at just by eyeballing it.

That will be included in the transfer to Maple 10 as well.

When you look at the pricing of the diamond offerings, the ceramic drivers immediately look more reasonably priced ..

One is truly slipping into the thrall of the Dark Side when considering the diamond tweeter pricing, and pondering the possibility of financial options to fund such insiduous purchases... such as selling one's offspring into white slavery.

In the mean time, I will have the droids order a pair of C90-T6-89 to take my mind off the D30N-6-36. It's one thing to have the necessary credits on hand, it is quite another to expend the price of a standard land speeder on just a pair of tweeters...

[jkrueger]

Following closely

Did somebody say diamond? Oh Great Leader, if such unobtainium is used you will lose many of those from the outer reaches of the galaxy who wish to participate in this "Uber" design. I hope that your plans still include the original budget. ;x( ;x( ;x( ;x( ;x( ;x(

[Evil Twin]

The "budget" for the M12ta-SD is set and firm- fear not. The final driver selection is not- but it is close- by the end of next week I should have the additional midranges to test and a choice ready.

Enclosure preparation (panel lamination) has already begun.

[Jed]

Does the Dark Lord ever sleep?

[jkrueger]

cabinet info

Happy New Year!!!! Here's to the most exciting project of 2007!

Edit* Are you almost ready to give out cabinet details Jon? ;x( ;x( ;x( ;x(

[Evil Twin]

My daughter gifted me with an unexpected present this holiday- a viral respiratory infection, which is not something to be taken lightly given the state of my lungs since those pivotal events after the Clone Wars.

Your patience will be suitable rewarded in due time... most likely just a little later this week.

Does the Dark Lord ever sleep?

Only when ill, it seems- 12 hours last night. However, more construction materials were purchased yesterday for the new design study, lest we fall too far behind the Emperor's target schedule.

To all, we wish a productive and noteworthy 2007, as you reckon time locally.

May the Force be with you.

[TacoD]

I am reading about a new Magnesium driver from Seas, W16nx001. It will be introduced on the CES (Jon if you go there, can you take some pictures?) Maybe this is the pistonic driver you need. The surround is different, it has some special relief (my english doesn't permit a better description). Also the basket has that grey coating like the Nexels drivers. Rick Craig has similar message @ madisound forum. But I have finally received my own copy of Voice Coil.

[Evil Twin]

This year I'm working at home on projects instead of going to CES. Usually a CES every few years get's the job done on seeing what's currently out there.

Hopefully there will be some announcements with specific information, and I'll have to add it to the Evaluation list. C89s ordered today. Almost went for a pair of D2904/71000 (slightly more expensive version of 66000), but frugality and pending Imperial budget restrictions won out.

[ssabripo]

I will be there next week Dark Lord....I got a list of vendors I'm gonna check out, but look at the CES website list of attendees, and tell me if there is one in particular you want me to check out.

signed,

[Evil Twin]

I will be there next week Dark Lord....I got a list of vendors I'm gonna check out, but look at the CES website list of attendees, and tell me if there is one in particular you want me to check out.

signed,

So, for once I can have the rebel spies working in my favor? Interesting... most interesting.

Seas will be there, at the Ventian Towers (many former exhibitors at the Alexis Park for Hi End will be at the Towers this year); reportedly they have a new W16nx001 midrange which might be of interest-

Parts Express usually has a large booth at the entrance to the Alexis Park; with Darren gone, it's hard to gauge what their particpation will be.

Depending on your tastes in recorded music, you might want to visit Elusive Disk and Acoustic Sounds, they usually have some interesting items at attractive prices, including vinyl, SACD, and XRCD.

Ayre should be at the show, and I expect them to be displaying the final version of the MX-R mono block amplifier (cutest little non-digital amp you've ever seen), I dont' have any word on whether they're exhibiting for sure or not, given Chareles Hansen's accident.




Cary Audio last year had some of the better sound, exhibiting with Dynaudio speakers. Worth a listen, especially the SACD player. Hmmm, I see they have an opening for a senior design engineer... I wonder if they have anti-Sith bias? It's so common these days... They're at the Venetian Tower.

Balanced Audio Technology is also a reasonably good bet. No news posted on their web site.

