Speaker driver measurements: 5-8" Woofers

[TMM]

Hi all,

Much in the same way as my existing 2-4" midrange/fullrange driver testing, I've decided to publish measurements of larger woofers I have on hand too.

The format is the same as previous testing of smaller drivers with the addition of a new test method to evaluate bass distortion.

To start us off we have a group of four 6.5/7" woofers. I'll hold off writing my interpretations of the measurements until I have a few more measured and a better idea of the baseline performance.

You can view the measurements and commentary here:
Timothy Feleppa's Pages

Any feedback is always appreciated. Cheers!
-Tim

[Bear]

Hi all,

Much in the same way as my existing 2-4" midrange/fullrange driver testing, I've decided to publish measurements of larger woofers I have on hand too.

The format is the same as previous testing of smaller drivers with the addition of a new test method to evaluate bass distortion.

To start us off we have a group of four 6.5/7" woofers. I'll hold off writing my interpretations of the measurements until I have a few more measured and a better idea of the baseline performance.

You can view the measurements and commentary here:
Timothy Feleppa's Pages

Any feedback is always appreciated. Cheers!
-Tim

If the Usher is a SS clone, then they seem to have made an improvement or two (reduced resonance ~1.2K, reduced break-up, all presumably helping the cleaner odd order harmonic performance...). Good to see, thanks!

[TMM]

The level of 5th order approaching 600hz is a bit high on the Usher though. The 3rd order rising around there is a bit odd to me - the 5th order is explained by the breakup node at 3kHz but the 3rd order isn't. 2nd order also peaks a bit around there at high volume. I'm wondering if what we are looking at is a combination of a resonance at 550-600hz and a less linear motor that allows the cone breakup to be seen in the 5th order.

I have another pair from a newer batch (2014), might test them to see if there is any variance.

Sent from my Nexus 5X using Tapatalk

[HiFiCompass]

Nice job! Thanks!

[TMM]

I've finished testing nine 6.5"/7" woofers and have added commentary and ratings for them.

Drivers tested:
ScanSpeak 18W/8545K00
Usher 8945A
SEAS CA18RNX (H1215-08 )
Vifa NE180W-08
Dali Ikon 6 MK1
Fountek FW168
Dai-ichi W6S-52
MCM 55-5665
Response CW2194

I've added an Le(x) test which gives some insight into intermodulation distortion, especially important for woofers which are used for both bass and treble, such as in a conventional 2-way with a dome tweeter.

It seems we have a large group of drivers over a wide price range which do some things well and other things not so well, but I'd say there are no particularly poor performers overall.
We also have a few very high performing woofers that are a significant step up from the rest.

Link to results

[nerd of nerds]

Thank you for posting this! The MCM drivers you've been measuring look very interesting. Do you have any plans to test the 8" version or the 5" midrange (I think 55-5660)?

[TMM]

The 5" has disappeared from element14/farnell/newark for some reason even though the 3,4,6,8" versions of that series are still there. I'm not really interested in 8" drivers at the moment but have 1 or 2 lying around that I might throw on the test baffle for fun. I think I've blown my budget for buying speakers for the next few months at least - I probably need to stop for a bit and build something to learn more about what works in practice, what doesn't, and gather ideas about what to investigate next

[Bear]

I've finished testing nine 6.5"/7" woofers and have added commentary and ratings for them.

Drivers tested:
ScanSpeak 18W/8545K00
Usher 8945A
SEAS CA18RNX (H1215-08 )
Vifa NE180W-08
Dali Ikon 6 MK1
Fountek FW168
Dai-ichi W6S-52
MCM 55-5665
Response CW2194

I've added an Le(x) test which gives some insight into intermodulation distortion, especially important for woofers which are used for both bass and treble, such as in a conventional 2-way with a dome tweeter.

It seems we have a large group of drivers over a wide price range which do some things well and other things not so well, but I'd say there are no particularly poor performers overall.
We also have a few very high performing woofers that are a significant step up from the rest.

Link to results

Thanks, Tim! The Vifa certainly looks good. And performs well, too! Have you considered doing a horizontal polar plot for a few of the better performers? It's obviously quite a bit of hassle to do whole cloth, but it's good data point for larger drivers to see how the sound field collapses with frequency as a driver gets closer to outright beaming, especially for these larger drivers.

[TMM]

The important polar information is mostly there in the frequency response plot with the 15, 30 and 45degree off-axis measurements. My test baffle is really too small to make an accurate polar plot at the frequencies where nulls begin to appear for drivers of this size - even 30cm away from the driver is getting fairly close and as I approach 90degrees at that distance the microphone becomes too close to the edge of the baffle to make an accurate measurement. We can see from the results that for all 6.5/7" drivers, there is going to be some significant droop in the response above 3kHz when you become more than 30degrees off-axis. Below 1kHz, all of them are near omni-directional.

[Bear]

I'd mostly agree with you, though for 18cm+ drivers, I do find comfort in a nice polar plot. But even Vance doesn't do it very often, and he's (presumably) getting paid to do driver testing.


I was/am considering a 2-way or 2.5 way with the SW223, and seeing the dispersion at 1500Hz beyond 45 degrees informs me (perhaps wrongheadedly!) that it's feasible with a decent off-axis (>45*) performance.

[TMM]

The polar response normalised to 0dB at 0degrees is going to look very similar for all drivers having the same physical cone size. Where they will deviate is in the presence of cone breakup as the cone stops radiating as a piston. This is just going to change the positions of the ripples/lobes slightly - the overall width of each curve is going to remain similar. For any decent design you should have your crossover significantly below where the cone breakup becomes severe anyway.

Smaller drivers tend to have smoother frequency response, lower midrange distortion, better polar response but poorer low frequency distortion and bass extension. Technically 6.5" is already too large for a 2-way crossed at 2kHz because if we consider that the cone is about 5"/130mm in diameter, a deep null will occur at F=340/(2*0.13)=1.3kHz at 90degrees off-axis, so our power response will have a dip. In choosing a driver for the bottom of a 2-way, always a compromise has to be found between frequency response, polar response and non-linear distortions, so less than perfect polar response may have to be accepted in favour of better parameters elsewhere. In my opinion, around 6.5" is the ideal compromise for a 2-way at normal home listening levels (peak 90dB SPL @ 1m).

The SW223 doesn't look ideal for 2-way for reasons other than just frequency response or polar response. It's intentionally marketed as a subwoofer. The breakup node at a relatively low 3kHz will raise distortion components in the midrange (3rd order at 3/3=1kHz, 5th order at 3/5=600Hz etc). In the datasheet it can be seen that the 3rd harmonic reaches 0.5% (-46dB) at around 1kHz because of this. The better 6.5/7" woofers I've tested/seen are maintaining about -60dB (0.1%) 3rd order there. Designs with a huge amount of xmax typically further compromise distortion in the midrange and treble because they use long coils which compromises Le(x). Sensitivity is also poor.

[Bear]

Yeah, the third order is quite good up until ~900Hz in Vance's measurements. It's definitely not a candidate for a two way, though there is a certain charm in having a sub + tweeter in a loudspeaker. Right now, I'm weighing which mid range to shove in between. Since I have both RS52s and ZA14s on my shelf, those are my default choices.

The rest of the Vance's measurements...

http://www.audioxpress.com/article/test-bench-wavecor-sw223bd03-8-75-inch-high-xmax-subwoofer