Zaph|Audio

The model number of this tweeter is JA-0516A, and is likely pulled from a NS500 system. There is a sticker on the back saying it was a genuine Yamaha replacement tweeter. Age is hard to know for sure, but somewhere between 15-20 years, and according to the owner it has been used for 6 of those years without issue. A much more common version is the JA-513 but it's not clear how they differ, if at all.

The response is nice and smooth on the low end, with a mild breakup rise at 10kHz and a rolloff after 15kHz. The harmonic distortion is nothing special, but usable down to 2kHz LR4 or maybe 2.7kHz LR2. There is no ferrofluid in the tweeter and as a result there is a very steep impedance peak at it's resonance of 770 Hz. I'd be willing to bet that the original crossover did not use a LCR trap to flatten the impedance, and as a result the tweeter could be easier to blow when a loud 770 Hz tone comes through, not to mention there would be a big peak there after the filter is in place.

Build quality seems good. The tweeter is large and heavy. It was properly countersunk for testing, which was a pain considering it's flange shape. The pure beryllium dome sure isn't helping the tweeter to have an extended and flat top end. Overall the performance is not bad, but it's generally outperformed by any decent tweeter from this millennium regardless of diaphragm material.

Usher 9980 Beryllium

Performs just like the current Dayton RS28, complete with dip in the top octave. Even the breakup is at the same frequency. Looks exactly the same except for a mild gray-purple tint to the dome. The distortion profile is a teeny bit cleaner, but it's close enough that it could just be batch differences. Hard to say. This Usher isn't available to DIY'ers and thus the listing on this page. June 2009 update: there's controversy about what constitutes "real" beryllium. You don't have to look too far to find it. Regardless, to date I've seen no real proof that any beryllium usage, either pure or heavily alloyed with other materials, is any better than aluminum or coated fabric. There's certainly a lot more to a tweeter than the chosen diaphragm material.

Vifa BC18SG49-08

As of this writing, 2 versions of this buyout woofer are on sale at Madisound for $13 each. Round and truncated frame. Not for long, they will be sold out soon. Similar to the older poly frame TC line, which was a bit overpriced around $30-$40 each. At $13 it's a very good deal and reasonable performer. Would work well in a sealed box.

Mark & Daniel DM4 results...

Lets start with the good: non-linear distortion is decent and these tweeters do indeed work very low. A steep 900 or 1000 Hz crossover would be fine.

Now for the bad news: everthing else about these tweeters was horrible. The response curve borders on unusable and the build quality is low. One tweeter arrived with a cracked faceplate, even though they seemed well packed. The damage was cosmetic only, however when the faceplate came off, a bunch of paint chips came with it, some of which got into the pleats. I was unsure of how that might affect the testing, so I only tested the other one. The other sample seemed fine. I inspected the pleats and there was no damage. The faceplate was loose however so I securely taped it down to prevent rattling before testing. The faceplate is just a 1/16" thick painted piece of plastic glued onto the stamped steel front. It looked like a cheap afterthought. The tweeter was tested on a proper countersunk baffle, open back. I did not test with an enclosure on the rear of the baffle, but of course you'll need one in a box system to prevent woofer pressure from affecting the tweeter membrane. I would expect that experimentation and lot of measurements would be needed to see what the optimal enclosure size would be.

Honkey low ends are a trademark of most AMT based drivers. I've seen response curves of a few, like the Eton version for example. I could hear the 500 Hz resonance of this tweeter just by lightly blowing on the pleats. My guess is that the tweeter filter's transfer function had better be at least 30 or 40 dB down at 500Hz. If this isn't done, there will be audible issues right in the vocal range. Simple first or second order filters won't work well with this driver.

Could this tweeter sound good in a system? Yes, if you could manage to hammer the response curve into a workable shape. I see it as not worth the effort however, unless you happen to have a DEQX box around to do the hammering for you in the active digital domain. Or at least, an HT receiver with DSP EQ that does not suck. But even with active EQ, the low build quality could still be an issue.

B&C DE10 horn driver w/ME10 horn...

