CatZ Tweeter Shoot Out to come...

[JonMarsh]

OK, this has gotten a little bit out of hand, but it looks like I'll be evaluating a potful of tweeters for various price points in this project... I suppose later, based on the test results, I could even put up a poll for people to pick their "no holds barred" and "bang for the buck" favorites.


Here's how the contenders are looking, and what I should have on hand... (let's keep in mind I DO have limited resources, guys...)


ScanSpeak Revelator 9900

Proposed as an old fashion favorite, and the favorite of some regardless of price, this classic Scanspeak tweeter has a shallow waveguide built in to the faceplate, which must be planned for in the crossover implementation. It does have more linear Xmax that many of the newer ScanSpeak tweeters, too!












ScanSpeak 7100-02

This newer ScanSpeak model got my curiosity raised because it seems to combine a front plate similar to the 6600 and 6640 with the motor and dome of the 6620. I think the short waveguide in the 6620 is a bit of a problem, personally. The 7100 is an upgrade candidate for the original NatalieP.














ScanSpeak 6620

This is in the group more as a reference, and because I have a pair left over from the original Ardent project. Published data looks good, but I wasn't enamored of the sound, and had some specific issues with measurements. Contrast and compare....












Now were going to move down market price wise....


Transducer Labs N26A

This is one I've measured and bought several pair of, even one set in an unfinished dipole with waveguide design. I think it's an excellent motor and fairly compelling value proposition at US prices. 0.5mm one way Xmax, and a very strong motor.









ScanSpeak Discovery D2608/9130

Let's not forget this was original a Vifa designed and built part- the first pair I bought came that way, and went into the Modula MT XE. So far, I've only used them in waveguide applications, where they work very nicely, but they also have 0.5mm Xmax and fairly solid performance, though rolled off in the low end below 2 kHz- but not due to Xmax limitations. This tweeter does have a minor cult following, and for good reason.












Now, we have a dark horse contender, with a strong recommendation from Mogens...

SB Acoustics SB26STAC

In some ways, this is what I call a typical 4 ohm wideband tweeter with low Fs; it could be dropped into a NatalieP quite easily, for example, or other designs where the RS28 or Vifa XT25 are used, as well as quite a few SS tweeters. Notable specifications are the 0.6mm one way Xmax, the use of a fairly high energy saturated flux gap design, a copper cap on the pole piece for reduced inductivity modulation and lower distortion, bringing a very flat impedance curve in the operating range. Published off axis response looks quite nice, and the claims of a low reflection rear chamber seem to be backed up by the impedance curve behavior. Oh, and probably the best specification of all is the price: under $50, probably even with shipping.


This should be an interesting testing round up- I expect to complete it by Thanksgiving weekend.

[Hdale85]

I bet this list grows a lot...lol

[5th element]

The only one I'd consider would be the metal domed SB. And only because in these measurements it has a nice flat and extended frequency response. It's breakup is almost non existent and it has lower than average HD2.

http://www.audiogurman.com/1.htm#sb26adc

[JonMarsh]

The only one I'd consider would be the metal domed SB. And only because in these measurements it has a nice flat and extended frequency response. It's breakup is almost non existent and it has lower than average HD2.

http://www.audiogurman.com/1.htm#sb26adc

You mean the SB Acoustics SB26ADC-C000-4? You must know I'm a light metal fan! (heavy metal has too much resonance!)

Well, I didn't go through the rest of the SB tweeters after Mogen'd recommendation, I like that it's a black anodized dome, and the apparent Q of the break up does seem reasonable, though it may be likely that the phase plug is affecting how that behaves and measures. But gee, don't you think it's kind of pricey? I mean, BEFORE shipping it's already $52.90! :W

And to think I thought after posting this there would be no chance I'd add another tweeter in to the mix...

Just ordered a pair. :B

[5th element]

Haha Jon. I think the 'You can never have enough drivers' signature needs to come back, at least for a couple of weeks.

[BobEllis]

It's too bad that Transducer Labs doesn't seem to be able to produce in quantity. North American stock is under 50 units. It probably shouldn't get too much consideration in a build as likely to generate as many builds as a new Jon Marsh design.