Rives Audio (acoustical engineering) will be demonstrating with Talon Audio, Wadia, VAC, and Shunyata at the Venetian 4606; that should be interesting.

Avalon doesn't normally exhibit at the CES, but often several firms use their speakers (Argento Audio, Hovland Audio, Axiss Distribution, and they do sometimes host a hospitality suite away from the show venues. Might be worth looking into those possibilities.

A number of companies who's wares I was quite interested in auditioning last year frankly disappointed- inlcuding the new Quad ESLs, SoundLabs, PipeDreams.

CES is rarely a good opportunity for seeing video products under good circumstances, but the new JVC and Cinetron LCOS 1080P projectors in the 6K price range should be interesting.

[Evil Twin]

Accuton mids

The C89's came in today, adding to my collection, so the measurements can start tomorrow.

Accuton and Madisound are perhaps not completely forthcoming, because the Madisound Accuton PDF describes the C79 as a 3" ceramic dome midrange, and the C89 as a 5" Ceramic dome midrange.

:nonod:

Both have 37 mm titanium VC.

The rated Sd of the C79 is 75 cm2, whereas for the C89 it's rated at 83.3 cm2.

The C79 has a rated Xmax of 2.6 mm, while the C89 is 3.1 mm. This is curiously at odds with the differences in rated sensitivity; which is slightly higher for the C89. The magnet structures appear identical externally.

Le for the C89 is rated at 0.4 mH, versus 0.12 mH for the C79. Curious. Consistent with a shorter VC, but more than I might expect.

The C79 has the "ears", which modify the diaphragm breakup characteristics, pushing it up in frequency and damping it.

The C89 has a longer VC, with the magnet structure set further back, and less rear masking- a benefit for baffle design and mounting.



Exhibit "A" (for "Accuton") is the picture below, showing from left to right, a C88, C89, and C79.



The C89 uses the newer coarser hex grid grille, as used in the C173N-T6-90.

Looking at and measuring the diameter of the diaphragms, it's obivous that the C79 actually has the larger diaphragm, with a diameter of 3.5" (89 mm) versus 3.25" (82.5 mm) for the C89.

The Emperor will not be pleased to hear about these discrepencies.


I clearly need to do a lot of measurements tomorrow. :yesnod:

[mmoeller]

The interesting properties of the Duelund approach to synthesizing the three way filter at once are several-

• The combined transfer function sums to unity across the frequency spectrum as does a simpler two way crossover, and starts from only one equation
• The transfer function is set by the center frequency Fc and the coefficient damping factor, aleph, and can realize a range of transfer functions with wider or narrower spread of the corner frequencies for LP and HP and midrange level just by adjusting aleph
• If aleph is reduced down to the square root of 2 (~1.414), the midrange portion by that point dissappears, leaving an LR-4 two way.
• For aleph = 2 or more, the step response has no ringing- fast transient settling.
• For aleph = square root of 3, the transfer function is a three way Bessel with the HP, LP, and peak of midrange combining at - 9 dB; midrange driver efficiency and output level requirements are the least, making the smaller wide range drivers (Tangband, HiVi, etc) suitable for a midrange filler while keeping frequency requirements reasonable for woofer and tweeter with Fc of 1400-1800 Hz.
• The complete transefer function from low to high exhibits only one phase rotation through the entire three way function, with lower group delay peaking than a single LR-4 two way.
• The midrange is wired out of phase to the woofer and tweeter (a minus sign in the complete expression of the numerator for the midrange)
• All three drivers are always in matching relative phase at any frequency, so the possibility of destructive interference or off axis lobing is minimized. This results in a reported sense of integration and not being able to identify sounds as coming from one driver or another

Note that there is a greater sense of driver overlap because the graphs are posted with a scale extending to -60 dB. I will post an example later with a more typical Y axis scale.

This is for another project in consideration with a dome midrangea and small form factor dome tweeter.



Note that the drivers are -6dB at the crossover frequency, not -3 dB as in most conventional crossovers. This may be adjusted to fill in the power response; it matters whether a BBC dip is preferred (better for most commercial program material), or something closer to flat power response, which could justify elevating the midragne level a couple more dB.