Some of these tweeter tests are different than what is normally seen here. The response curve is smoooth enough to be controllable with a gradual rise based on the horn loading. What makes it less controllable is the twin peaks in the impedance curve. Thanks to these twin peaks, a simple single cap filter is not usable. Your options for avoiding the impedance issues are to cross over very high, use multiple impedance notch circuits, or go fully active. To be honest, I don't like any of these options.

This horn is labeled as having a 90 x 60 degree pattern. Since the horn has a square flange, it's not exactly clear which way the horn should be oriented. There are some line markings but no documentation. I did some google research, and found a bunch of people that suggested the lines on the horn be mounted vertically. They were all wrong, at least if you're shooting for the widest horizontal coverage. That happens when the markings are placed horizontally.

Harmonic distortion is excellent at the test level. Most traditional dome tweeters scale their noise floor with the fundamental in a fairly predictable way across their bandwidth, until excursion becomes an issue. With this DE10+ME10 combo, I noticed a fast increase in 3rd order distortion at 2kHz that was not consistent with the rest of the noise floor increase. For most of the bandwidth, an output level increase has close to a similar noise floor increase. At 2kHz however, every output level increase cranked up the 3rd order harmonic distortion by a factor of 3 or more. What does this mean? If you plan on using this combo at high levels, the crossover should be way down by the time you get to 2kHz. A crossover point of 2.5 kHz with a LR4 slope should allow sufficient high level usage. In a lower output situation such as near field, this combo could be crossed over quite a bit lower.

There was a minor consistency issue between the tweeters. Through some experimentation, I saw that it was inherent to the tweeters themselves, and not the fitup with the horn lens as might be expected. One has a minor issue going on at 2.5 kHz, which was somewhat more visible in the impedance curve with the horn lens in place. It's still a minor blip without the lens however. If I were giving star ratings, there would be one off for that.

Very few horn tweeters come through Zaph Audio testing, so I can't say if this is normal horn behaviour. Do not assume that all horn tweeters perform like this. I'll get my hands on some other horn drivers over time.

Seas H1333...
Also known as the T18RE/XFCTV2 7" coaxial.

It's certainly usable as a 2-way coaxial, with a little on-axis tweeter raggedness being it's only downfall. I'd rate build quality slightly low for a Seas, but this driver combo looses a point because I don't think enough attention was given between the tweeter and woofer fit-up. The coil former is longer than it needs to be, and the edge of the tweeter is not level with the edge of the cone. This is likely somewhat responsible for the ragged tweeter response.

Some multi-tone spectrums were done with very low frequency running on the woofer, with the tweeter playing a different frequency independently. The 1kHz was the worst case, showing a little grunge around the fundamental. I actually swept a tone through the entire tweeter range while cranking 30Hz out of the woofer and watching it all on the spectrum. The tweeter was relatively clean for a small neo and maybe that helps the situation. I'd say, at least for this driver, amplitude modulation doesn't seem to be an issue, though I'm not convinced this multitone measurement is even the proper way to measure that.

Anyone wishing to use this should pay particular attention to both driver's off axis response curves. The tweeter response cleans up off axis, and will need to be crossed over below the point where the woofer begins to beam.

The real issue with the driver is value. There is absolutely no way it's worth it's current cost nearly $160. (late 2008) For those considering a center channel, at that price you can build a 4-driver 3-way ZDT3C and it will outperform it in every way. The coaxial isn't a bad driver, but the cost kills it. I am a fan of many Seas products, but not this one. There could be some specific installation requirements that still make this driver a good choice regardless of cost.

Tall enclosures...
I've often told people who want to use tall enclosures that there needs to be a lot of extra damping at the bottom. See the vented option in the ZD5 project for example. Sometimes I wonder if people think I'm just pulling advice out of my ass. Rest assured however, that almost everything I do has a reason. (Heheh, I stuck "almost" in that statement because I'm human)

The following impedance curves are from a single woofer in a vented 38" tall tower. We have two different amounts of damping at the bottom. All other surfaces are well braced and lined with Whispermat.

