Linearity with level issues with AMT tweeters: just Goldenear or all of 'em

[Pallas]

Soundstage recently published measurements of the Goldenear Triton One taken in the NRC's anechoic chamber.

True, the Triton One is astonishingly expensive for a speaker so cheaply finished (black sock over painted MDF, with small polished endcaps) at $5000/pr. But for this thread Sandy Gross's chutzpah is neither here nor there.

While ever since I first heard Goldenear speakers I thought they were way more hype and ad dollars than engineering-driven sound quality, this dataset exposes some truly eyebrow-raising flaws. What really shocked me is what a hapless dog this driver turns out to be:



I say that because of the linearity curves the NRC takes, that nobody else does. Here's the difference between the speaker's FR at 70dB and at 90dB


Compared to the 70dB level, you can see massive amount of compression at the bottom of the tweeter's range: almost 3dB where the ear is most sensitive!

At 95dB it's even worse yet:


Now, for comparison, here's a $170/pr Infinity Primus:
90dB


95dB


Is that a common issue with AMT's, or did Goldenear just cheap out more spectacularly than most?

[Pete Basel]

Re: Linearity with level issues with AMT tweeters: just Goldenear or all of 'em

Very interesting!
Here is the Guru QM10 that uses a low cost 14mm Visaton tweeter:

http://www.soundstagenetwork.com/measurements/speakers/guru_proaudio_qm10/

It looks as if it has very slight gain at 90 dB, and then significant compression at 95, still
it is a small tweeter. Also, it uses a poly-fuse and I'm not sure if that might contribute
significantly if it is close to the trip point.
Do you know if the GoldenEar employs a poly-fuse?

[Pallas]

Re: Linearity with level issues with AMT tweeters: just Goldenear or all of 'em

It looks like the Raal is quite good in this regard, at least up to 90dB. Here's the 70dB-90dB difference measurement for the Selah Audio Verita, which has clear midrange mushroom cloud issues as one would expect from the driver complement but up to 90dB at least very consistent FR.



Alas, they didn't publish a 95dB measurement, which is why I wrote the "at least up to 90dB" above. The upper bass compression is a little surprising, considering that's a Scan Illuminator. Maybe a good indicator that one should ideally have more cone area for that region, if even the best drivers compress a dB or so at high-but-not-obscene SPL...

[Pete Basel]

Re: Linearity with level issues with AMT tweeters: just Goldenear or all of 'em

There is a fad for non-thermally conductive formers these days and thermal capacity takes a hit.
Interesting that the high end Vifa drivers, employ a titanium former - nice!

http://www.parts-express.com/tymphany--ne180w-04-6-1-2-full-range-woofer--264-1096

[Pete Basel]

Re: Linearity with level issues with AMT tweeters: just Goldenear or all of 'em

Interesting that you bring this up, because I was going to start a thread "What is Wrong With This Picture" and link to this test data:

http://www.soundstagenetwork.com/index.php?option=com_content&view=article&id=1043:nrc-measurements-magico-s5-loudspeakers&catid=77:loudspeaker-measurements&Itemid=153

Main review:

http://www.soundstagehifi.com/index.php/equipment-reviews/657-magico-s5-loudspeakers

My guess is that these speakers use a poly-fuse, and perhaps they shouldn't.

[dcibel]

Re: Linearity with level issues with AMT tweeters: just Goldenear or all of 'em

Keep in mind that these compression measurements are at 2m, not 1m. Not many home speakers will be compression free at this volume.

For some additional reference, the NHT Xd is here, which uses a Seas Excel 5.25" midwoofer and a nice Seas 1" aluminum dome tweeter with a neo motor:

http://www.soundstagenetwork.com/measurements/speakers/nht_xd/

This is what I would consider a "high end" speaker, but at [email protected] you can see that the small 5.25" woofer has run out of steam (its a small sealed cab), and the nice tweeter is a couple dB down as well. The NHT Classic Three performs a lot better in this regard, utilizing a unique 3-way design which takes a lot of the power load away from the tweeter.