This suggests a two or three way switch control- "Presense".

Sorry to throw you off track again. Mabey I should make a new thread for this question, but I'm gonna ask anyway :P .

I hope I'm asking the right questions here. I've been refreshing my knowledge of LaPlace and filter design. Duelund does a great job of combining the filter transfer functions of the 3 way design. His choice of filter profiles will limit the driver choice with its realtivly shallow rolloffs.

Would it be feasable to use these techniques to create entirely different filter functions and combine them it the same elegant form?

For example, if one wanted to increase the high pass filter order, or include an equal ripple element, would adding the appropiate zeros and poles to the transfer function work out. I assume with the proper math it would be possible.

I understand that the aleph in Duelund's transfer function controls quite a bit. I'm guessing that he used aleph to substitue for the zeros to controll the overall responce with one variable. The aleph substitution may change due to the different filter overlaps with different filters. How would one go about determining if the driver overlap was reasonable? sumed transfer function?

Would changing the filter orders effect the overall summation to 1, and result in a phase differential that Duelund seemed to conquer, or would the equations account for that with sign changes in the transfer function?

Mabey I just have to do some more math, and it might become evident.

[FrancoB]

Could an Eton 4-300 perform midrange duties in a system like this? Together with a Scan Speak 9800/6600/7100?

Thanks,

Franco

[Jed]

The Eton 4-300 has a spike in 3rd order distortion at 2.3-2.5K so I wouldn't go up to 3.5K with it, without some kind of BBC dip that is. The cone breakup is at around 7K. At around 90 DB, distortion was at 1.2%-1.5% or so if I recall- with the peak at 2.3K. I've used the 4-300 in several systems and it seems more comfortable in pairs/MTM arrangement, and crossed around 2K- 2.5K 4th order.

[Evil Twin]

Sorry to throw you off track again. Mabey I should make a new thread for this question, but I'm gonna ask anyway :P .

I hope I'm asking the right questions here. I've been refreshing my knowledge of LaPlace and filter design. Duelund does a great job of combining the filter transfer functions of the 3 way design. His choice of filter profiles will limit the driver choice with its realtivly shallow rolloffs.

Would it be feasable to use these techniques to create entirely different filter functions and combine them it the same elegant form?

For example, if one wanted to increase the high pass filter order, or include an equal ripple element, would adding the appropiate zeros and poles to the transfer function work out. I assume with the proper math it would be possible.

I understand that the aleph in Duelund's transfer function controls quite a bit. I'm guessing that he used aleph to substitue for the zeros to controll the overall responce with one variable. The aleph substitution may change due to the different filter overlaps with different filters. How would one go about determining if the driver overlap was reasonable? sumed transfer function?

Would changing the filter orders effect the overall summation to 1, and result in a phase differential that Duelund seemed to conquer, or would the equations account for that with sign changes in the transfer function?

Mabey I just have to do some more math, and it might become evident.

The Duelund starts from the most basic approach to a two way and is built up from there by direct manipulation- it always has just a single phase rotation in the all-pass, always sums to one, regardless of the value of aleph. If it's possible to derive a version with higher orders than 4th order LP/HP and 2nd order bandpass, it would have a much greater phase rotation through the all-pass transfer function.

Now, a practical implementation could probably have some adjustment of behavior once you get outside the -18 to -24 dB range of driver filtering- at this point, the impact on the total transfer summing function is fairly low. So the inclusion of notch filters or transitioning to a steep roll off rate may be reasonabl, IF the phase in the main amplitude region (from -18 dB and up) isn't disturbed significantly.

Any other comments would just be informed speculation on my part, until I have some real world experience with this. Hopefully fairly soon.

[Dennis H]

If it's possible to derive a version with higher orders than 4th order LP/HP and 2nd order bandpass, it would have a much greater phase rotation through the all-pass transfer function.

Jon, a serious question (displaying my ignorance here). What's so special about a 360 degree phase rotation like the Duelund? A linear-phase (transient perfect) system would have zero degrees of rotation. So why is 360 better than any other (bad?) number.