A little clump wasn't enough to kill the lengthwise pressure node at 140hz, it took nearly a pound of Acousta-stuff densely packed into the bottom area. (and even then, it's not completely gone, but good enough for me) The undamped pressure node is visible as a peak in the impedance curve and also as a dip/peak combo in the frequency response. Audibly, I think it just sounds like a little lack of definition or "mud" in the midbass. Properly damped, it's clean and tight. Consider this verification of something I've always recommended. Single volume tall enclosures work fine, but you better pay attention to that lengthwise node. As the enclosure gets longer, the more damping you need.

More waveguide experiments...

I'm still having the urge to do a BIG waveguide 2-way someday and so these tests continued. The hope was that without the grill and phase lens in drivers like the 27TBFCG, the response would be smoother. Indeed is was, but only off axis. In fact it looked nearly perfect off axis. On axis remained a bit ragged. Some directivity-minded folks might even suggest that the off-axis is more important than the on axis but I think they are both equally important. If I were to do a design with this combo, I would likely add a mild notch at 8.5kHz to smooth the on-axis a bit. I would not mind the trade-off of a mild null at 8.5kHz off axis. The combo is looking good, but I've got a few more modifications up my sleeve.

Tymphany LAT-250 test results...

Results are in and they are kinda strange. Not surprising, since the driver itself is kinda strange. Fs was kinda high, and not very pronounced, and this was after a substantial break-in and cool down period. It's way higher than the spec sheet says. This "subwoofer" doesn't have much low end extension at all. Note that I could not get accurate T/S numbers on this driver. With an almost non-existant Fs impedance peak, I could not get a good delta compliance Vas number. There's no way to do a delta mass on this driver either. Anyone trying to use this driver will likely need to just put it in some experimental boxes and see what works.

With the lack of low end extension, I don't really see any effective use of this driver without active response shaping. Distortion is good above 90 Hz and excellent at 150 Hz, but slightly high lower in frequency where cleanliness is really needed. The bottom line is that this system has seriously limited usefulness - it's your only option if you absolutely must have a woofer that is 3" wide. This is essentially ten 2-1/2" woofers, but performance wise, it will be drastically outperformed in almost every way by just about any 8" woofer.

The test results of this small version may not represent the usefulness and effectiveness of the larger versions, but I do have my expectations. I think the real problem with the LAT woofers is that they still have traditional Thiele/Small parameters. A certain volume displacement is going to require a certain cabinet volume. That renders any slim profile almost useless if it has to go in a big box anyway. The LAT has some interesting engineering and features but Hoffman's Iron Law remains in full effect.

$13 MTM Buyout Plate from PE...

Vifa XT15 test results...

Peerless 2" driver...

The testing level completely overwhelmed this speaker on the low end, and the harmonic distortion plot reflects this. Aside from that the harmonic distortion is not bad but only average. I honestly expected better given the copper cap that's supposed to be on the pole piece.

Note: I had some problems getting a reliable Vas number for this speaker. It's left blank at 0.0 in the T/S parameters as are the other specs that come out of that portion of the test. With delta mass, it's hard to find the right mass to have a good delta rather than too much or too little which throws off accuracy. Same thing with the delta compliance method.

AR bargain bin speaker...
This discount pair of Acoustic Research speakers has been tested. Model ARXP52. They are available for $99 at various web sites.

Overall not too bad and I'd say it was a step up from the Insignia, at least in midrange smoothness. The difference between the Insignia's 800 Hz peak followed by sharp dip and the AR's smoothness in the same area is clearly audible. Performs better with the grill off but looks better with it on. Definitely worth the money.

As far as improvements, I don't see an easy way to get to the inside. It appears everything was glued together and there are no visible screws. The front panel has a formed foam cover which may hide screws underneath, but the foam is glued on and likely to be destroyed if torn off.