[JonMarsh]

It's too bad that Transducer Labs doesn't seem to be able to produce in quantity. North American stock is under 50 units. It probably shouldn't get too much consideration in a build as likely to generate as many builds as a new Jon Marsh design.

Not without alternates... but frankly, I wouldn't be surprised if it's the last man standing when looking at overall distortion performance. It has a pretty high performance motor with monster magnets and plenty of Xmax.

I will test all drivers from scratch at multiple power levels, starting at 2.83VRMS nominal (we'll call that 1W, though for 2 ohm tweeters it's 2W). Most of the tweeters in this series are about 90dB/2.83VRMS, so any deviation will be noted. I suppose what I could do is test at absolute SPL, but then you'd have to decide, at what frequency, and for which tweeter? There's easily 3 dB variation in the passband response curves in most of these parts- so which frequency do you choose? That's a doubling or halving of power...

The way I think this will play out, there will probably be 3-4 designs developed, with the difference being the tweeters- flavor to taste or availability... if I had to guess, I'd say that will probably be

• one design with the SB26 tweeters, (either flavor- cloth or metal) (this will be the budget build);
• one with the 7100-02 for the hardcore no holds barred Scanspeak guys; drop your 6620 into it if you must;
• one with the 9900 for the retro ScanSpeak fans; will take some special tuning to work with the faceplate waveguide;
• one with the TL N26, for the esoteric USA tech spec guy(s), which might only be me...

I'm picking up more tools from the storage lockers today to get ready for this...

It's going to be a CatherineZ and GTS1000 Christmas... Yoshimura slip on exhaust on the way, DMV appointment made, and the Helibars for K1600GT arrived last night, and they look great- all I need now is the fork top baseplate from Germany!

[Steve Manning]

Jon, I can't speak to the other drivers you have listed, but I have the 9900 in my 1071's and I love them, very clean no harshness at all. I'm curious how it makes out in your testing against the others.

[JonMarsh]

I'm curious, too. I'm also hoping that Mogens's recommendations for the SB26 full frame models wind up being a sleeper performer at a budget price. I'm not expecting any surprises with the N26, but will re-test anyway, for consistency.

I just got through picking up tests cabs and baffles from the storage unit and home, as well as another pair of N26. With luck I'll route a few of the test baffles this weekend- I'm going to test them the same way I did the 6640, which is mounting them on a clean 1cu ft PE MTM cabinet I have for that purpose. Yes, there will be some baffle step introduced, but it's in the range below 1 kHz so it really doesn't impact what we care about, and the measurements can also be used for design work, then.

[---k---]

It's too bad that Transducer Labs doesn't seem to be able to produce in quantity. North American stock is under 50 units. It probably shouldn't get too much consideration in a build as likely to generate as many builds as a new Jon Marsh design.

I don't know. 50 units sounds reasonable to me. That is 25 builds. I'm mean, really how many of us crazies do you think there are? Especially at this price level. All the builds won't happen at one time. I'm sure they can order another batch when they get low. It would be very interesting to know how many ES180 and ES140 PE imported, since those went out of stock so fast.

[BobEllis]

That was 50 units including metal and carbon dome models at Solen. Meniscus has only 4 of the models Jon is testing. There are 25 of two other models in stock there, but they have slightly different characteristics and may not be a drop in replacement.

[JonMarsh]

SB26ADC specs

Just updating to include the SB26DC specifications













The weather is somewhat overcast but mild today, so I'm going to focus more on some woodworking tasks on the patio, but I may get to a couple of tweeter measurements today. Not the new ones on order, though! But maybe the 7100-02 and the N26A.

[Jonasz]

Does the SB26ADC have the same "raised" voicecoil as the SS9800?

[5th element]

It's funny you should mention that, because in the lower resolution images of those tweeters I had always assumed, the way they looked, was because of a transparent dome of plastic that was held in front of the dome as the phase shield and it extended right across the total diameter of the dome. Looking closer at the PDF for the 98 shows that that's an optical illusion and that it is in fact the voice coil sticking forwards that creates that interesting 'look'

The SB metal dome certainly has a very healthy amount of linear excursion!