[augerpro]

May I also bother the Dark Lord with an unworthy question? Would driver placement be more critical with this type of XO? The reason I ask is with everything but the most "perfect" mid and a large bandpass, there is alot of sharing between drivers. For example, in the center of the bandpass the HP and LP meet only 24 dB down for standard iterations of this method. I wonder how much they contribute to the volume of the mid, ie the bandpass? Or do they just muddy the sound? Would placing the those drivers with a ctc equal to the mid center frequency help to cancel out their (poor) contribution to the overall SQ?

I just discovered this ancient Sith text on Dueland filters a last week, so forgive me if that's a silly line of reasoning...

[Evil Twin]

The midrange level and the HP/LP level are all a function of aleph, as is the spread of the HP and LP corners. Placement in the Z axis is important, as well as in the Y axis- I would use as tight a spacing as possible between tweeter and midrange, and consider using the -9 to -12 dB frequency on the LP for the CTC spacing woofer to mid.

In the posted example with aleph = 2*square root of 1.8, one might choose to use the -9 dB point for LP as the CTC frequency; this would be ~700 Hz, suggsting less than 19". If one chooses the -18 dB point, at the peak of the midrange output, that is 1 kHz, and a CTC of ~14". At this point, the woofer contributes ~ 2.5 dB of the midrange output; not very much, but still worth considering.

And yes, the woofer and tweeter may be only 18 to 24 dB down, or even less, depending on the aleph coefficient; note the plots previously posted for differetn configurations.

With larger aleph values, the area covered by the midrange increases, the HP and LP move further apart, and their crossover frequency output drops in level. But the demand on the midrange becomes considerable in this case- some drivers can handle the lower end better, some the upper end.

It's conceivable that a WMTMW configuration would provide the ultimate in output, or perhaps a WWMTM, but at what price? Is more than 106-108 dB really needed in a home system?

One could define a 106 dB/speaker system with a single 12, a 3.5" midrange, and tweeter; a 112 dB system would require dual 12's, a 6" midrange or dual 3.5" midrange, and one tweeter, perhaps with a slightly higher crossover frequency. But considering the original budget target with a single 12", I only see the midrange system being required to provide ~106-108 dB @ 1 meter per cabinet. Adequate for many rooms. Especially since this is the "classic" 20-20 capability, in truth.

These considerations suggest a first pass design with the woofer positioned relatively high in the bass module, the midrange and tweeter close together in the short triangular upper module.

Concept drawings are underway, but there have been many activities to attend to this time of year.

First, an examination of proposed transfer function plans with the attendent driver requirements for output analyzed, before woodworking and driver selection for the first pass is finalized. OTOH, cabinet drawins are already in preparation, as the first midrange module will have a removable panel and be able to test several drivers.

[mmoeller]

The Duelund starts from the most basic approach to a two way and is built up from there by direct manipulation- it always has just a single phase rotation in the all-pass, always sums to one, regardless of the value of aleph. If it's possible to derive a version with higher orders than 4th order LP/HP and 2nd order bandpass, it would have a much greater phase rotation through the all-pass transfer function.

Now, a practical implementation could probably have some adjustment of behavior once you get outside the -18 to -24 dB range of driver filtering- at this point, the impact on the total transfer summing function is fairly low. So the inclusion of notch filters or transitioning to a steep roll off rate may be reasonabl, IF the phase in the main amplitude region (from -18 dB and up) isn't disturbed significantly.

Any other comments would just be informed speculation on my part, until I have some real world experience with this. Hopefully fairly soon.

Thanks for the reply. It's sometimes hard to phrase these questions to make some sense. I think you answered my questions very well. Thanks for your time.

What effect could the greater phase rotation have? Assuming that it was consistant and accounted for, done correctly.

[Evil Twin]

What effect could the greater phase rotation have? Assuming that it was consistant and accounted for, done correctly.

That is a matter of some question- it implies greater total group delay, and past a certain point, this could be audible. To what degree is open- Sigrfried Linkwitz once thought the group delay of a an LR4 network should be avoided if possible, but his own listening tests concluded it was not an issue, at least not at lower frequencies.


On another topic, the matter of midrange selection remains. At the urging of the Emperor, some preliminary tests have been completed- not completely with the results expected, though perhaps surprises should be expected more routinely...