This system doesn't appear easily modified via and external notch circuit as there are no major peaks to take care of, only dips. Dips are less audible than peaks and there are no major issues in the harmonic distortion, so I'd say modifications may not be worth it anyway and the speaker can be enjoyed as is. Here's the measurements:

The drivers seem to be of average low end construction. The tweeter almost looks like a Vifa DX19, with a wide surround for a 3/4" dome. I can see the magnet through the port and it's not a Vifa, but something out of Taiwan. It does seem to have the rising 20+kHz response just like the ring radiator though. The woofer has an aluminum cone and has a curved profile. Judging from the harmonic distortion curves, there is a breakup somewhere between 4 and 5kHz and it doesn't seem too sharp. Other information on the drivers is limited since as mentioned above, I was not able to get inside to take additional measurements on the individual drivers or to inspect the crossover topology. It's difficult to say how much of the dip at 3kHz is due to diffraction, or driver response issues or crossover integration. It's a fairly narrow band dip and is not real noticable since the rest of the speaker is smooth.

2007 Seas W15CY001...
In my 5.5" woofer comparison, a different earlier version of the Seas W15CY001 was tested. It was similar, but had a lower frequency breakup that was also lower in amplitude. Still, it was a great driver and I believe one of the best small woofers on the market. I've used it in a system but never posted the design because of the changing breakup node over the years. (not to mention the T25CF001 is discontinued) That may change with the latest test results:

Hemptone driver tests...
These Hemptone speakers from A Brown Soun have been tested. The first is a 5" full range, reportedly a drop in replacement for some Fostex drivers. The 2nd is a 7" woofer. These sell for $85 and $109 each, respectively. I'll let you all take a look at the measurements and draw your own conclusions.

Hi-Vi S1 Coaxial...
I finished testing today. It certainly is a unique driver. (picture below on 8/4) It is basically a radial planar tweeter mounted inside a large ring radiator mid. The tweeter's effective radiating diameter around is about 2.5cm though it is about 3.5cm diameter including the phase plug. The phase plug appears to fasten down a light mesh grill in addition to the copper/film planar coil under a spacer. The mid apears to be typical ring radiator construction. In other words, all surround. OD is 8.5cm, ID is 4.6cm. In determining the radiating surface, A good estimate might be 20.5 sq cm using the center of the surrounds, inside and outside. In other words, roughly 4 times the area of the average 1" dome tweeter or just slightly less than an average 2" dome mid. Connections are spring loaded terminals in back. For testing, the driver was countersunk up to the metal trim ring, with the rounded over part being above the baffle. That's the way it seemed to be designed for mounting.

The mid on the other hand performs a lot better, with smoother response and decent harmonic distortion. You can see the typical ring radiator distortion profile, with rising 2nd order HD on the low end. Other than that, you can use these as low as your required output level allows. I think for most people, that's going to be around 800Hz LR4. The mid will definitely require an impedance flattening circuit. Also, given it's drooping top end, when it is shaped back down to flat the sensitivity will be closer to 85dB/1m. The good news is that you probably can cross the mid over high enough to avoid the tweeter's issues. It's going to require some inventive crossover topology, but it's certainly doable. The ideal usage of this system would be between two woofers, making it essentially a 3-way MTM format. Hi-Vi's D6.8 woofers would seem to good match, with their low sensitivity and decent performance. It would probably make a good sideways center channel too, since horizontal lobing could be minimized. Overall I would have hoped for a better tweeter inside the ring mid, but it's usable as is.

Usefulness of the CSD...
As I've said several times before, the cumulative spectrum decay (CSD) is a highly overrated form of measurement and is merely a different way of looking at the frequency response. It's all generated from the same impulse and it's all linear distortion. Yet, the CSD is often misread. To make a point, I'd like to present two CSD's done under the same conditions - one for a poly cone driver with a well damped breakup, and one for a metal cone driver having a harsh breakup 6.5kHz.