Here's a quick link to a lot of speaker measurements available at Soundstage:

http://www.soundstagenetwork.com/measurements/speakers/

[Pallas]

Linearity with level issues with AMT tweeters: just Goldenear or all of 'em

The NHT Xd performs to my expectations for a 5" Seas Excel - tiny cone area, typical Seas not-very-good suspension; I've always thought the Seas cone drivers overrated - and the tweeter looks a whole lot better than the Goldenear's AMT.

However, the Revel Performa3's, with the aluminum SB Acoustics tweet in a small waveguide, have surprising to me levels of nonlinearity with level. I've only heard the M105 but I thought it was excellent for a little pea-shooter. Way better than, say, a KEF LS50. It is a pea-shooter, and a dealer demo (with the Teac DAC/amp using Abletec modules) so I didn't push it.

[johnnyrichards]

Re: Linearity with level issues with AMT tweeters: just Goldenear or all of 'em

The big Cerwin-Vega measured better than anything else on that page, except for being super hot on the bottom end.

Glad I don't listen to stuff at 90-95db peaks anymore. Common sense prevailed and I realized driver behavior above 90db us a non-issue for anyone that wants to preserve their hearing.

[Jay1]

Re: Linearity with level issues with AMT tweeters: just Goldenear or all of 'em

Some of us sit more then 4 meters from the speakers, and would like to watch movies at reference level. I feel even better about my current build direction. Nice thread Pallas.

[donradick]

Re: Linearity with level issues with AMT tweeters: just Goldenear or all of 'em

Thanks for the post Pallas!

I've often wondered about the vast differences in speaker sound, and this is not only a very good aspect,
it will also be very easy to measure for my DIY designs. I've been musing a lot lately about the dynamic
range of a live orchestra and just how limited our usual reproduction is.

[dcibel]

Re: Linearity with level issues with AMT tweeters: just Goldenear or all of 'em

Some of us sit more then 4 meters from the speakers, and would like to watch movies at reference level. I feel even better about my current build direction. Nice thread Pallas.

I hope you chose to go pro ;)

Looking through the sets of measurements, Its clear that 5dB in volume can be the difference of several dB compression loss due to drivers hitting their limits. So it makes sense to go pro in this regard when high volume is required, as pro drivers have a two fold advantage here of 5dB+ extra sensitivity and several times the power handling as well. A lot of pro drivers even list power compression figures at full rated power.

[Jay1]

Re: Linearity with level issues with AMT tweeters: just Goldenear or all of 'em

I hope you chose to go pro ;)

1.4" CD 15" 2 way. Still not settled on a woofer, need to build a test box before I order a woofer to try.

[carlspeak]

Re: Linearity with level issues with AMT tweeters: just Goldenear or all of 'em

Does anybody know if any of the speakers tested at soundstagenetwork have a Mundorf AMT19CM1 tweeters?

[Pete Schumacher]

Re: Linearity with level issues with AMT tweeters: just Goldenear or all of 'em

And then there's this anomaly from the little Harbeth. All the compression is from the woofer. The tweeter is performing admirably at 95dB relative to 70dB. From the impedance plot, crossover to the tweeter looks to be at 2000Hz or below. Midrange mushroom clouds aside, that's a nice performing tweeter; much better than the KEF Reference 201/2 for example. It makes you wonder what types of padding resistors they might be using that could be causing some of the compression by changing value with heat.

[fdieck]

Re: Linearity with level issues with AMT tweeters: just Goldenear or all of 'em

There is a fad for non-thermally conductive formers these days and thermal capacity takes a hit.
Interesting that the high end Vifa drivers, employ a titanium former - nice!
http://www.parts-express.com/tymphan...ofer--264-1096

It is not a fad but a return to higher Qms values. Some of the order drivers that used aluminum formers used very thin formers which had higher resistance and reduced eddy currents. It is really interesting to look at some classic drivers (Wharfedale, Bozak, vintage JBL for example) and see how many features are returning in the new high performance drivers. Paper voice coil formers were still used on many drivers around long after aluminum formers were introduced. Ever ask yourself why?