The published data for the C79 shows extension on axis to approximately 10 kHz before peaking and rolling off. The C89 shows mroe ragged response, possibly with multiple resonances, extending to about 8 kHz.

The other concern about the C79 is the small spacing between the front mounting plate and the manget mount, which tends to maks the driver and may be responsible for some of the lift in the 1400-2 kHz area usually seen in raw C79 plots. The C89 has a longer VC former and about 50% more rear window.

After completing some measurements, the topic of driver variabilty must be raised.

The mids were measured in 24" X 24" open baffles, at several distances, but shown here are 6" measurements of the C89 on axis and 30 degrees off axis, and a single 6" on axis measurement of the C79.


C89 on axis, 6"



C89 approximately 30 degrees off axis, 6"



C79 on axis, 6"


The dip in all measurements at ~600 Hz is largely a function of the dipole baffle dimensions. The roll off below 300 is a function of both driver and baffle; note that the C79 rolls off at a faster rate.

Neither resemble their published curves very much above 5 kHz.

Considering the rear window issue, LF extension and Xmax, and overall smoothness, the C89 will be evaluated first.

[jdybnis]

The graphs would be much clearer if you reduced the measurement window to 1.25ms and got rid of the baffle effects. And why are you using such a small baffle for measurement? With the driver mounted at the center of a half sheet of MDF, the mic 12" from the baffle, and the window set to 2.5ms you would be quasi-anechoic down to 400Hz.

[Evil Twin]

Not shown were 2" and 2" nearfield measurements, which overide the baffle effects but distort the high frequency response curve somewhat. A 4' X 8' baffle would be rather unwieldy to use and store in the available space. A 6 msec window is short enough to avoid nearby boundary reflections and still give some detail down to 200 Hz or so; some of the measurements were made also with 2 msec windows, but no advantage in boundary pickup were noted. A sealed box test would be useful next, but that's why this is just a process to select for the next evaluations.

The real point behind this is that both C79 and C89 measure quite similarly as pairs, and that the HF performance is not as extended as "factory plots" would indicate-

Practical implementation challenges with raar masking due to the size of the magnet plate and it's distance from the front plate raise concerns about the C79, while it doesn't seem to offer a clear HF advantage in response. To avoid rear masking of the C79, I believe it would require a sheet aluminum (3/16") mounting plate. The C89 may be feasible with just a carefully back beveled HDF layer up. Even there I am considering other options.

[TacoD]

Like I said in the past, factory plots of Accuton are not as acurate as Danish stuff. Can't you use the Seas Excel W12CY001?

[Mark K]

This year I'm working at home on projects instead of going to CES. Usually a CES every few years get's the job done on seeing what's currently out there.

Hopefully there will be some announcements with specific information, and I'll have to add it to the Evaluation list. C89s ordered today. Almost went for a pair of D2904/71000 (slightly more expensive version of 66000), but frugality and pending Imperial budget restrictions won out.

Probably a good thing.

The 7100 is a fine performer- in my tests the 6600 is somewhat better. I sent bothe the 7100 and 6600 to DLR and hopefully after, Zaph. I'll post after they've got a chance to look at them as well.

[Evil Twin]

Like I said in the past, factory plots of Accuton are not as acurate as Danish stuff. Can't you use the Seas Excel W12CY001?

I have seen this before with the C23 tweeter; the C13 appears much closer to factory specifications.

The W12CY001 is not without it's problems... including low overall sensitivity, frequency response which begins drooping above 2 kHz (due to cone flex, I believe, as the W22 does above 1 kHz), and a wonderfully wide band cone break up mode commencing at 5 kHz with significant output to above 15 kHz. It is most curious how Seas doesn't provide an impedance curve for this component that extends above 5 kHz- that's undoubtedly where the issues begin to arise.

On the other side, it appears that using the C89 to 2 kHz would be rather straight forward, possibly with an even simpler crossover than the NatalieP. How it sounds may be another matter.

There are many issues to research and test- that is why this is described as a design study.