Aside from harmonic distortion which is a form of non-linear distortion, it's all about how easy the driver's response curve is fixed in the crossover. Don't overestimate the difficulty of metal cones - this one is easily controlled with only 3 components. Actually, both response curves for these drivers are very good and easily controlled in the crossover. Below we have CSD plots of each driver with a filter in place, giving us well shaped LR4 rolloffs at around 2kHz. They are not exactly the same, but they are close enough to make a point. The metal cone's breakup has been dealt with, and the poly cone's shelf between 1 and 2kHz has been smoothed out. The response curves, and thus the CSD plots, now look very similar.

I hope this helps solve some of the misunderstanding centered around the subject. I'm not sure if I can think of any other ways to further clarify this.

Le(X)...
This section has been moved to it's own page and updated with more information and more curves.

Front port mounting...
I have to let you guys in on a little something I found out. In a recent design a front mounted port. Gasp, shock, I heard that was bad. Front mounted ports are not nearly as bad as some might have you think. Yes, some midrange can sneak out the front, but it's often down in level enough to not be noticed. But what I'm going to show you here is that how much midrange escapes the port is dependent on the port's length and implementation. Below is a couple of response curves, same enclosure, same front mounted port, same measurement technique, different tunings and port lengths.

Aside from the port length, the primary difference was that the longer port stuck through shelf brace in the Parts Express box, effectively denying the woofer a direct path to the inside opening. Something to think about if you ever want to use a front mounted port. Lucky for me, I prefer the lower, leaner tuning in my listening room.

Using trending in HD sweeps...
I briefly mention using trending in the documention for many of my harmonic distortion tests to get an idea of the very tall order harmonics generated. Looking at 2nd and 3rd order distortion is great information. But looking at 4th and 5th in relation will give you a bigger picture of the non-linear distortion performance of a driver. Here's a little more information on it. For this particular example in the first three images below, I'll use the recently tested Aura NT1 tweeter. The 4th image is a different defective driver.

I've included a trending image of a defective driver with a rub and buzz problem. It points out the following fact: Even if the 2nd and 3rd order harmonics are relatively low, if the 4th and 5th are both higher, there is going to be a serious distortion problem at that frequency. That's probably why Germany's Hobby HiFi HD sweeps include the 5th order harmonic - to give readers an idea of what else is going on in the tall order stuff.

There are ways to interpet this depending on certain circumstances. I'd like to suggest a few examples from my past measurements. First, lets have a look at the Seas 27TBFCG and focus on 2300 Hz. The 3rd order is the same as the 5th order. Is this a problem? No, because the level where they intersect is very low and therefore the upper harmonics are also very low. On the other hand, looking at the Tangband 25-1166 at 2300 Hz, the 3rd order intersects the 5th order but does so at a significantly higher level. This points to quite a bit of odd order tall spectrum noise going on at that frequency. Now, you also have to consider the bandwidth of these interections. A few problems at select frequencies is not going to be nearly as noticable as poor tending across a wide range. This unnamed midwoofer has trending issues accross the spectrum. The resulting tall order junk is going to be audible just about all the time in addition to having IMD problems.

Now the real difficult part of all this (and all forms of testing) is deciding what's audible, what's acceptable and what's troublesome. Unfortunately some experience is required for that, and worse yet it's a subjective opinion that may be different for various people. Cleaner is always better of course. But when a driver strays from clean, how and where it does it could be percieved as more acceptable or less acceptable by different people. I hope you guys are reading between the lines with what I'm saying here. There is undeniable truth in distortion measurements. But you still need to figure out for yourself how to use them.

Vifa D26NC55 mounting alternatives...
Ok, so the D26NC55 is a great tweeter. But when a DIY'er sees the flange, it's easy to change your mind about the tweeter. It's a weird shape that is a complete pain in the ass to countersink. I'd like to show you something else that works well. Basically, it amounts to a half inch waveguide.

The resulting response curve gives you a 3dB rise centered at 3.5kHz. Might not seem to useful at first glance, but what you'll find is that once it's on a normal sized baffle, it's quite easy to work into a usable LR4 slope with minimal component count. I can guarentee that if I ever publish a design using the D26NC55, it will use this mounting format.

