New test data on the RS180, RS225, and more

[Mark K]

I've updated my site to reflect

Eton 19sd. (Finally, How long has it been Paul...?)

The newest version of the Seas L18

Dynaudio MW180-and 8" car version

I also fixed the link to the RS28A

[JonMarsh]

Cool, Mark, and as always, thanks so much for sharing!

~Jon

[Jed]

Thanks for taking the time to do all that work and sharing with us. The RS drivers seem to be a fantastic value, and when used in a more limited range of frequencies, competitive with drivers costing hundreds more. Am I right in saying that a RS 3 way with more conventional crossover should compete or even surpass the mega dollar euro drivers like SEAS-Eton, Scans, etc?

Actually, the more tests and reviews I see of the Eton and others, the more I think there is a lot of hype for the expensive stuff. I've been a bit of an Eton brand snob for years- might have to break out of it and try some of the new stuff offerred from parts express.

[JonMarsh]

Eton makes some great cones- used several of their drivers in the late 90's, early 2000's. But their motor designs are not state of the art.

Scanspeak has rather good motor designs, and so do some Peerless models, but I don't agree with the reasoning behind some of the SS cone designs.

It is good to have a moderate cost alternative which works so well!

~Jon

[TacoD]

Thanks for the measurements, maybe I should give those Daytons a try. I like those Eton cones for bass very stiff and low energy storage. A 11" Eton sounds different from 10" Scan or Seas Excel, I prefer those Eton's, very hard to explain why those units sound different .

[capslock]

Eton makes some great cones- used several of their drivers in the late 90's, early 2000's. But their motor designs are not state of the art.

Scanspeak has rather good motor designs, and so do some Peerless models, but I don't agree with the reasoning behind some of the SS cone designs.

It is good to have a moderate cost alternative which works so well!

~Jon

I do subscribe to your point of view!

[capslock]

Mark, while I can follow your comments about the L18 having lower distortion down low, and the RS180 doing better above, I don't see why you say the L18 has the better cone. I might follow you based on the FR, but I don't quite buy those near field measurements. The new L18 has less of an 4.8 kHz peak compared to the old one by using a curved rather than straight profile, but it is still there in the Seas data sheet but does not show in your measurement. On the other hand, from your linear distortion tests, the RS180 seems to be doing much better than the L18, which I find surprising, because both use very similar if not the same cone profiles, and the L18 has a more trustworthy looking surround (rubber vs. Santoprene or similar).

That the L18 is doing well in LF distortion does not come as a surprise. After all, it uses the W18's iron parts (T-pole + extension) which make for the most symmetrical static B-field I've ever simulated. Higher up in frequency, the inductance vs. excursion issue becomes more relevant, and the RS180 which uses Faraday shielding (possibly in the form of a full-length copper cylinder) has an edge. I suspect the RS has a more simple pole piece (straight cylinder or T-pole without extension), which explains its disadvantage down low (anybody taken one apart?). By the way, take off the bottom plate of an L18, add the copper rings below and above the pole piece and you've got a W18 motor!

The RS180 and shielded version of the W18 were reviewed in the current 5/05 issue of HobbyHifi. The RS is clearly more floppy below 4 kHz, and it has more severe break up nodes above. The old L18 was as well-behaved as the W18 below, but much worse above. I expec the new L18 to be similar to the RS180.

I can't compare harmonic distortion yet because they printed the wrong plot for the RS. They will correct that in next issue, and they also post corrections on their homepage. I will post the link when the correction comes in.

[Jed]

Thanks for the measurements, maybe I should give those Daytons a try. I like those Eton cones for bass very stiff and low energy storage. A 11" Eton sounds different from 10" Scan or Seas Excel, I prefer those Eton's, very hard to explain why those units sound different .

Does anyone have any experience with the Scan 25W8565? It was recommended to me for use in a 3 way speaker. Of course, the cost is about $180each- so if it is not much better than an RS ????

Taco- if you purchase some RS drivers, what type of setup would you use? Maybe if you have some of your eton mids laying around you could model up a 3-way using RS28a, 7-372, and 10" woofer (eton, scan, dayton, usher, seas?), if you decide to go that root.

I've also got some eton 7-360s.

Basically I need to decide what I want to do after my eton ER4/ 8-472 project is finished. I have RS28a, eton 7-360 mid or 7-372 heatpipe mids to play with plus ???.... which woofer for a 3 way or 2.5 way? I think I will want more power/bass than an ER4/7-372 2.5 will offer. Just exploring options at this point ....Suggestions?

[Mark K]

Hi Eric,

I think you're reading my linear distortion plots wrong. In every case but 1k, where they are equal, the L18 decays faster/closer to ideal. So the linear distortion numbers are better for the L18.

As to the FR. Well, this kind of quasi nearfield has to be taken with a grain of salt. I think that below 1k, it's very useful and tracks far-field well. Between 1 and 2 k, pretty good, and as you go above 2k, the correlation between a near or quasi near and a far field measurement starts to diverge quite a bit. There is still useful information there, but it's superimposed on the peaks and nulls, as well as the 6dB dropoff of the nearfild. (Since it's not nearfiled, it probably doesn't drop off at 6dB.)

Anyway, I think this method is useful for this size of driver since you'd really only use a metal 7 to 2k max. I'd go a little lower anyway.

As far as the w18 and L18-It isn't just the copper. The L18 had better low end distortion numbers than the w18. It just doesn't have the copper rings and so has a higher midband distortion numbers. So I'd be very surprised if this were the only difference. Unless the W18 I tested were different/older...

[Dennis H]

It just doesn't have the copper rings and so has a higher midband distortion numbers.

True but, according to Zaph, the L18 doesn't have the 3rd harmonic spike at 1500 like the W18 so it's cleaner sounding in the crossover region. I think we should be measuring non-linear distortion up to at least 2000 to get a handle on what's going on up there. Even a simple harmonic distortion sweep gives us some useful info.

[Mark K]

True but, according to Zaph, the L18 doesn't have the 3rd harmonic spike at 1500 like the W18 so it's cleaner sounding in the crossover region. I think we should be measuring non-linear distortion up to at least 2000 to get a handle on what's going on up there. Even a simple harmonic distortion sweep gives us some useful info.

Well, in this case you don't have to even measure it. Any large ringing/resonance will cause the corresponding xth order HD plot to peak at (1/x)*f

That is, the L18 peak breakup is around 7k. So the third order HD graph will have a peak at approximately (1/3)*7k or around 2.3k, give or take. The second order, 3.5k. The fourth order, 1.75k, etc. Now usually the orders higher than 3 are fairly low, so you can usually ignore them.

I have been thinking of doing close thirds sweeps though. I think these may be more interesting in regards to breakup and a bit more immune to echo contamination.