[JonMarsh]

Now, you know, I've used the 9800 a couple of times, and I like a lot of things about it... EXCEPT for the 0.1 mm Xmax! That doesn't float my boat, unless you were crossing it at 3000 or more.

Now, as the VC former and positioning, I think you're right about that- the similarities when you examine them closely, or look at the SB26ADC cross section, are striking. OTOH, the moving mass on the SB26ADC is quoted as about 1/3 less, about 0.4 gram versus 0.6 gram.


Another annoying thing about the 9800 is that the VC assembly wasn't replaceable- still doesn't seem to be...

[5th element]

I was quite surprised that the 98 didn't have more xmax, especially when compared to all the other tweeters of similar outwards construction. You can get a replacement dome/coil assembly from here.

http://www.rumoh.eu/en/spare-parts/3954-replacement-kit-d2904-6000.html

Not cheap mind you, but much cheaper than buying a new tweeter.

We'll have to see how the SB one measures because according to that Russian website it's every bit as much a good tweeter as the good old D2905 range are/were. The only area it loses out on is it's slightly rising 2nd order towards the low frequencies, but if crossed appropriately that shouldn't really be an issue.

It's certainly one on my hot list, although I don't actually have a use for a tweeter at the moment. I've got plenty on the shelf!

[JonMarsh]

As of this morning, both pairs of the SB tweeters have shipped! :B

[Jonasz]

Sweet!
Any idea how the good old Seas Millennium would compare to these younger, apart from the 9900 oc, tweeters?
It sure has a healthy 0,5mm xmax

Btw, where do you predict the crossover to end up, around 1800hz? Will you use the same headscratching :stupid: topology as the original NatP crossover?

[JonMarsh]

You know, I have a pair of Millennium Excel's around, there's some measurements I've done already- just 2.83 VRMS drive-




And some axis plots....




I've always felt like the pair I got wasn't particularly smooth, though they average out OK....

I tried them in a waveguide, too--



0 - 30 degree range



25 - 40 degree range



Millenium Excel distortion in waveguide:



As you might expect, distortion lowered in the 2 - 5 kHz range.


As to the network design plans, yes, this will use the "classic" NatlieP quasi 3rd order Linkwitz concept (not normally a valid concept, but as explained in Wavecor Ardent and Isiris threads, with some tweaking works very well with drivers with a time offset; i.e., mid woofer and tweeter mounted on same baffle). I expect to keep the same combination series/parallel, it has some benefits over a straight parallel in assuring the phase matching in spite of component tolerances. :W

[Hdale85]

Will there eventually be matching surround speakers?

[JonMarsh]

Will there eventually be matching surround speakers?

No reason why not- just a Modula MT equivalent. Probably the SB tweeters would be a preferred solution for that. Have to see which enclosure models best for that.

[Hdale85]

Fantastic, I'll likely build these next summer to upgrade my HT. What would you do for a center though? I know generally a WTMW is ideal.

[PMazz]

Jon, not to alarm you but eye witnesses have spotted this over your storage space the last few nights....

[5th element]

Well the sun has been putting out lots of solar radiation and has an abundance of sun spots at the moment too. It is no surprise that the aurora is going wild

[JonMarsh]

Jon, not to alarm you but eye witnesses have spotted this over your storage space the last few nights....

It's probably due to the interaction of all that high level solar radiation we're getting lately that Matt mentions, and the magnetic field of all the drivers in there.... I've heard some people talking about an anomalous disturbance in the Earth's magnetic field in Livermore, CA! :B

Just imagine what will happen when I start storing some RAAL drivers there... 8O

[JonMarsh]

Slightly OT, but cool blog from the creator of Fuzzmeasure, my favorite measurement system (Mac only) reflecting on the 10 years since Fuzzmeasure was first created in 2004. Also and a bit about the history and response to it.

http://supermegaultragroovy.com/2014/11/19/reflecting-on-ten-years-of-fuzzmeasure/

This is one of the pics from the blog; wish my family room looked like this! :W



And the best thing about his blog is that he's re-organized his company and is well on the way to completing Fuzzmeasure 4!

[Hdale85]

Pretty sure the cost of that room alone could buy a pretty decent sized house, heck here in Ohio it could likely buy a very large nice house!