[Hank]

Wow, I get buried in work, and look what happens: a new mega-project by Darth! I was at CES, but alas, in many meetings with projector companies, touting our patented Vikuiti Super Close Projection tm light engine. MANY meetings, so couldn't spend but a few hours at the main show. To show how out of touch I've been, I took a cab to the Alexis Park hotel, expecting to see the good stuff as in years past. It was then that I discovered it was moved to the Venetian (I thought that venue was mass-market 'high-end' stuff). I had a good two hours, so only flew through it. It seems that many of the smaller companies that used to be at the Alexis Park were missing. Rent too high at the Venetian? I stopped at the Parts Express booth and picked up a show special flyer (GOOD prices) and this month's audio Xpress. In it there's a Madisound ad for a new SEAS driver, the 16cm mangesium cone W16NX-001. Probably overkill for this project, as it goes to 50Hz, and probably at a dear price.
I was going to ask why the Woofer Tester disappeared from the PE catalog, but I found it in it's own 'booth'. The company is on its own and has a couple of step-up models. VERY interesting and tempting: www.woofertester.com

BTW, I have a pair of the Tang Band W4-1337s titaniums. Mark is going to bring his test gear in a few weeks to measure drivers and these will be tested. Maybe TB in a system with high-end woof and tweet would be embarrasing, but imagine if the little bugger worked! You could always tell your friends it's a new custom driver designed on this forum :twisted:

[JonMarsh]

Tang Band is a company that could be pretty scary if they really got their act together. They make a lot of stuff for other folks (can you spell Adire, Ascendant?), I'm also curious how the 25-1414SC tweeter plays out, for the small form factor applications.





Those glitches in the impedance curve would naturally worry me, could be just surround resonances. The FR plot DOES look pretty amazing, especially if it really hits 87 dB/watt with a DCR of 6.8 ohms (which would make it a fairly "real" 8 ohm driver). I wonder of the response plot is smoothed overall? That hump centered at 15 kHz looks too clean to be a "normal" metal cone breakup.

With 3 mm it ought to be usable above 250-300 Hz. And paint it black to disguise it, huh? ET would like that, I bet.

OK, you talked me into it, I'll try to get a leg up on this, beat ET to the punch.

Order placed....

Sorry you didn't get hardly any free time at CES. I think the Venetian thing caught a lot of folks by surprise, though I'd informed a few that were going.

Looks like Smith and Larson are doing some interesting things... just what I don't need, more things to spend money on!

[cjd]

Yeah, T-B does some interesting things. I eye them now and again, and then my eye wanders...

C

[Jim Holtz]

Tang Band is a company that could be pretty scary if they really got their act together. They make a lot of stuff for other folks (can you spell Adire, Ascendant?), I'm also curious how the 25-1414SC tweeter plays out, for the small form factor applications.





Those glitches in the impedance curve would naturally worry me, could be just surround resonances. The FR plot DOES look pretty amazing, especially if it really hits 87 dB/watt with a DCR of 6.8 ohms (which would make it a fairly "real" 8 ohm driver). I wonder of the response plot is smoothed overall? That hump centered at 15 kHz looks too clean to be a "normal" metal cone breakup.

With 3 mm it ought to be usable above 250-300 Hz. And paint it black to disguise it, huh? ET would like that, I bet.

OK, you talked me into it, I'll try to get a leg up on this, beat ET to the punch.

Order placed....

Sorry you didn't get hardly any free time at CES. I think the Venetian thing caught a lot of folks by surprise, though I'd informed a few that were going.

Looks like Smith and Larson are doing some interesting things... just what I don't need, more things to spend money on!

Hi Jon,

I'm also very interested in the TB W4-1337S and have been doing a little research on it. I don't remember where I found these measurements but they're not quite as nice as the published ones. I'd really enjoy hearing your thoughts after looking at them.

Thanks!

Jim

[dawaro]

Hi Jon,

I'm also very interested in the TB W4-1337S and have been doing a little research on it. I don't remember where I found these measurements but they're not quite as nice as the published ones. I'd really enjoy hearing your thoughts after looking at them.

Thanks!

Jim

If I remember right Darren popped in on the PE board one day and posted some.

[Jim Holtz]

If I remember right Darren popped in on the PE board one day and posted some.