Tweeter IMD plots...
I don't normally do IMD plots because I find them to be less informative than harmonic distortion sweeps. However, I did pick some of the best performers along with some of the most expensive tweeters and did a few at 1kHz and 2kHz. There are no real surprises here but this might be informative to someone who simply wants to relate my normal HD sweeps to IMD plots.

Inside HD sweeps...

The THD reading of .68% is correct. It is audible, as I suggested in the previous blog entry about the importance of non-linear distortion. I can sit in front of the testing baffle and put my head where the mic normally goes, and hear distinct changes in tonality as the sweep travels from 1kHz to past 2kHz, passing the 3rd order and 5th order HD peaks on the way. Of course it's a bit harder to notice with music, but it is noticed over time when the right music comes along. For anyone that doubts that THD under 1% is audible, you owe it to yourself to start doing some testing and listening of your own.

When I'm running these sweeps, I typically sit here and observe what's going on in the tall order harmonics F6 and above that aren't tracked. I pay particular attention closer to the low end of a drivers range. If something looks out of place, I take note of it. In almost all cases however, the upper harmonics are right where I would expect them to be based on trending of the lower harmonics.

Review: Tangband WT-1427B...
The Tangband shallow mount neodymium subwoofer - here's a few words on it.

First off, I found it very difficult to test using a similar method as the 2 subs discussed earlier on this page. The subs previously tested had enough weight that placing them on the concrete blocks using gravity was sufficient. But this driver is so light, the same test setup caused it to hop around madly. Finally, I had manually put a ton of force down on it to keep it still while the sweeps ran. Unfortunately, a person that close to the mic makes the results less trustworthy, so I'm not going to post them here. I'm pretty confident sound waves bounce off my shiney hairless forehead and do something bad.

As most people would probably guess, low distortion is not this driver's specialty. In a quick comparison with an RS270 I had laying around, harmonic distortion was an average of 10-20 dB higher thoughout this woofer's usefull range. (note that the RS270 measures really well in the bass dept) Additionally, the WT-1427B is not the quietest woofer around, with some wind noise in the motor and a mild clicking at medium and high Xmax. The response curve and T/S parameters were very close to TB spec. That in itself is a little surprising, as many TB speakers have been way off.

The T/S parameters seem to fit the woofer's design well. A 14 liter sealed gives a very shallow low Qtc rolloff of about a 60 Hz F3. Sealed will likely need some EQ low end assistance. A vented 20 liter box tuned to 30 Hz gives a satisfying low end with an F3 of about 35 Hz. Venting a box that small is not easy however, and the only solution will be a 2" flared port positioned down the long end of the enclosure. A non-flared port will certainly huff and puff at higher levels. Specifically, the large flared 2" "Precision Port" is the only port I'd recommend as it's flares are big enough to reduce turbulance for that diameter and it's 11" length (12" to flares) should be just about right. Otherwise, a larger and longer rectangular port can be designed into the enclosure.

Overall, the WT-1427B is a woofer that could work well enough if you have depth and space limitations, but it's not a woofer to choose if you have plenty of space for a normal woofer. Potential uses are slim subs for behind furniture, side mounted subs in super slim floor standing enclosures, or in-wall mounting. For in-wall mounting, you had better build a reinforced enclosure however, because drywall will surely rattle apart.

Sub test: Dayton vs Skaaning...

Both drivers are 12" and 4 ohms but the similarities end there. These are two vastly different drivers. Here are some differences:

The comparison starts with harmonic distortion sweeps and continues with 3-tone IMD plots at 50, 60, 90 and 150Hz. I'm not specifying a dB level except to say that at 50Hz, Xmax was roughly 6mm. The tests are done free-air with the mic at a nearfield position of 4" from the cone. The woofers are supported on the magnet side by 140 lbs of concrete blocks, arranged to not block the magnet venting. Note that unbaffled and nearfield are not the best ways to test drivers, but it's not feasible to build an infinite baffle of the strength and size required for 12" woofers, nor is any home environment accurate enough to give a realistic far field response curve in the bottom 2 octaves. So, there are major defects in this test, but the best we can say is that both woofers are tested the same way for comparison purposes.

What does it all mean? Well, in a complete shocker, the $700 woofer outperforms the $130 woofer. Not by much, however, and the lead disappears above 90 Hz. Both are excellent performers, and we're basically comparing goodness at a price that's a steal with greatness at a price that's outrageous. Note that there's enough difference between these woofers that these results are not an argument in favor of my Woofers with lower Xmax sound better rant, but I'm sure some aspects of that argument are true.

The Dayton RSS315HF subwoofer remains my favorite sub that I use very often in my home theater. I don't think I'm about to toss my Skaaning SK300 on ebay however. It would make a great music-only sub in my listening room, where I don't have to reproduce explosions and sound effects.

Rub and buzz...
It's been suggested by a few that we need multi-tone IMD plots to see rub and buzz problems in drivers. I'm here to set the record straight and tell you that rub and buzz, or other issues that make a driver defective, are clearly visible in swept harmonic distortion plots.

When running a sweep on the tweeter above, I don't need to see a measurement to know that something was wrong. The tweeter practically screamed at me. The F6+ tall order harmonic products are very audible. Still, even if I was not able to hear the test, trending in the low order harmonic products still points to a problem. This is one of the reasons I don't publish multi-tone IMD plots. Issues are clearly visible without seeing the whole F6+ spectrum.

Parts Express cube speakers...
I tested some of these super-cheap mini speakers from PE. I'm not sure why, except that I needed to boost my last order over $100 to get free shipping, so I said what the hell. They are $26.

The insides of this are kind of surprising. At this price point, I would never expect to see a 6 component crossover. There is actual damping material and a relatively rigid enclosure is a bonus at this price point. The driver is a poly cone with rubber surround.

All said, if someone absolutely has to have SMALL, this is a high value and worthy alternative to the crappy performing and overpriced Bose cubes.

I would not consider dropping a pair of B3's in this speaker a worthwhile upgrade. The main Bose compromise is still there - a too-small enclosure. If crap (of a small size) is what you want, don't waste your time trying to improve it, because you can't.

Three 15cm Surface Mount Midwoofers...
Consider this the battle of a soft-cone surface mounts. Popular with the router impaired, these frames are not designed to be countersunk. I do however still recommend rounding over the inside of the woofer hole for airflow, particularly for the Hi-Vi W5 with it's massive magnet.

These drivers have a lot in common. They are all highly damped cones with curved profiles. All have very smooth response through the midrange. As far as distortion levels through the midrange, there are no standout performers - all three are reasonably good, but probably not up to high end driver standards. None of these are bad drivers and any one could suit you depending on your needs and preferences. But if I had to choose a winner, I think it would be the Hi-Vi W5. I look at the TC14 and M130 and think to myself: "I'd pay 10 extra bucks to get a metal frame". Well, the Hi-Vi answers that wish and costs the extra bucks too. Overall build quality of the W5 is the best of the group. Of these three drivers, the Hi-Vi is the one best suited for 2nd order crossovers with it's smooth top end response.

The Hi-Vi is a Chinese made driver. It looks like Asian-made drivers have come a long way. The W5 wins points for specs that are very close to advertised. It's website will make you laugh a little with what looks like the poorest Chinese to English translation I've seen. Apparently, the "PPF film" cone is made in Germany, and the coil former is Kapton.

Note: response curves are far-field merged with near field for accuracy. Be aware however that a near field method was used for the harmonic distortion. This means that there will be less room interaction but they are only accurate up to about 1000 Hz, with accuracy slowly falling off above that.

Wide Range Minispeakers...
A fun little comparison of tiny drivers was the first one posted on this web site years ago. Results are generally accurate, however there were some limitations with the harmonic distortion plots done at the time. First, this old version of SE only plotted down to -55dB, showing only the peaks and higher bass distortion. Second, the plots were done nearfield, limiting the top end accuracy. Based on the driver size, the top end HD plots will be good up to about 2.5kHz, slowly losing accuracy above that.

Yours Truly, John "Zaph" Krutke © 2007
Back to -Zaph|Audio-