[wkhanna]

It's probably due to the interaction of all that high level solar radiation we're getting lately that Matt mentions, and the magnetic field of all the drivers in there.... I've heard some people talking about an anomalous disturbance in the Earth's magnetic field in Livermore, CA! :B

Just imagine what will happen when I start storing some RAAL drivers there... 8O

You are the verge of creating a Black Hole!

& everyone thought CA was going to drop into the Pacific!

[mkc]

And they say the magnetic north pole is moving. Ha! It's Jon moving drivers around

Jon, I seem to remember you like the Hammond organ or am I wrong? These guys are good to listen to while working in the workshop. Very simple, but I think it's groovy

https://www.youtube.com/watch?v=0epaYGk9SfM

Cheers,
Mogens

[Jonasz]

And they say the magnetic north pole is moving. Ha! It's Jon moving drivers around

Jon, I seem to remember you like the Hammond organ or am I wrong? These guys are good to listen to while working in the workshop. Very simple, but I think it's groovy

https://www.youtube.com/watch?v=0epaYGk9SfM

Cheers,
Mogens

Mogens, that music is awesome, did they record a cd? :P

[mkc]

I think you can find something on their website.

http://michaelblicher.dk/

[JonMarsh]

You can get this on Amazon and iTunes in downloadable format, the album is "Blicher Hemmer Gadd". But so far I haven't found a CD version. I did find the flac download link from their own site- will buy that today if possible!

And yes, Mogens, I'm a Hammond organ (B3 particularly) kind of guy, though at this time I only have an electronic "clone" of it from KeyB. It's a very good clone, though, actually based on samples from three different era's of B3. This is some nice stuff! Thanks for the link!

OK, just purchased, upped the price I paid to 89 DKK, it's worth it. :W

[Jonasz]

Thanks guys!

[BobEllis]

Out of stock at amazon, but http://www.barnesandnoble.com/w/blic...=5706274006551

[JonMarsh]

Direct download, various formats including ALAC and FLAC. Since I go mostly with a music server, or Mac, this just saves me the ripping stage!

http://michaelblicher.bandcamp.com/album/michael-blicher-dan-hemmer-steve-gadd

Thanks for finding them at Barnes and Noble, but $32.99 is a bit dear! I'm glad I found the sponsored download site! Just paid 14.99, and that was paying them a bonus or tip, as minimum purchase price was 70DKK, but I chipped in 89.

[BobEllis]

I'm with you there. Download is just perfect.

[DS-21]

Slightly OT, but cool blog from the creator of Fuzzmeasure, my favorite measurement system (Mac only) reflecting on the 10 years since Fuzzmeasure was first created in 2004. Also and a bit about the history and response to it.


http://supermegaultragroovy.com/2014...f-fuzzmeasure/

Thanks for drawing my attention to that post. I hope FM4 includes the capability to make polar maps, because from my perspective that's the only thing really missing.

[JonMarsh]

That's a biggie, for sure. I'd also like to see HD4 and HD5 added to distortion plots, but I'm not holding my breath. I'm very pleased to see a renewed focus on this, but hey, I like Capo, too, (as a musician) and I'm glad it's success has given him more organizational freedom.

[JonMarsh]

Sweet!
Any idea how the good old Seas Millennium would compare to these younger, apart from the 9900 oc, tweeters?
It sure has a healthy 0,5mm xmax

Btw, where do you predict the crossover to end up, around 1800hz? Will you use the same headscratching :stupid: topology as the original NatP crossover?

Here's the refresher on my "signature" crossover technique, I guess... considering I've adopted it for the big speakers now as well as the little ones...











So, it's actually dependent/works best with a time delay offset between the high pass and low pass, with the low pass time delayed- picking the right drivers and mounting and crossover frequency is part of the effort. That, and nudging things along a little bit... :W

[JonMarsh]

Work stuff today, but SB tweeters did arrive! No 9900's yet, though.

Tweeter testing and other audio work tomorrow including a look at Tubepit's RAAL project.

[JonMarsh]

OK, today hasn't turned out quite like I hoped, but I do have fresh data on three parts, the SB26STAC, the SB26ADC, and the SS 7100-02.