Hi David,

Yes, I think you're correct. Darren's measurements aren't smoothed nearly as much as the one on the TB website but it's not too far off. They should work well as a mid range I'd think. It looks like they could be easily pushed above 3K which would open up lots of tweeter possibilities. :T

If they sound as good as the Visiton Ti100's John Pastuck brought to the Iowa DIY event, they're a killer mid. The drivers look like twins with perhaps a better motor design for the W4-1337S.

I know the distortion measurements from KT weren't that great on the Ti100's but the SQ was awesome. The mid range had a mesmerizing holographic sound to it. The Accuton 3-way that followed it was similar in the midrange. That is a huge complement because the Accuton design was simply superb sounding. It was easily one of the best sounding point source speakers I've heard.

Thanks for jogging my memory!

Jim

[jdybnis]

some of the measurements were made also with 2 msec windows, but no advantage in boundary pickup were noted.

Calculation of path length difference between direct sound and baffle edge reflections indicates that the window has to be < 1.3ms to avoid baffle effects (3" driver centered on a 2'x2' baffle with mic at 6").

[JonMarsh]

I think Jim's comments could be interesting, though I wonder if that's Darren's measurements if that's relative to a 1W drive level. That would be a lot lower sensitivity, which sometimes is the case for the far eastern "full range" drivers. But I wouldn't think so for one with a relatively low Xmax like this.

I agree with Jim that it might be pretty useful, though I would hope it isn't naturally rolling off below 500 Hz as appears to be the case in this plot.

[dawaro]

I think Jim's comments could be interesting, though I wonder if that's Darren's measurements if that's relative to a 1W drive level. That would be a lot lower sensitivity, which sometimes is the case for the far eastern "full range" drivers. But I wouldn't think so for one with a relatively low Xmax like this.

I agree with Jim that it might be pretty useful, though I would hope it isn't naturally rolling off below 500 Hz as appears to be the case in this plot.

IIRC Darren's only major concern was with the cone of the driver. Apparently it is paper thin.
What I would really like to see measured is their new dome mid. It has either a very nice response or a very smoothed one. If someone wants to test it let me know and I will see if I can get my hands on one.

[Jim Holtz]

I think Jim's comments could be interesting, though I wonder if that's Darren's measurements if that's relative to a 1W drive level. That would be a lot lower sensitivity, which sometimes is the case for the far eastern "full range" drivers. But I wouldn't think so for one with a relatively low Xmax like this.

I agree with Jim that it might be pretty useful, though I would hope it isn't naturally rolling off below 500 Hz as appears to be the case in this plot.

Hi Jon,

Actually Darren had forwarded the email and measurements to Curt who posted it on the PE board. Here is Darrens message that should help clarify things a bit.

(paste)
I finally got around to running a couple curves on that TB Titanium driver
that I have in my basement. Unfortunately, I don't actually have any hosting
service right now, so thought maybe you could upload these somewhere and put
a post on the Tech Talk board referring people.

The graphs aren't really good below 500 Hz or so because of my
hacked-together setup.

Overall, it looks pretty good I think. In the waterfall, some of those
little zigs above 3k are indeed a ringing artifact of some sort, and not
just ripple in the measurement. But, I suspect that they are IM distortion
products from that huge peak, and with the proper filtering wouldn't be an
issue. The inductance is also very low, and as John pointed out those
impedance zigs looked bigger in the TB graphs because of the scale. Within
normal mid/woof working range, these look pretty flat, and it appears that
these could be a very nice mid or small woofer.

I'll also point out that the titanium is VERY thin, so anyone who buys one
should be careful when touching the cone.

(Feel free and post some or all or a modified version of my comments if you
want.)

Have a good one!

Darren

A couple other comments from posts:

Even in my hacked-up setup, it was essentially infinite baffle. I think the baffle was 4' by 6' roughly. I had a bunch of junk layin' around it, so had to cut my time window a little short.

And the driver was not properly flush-mounted, which I usually used to do on any true "mids".

I would say that the dip around 1k is real, though it may not be quite as deep when properly set up. Even so, I don't think it is anything too severe in the grand scheme of things.

Also, I forgot to mention to Curt that the drive level was arbitrary, so don't take into consideration the overall sensitivity.

(end of paste)

I hope that adds a little clarity to the mesurements.

Jim