First, a look at a family of curves for the SB26STAC from 0 degrees to 45 degrees, 15 degree steps. These are mounted in the center of a 1 cu ft PE MTM cabinet, so they are representative to what they'll do in the "real world".



The peaking in the top end doesn't exactly float my boat; it's one of the reasons I abandoned the SS 6620 for the original Ardent design, and switched to the 6640 Be tweeter.

Distortion plots were also taken, at 2.83 VRMS (blue), 4.0 VRMS (green), and 5.6 VRMS (magenta), monitored with an HP scope. The test levels are normalized, so you can see the increase of distortion with harder drive;



Here, we see there's a pretty dang decent motor design at work, and the robust Xmax (for a tweeter) keeps the distortion low all the way down to 1500 Hz. OTOH, there is the rising second harmonic above 5 kHz, I'm not sure how audible that will be unless it also contributes to intermodulation distortion. Still, a respectable performance for a $40 tweeter!


Next up is the SB26ADC, with it's somewhat SS 9800 like construction in the VC and dome, and the use of a phase diffusor.



I like this frequency response plot better, and have to say that the phase diffuser shield is well implemented- the off axis behavior is pretty nicely controlled, with plenty of energy out to 17 kHz and 45 degrees.




Distortion is a bit more of a mixed bag, not as clean in the 2-5 kHz area, but still fairly respectable, and a bit better in the extreme top. Pretty solid performance for a $50 tweeter; will have to go back to the PE RS28 and see how they compare.

HD2


HD3


I don't have data on this at higher drive levels, but at 2.83VRMS the RS28a looks to still be the winner. I'd say the build quality has the advantage, too. But the SB26ADC has very nice SPL response, and nice off axis power response. One shouldn't rule it out at all....


OK, only one high buck tweeter in the mix today, the 7100-02.



Nice on and off axis response, good consistency, and about 5 dB more efficient than the SB tweeters.



The distortion characteristics are pretty classic ScanSpeak- moderate 2nd order distortion, but very low 3rd order, and it stays consistent up though 10-20 kHz. You do get something for your money with this one besides a pretty face...




I have to cut a new test baffle to mount and re-test the TL N26A; hopefully by next weekend the 9900's will have arrived, and I'll test both of these under the same conditions.

Out of sheer perversity I am going to repost my earlier distortion measurement at 2.83 VRMS for the N26A.



The SS is better on 3rd Harmonic, but isn't on the same block for 2nd harmonic, a measure of motor symmetry.

[BobEllis]

Thanks for that. So far I'm feeling pretty good about my N26 purchase.

[5th element]

Interesting results. The really high HD2 of the scanspeak tweeter really does jump out and just doesn't sit right with me. I mean sure, they have nice low HD3, but I tend to think that I'd rather have lower HD2 at the expense of a little extra HD3. I like my HD3 to sit at at least -50dB, -55dB is getting on for a very nice driver and any more is just awesome, but beyond around -55, if the choice was HD2 at -40dB and HD3 at -70dB, or Keep HD3 at -55, but have HD2 similar to HD3, I think I'd rather go for the latter.

It's interesting that Scan decided to go this route though, from the Russian website the Revelators HD2 is much lower, but it also has very low levels of HD3. You'd figure that they would try to improve upon both, but maybe the wide surround does something to the HD2 profile.

The Alu SB is a nice driver though. Nice flat and extended, with what looks like the capability to cross at 1.5kHz 4th order. What brought my attention to it was the fact that many people have remarked on how nice the 9800 sounds and what with SBs heritage and the tweeters visuals, the two tweeters probably have quite a lot in common.

I'm looking forward to your next lot of measurements!

[JonMarsh]

Looks to me like the SB26ADC would be a fine budget tweeter in the 2kHz and above range, and I'll do a crossover pass for it.

The high HD2 is something I've seen in most of the recent big SS tweeters- here's a measurement back around 2010 of the 66020-02.



The 6600, and 6640 are quite similar.

[Jonasz]

The Satori also got some elevated second order distortion. Seems to be the current trend. Jon, what's the mechanism that make 2nd ord dist to cause IM distortion? I've seen you mention this a couple of times. :spin: