Seas Excel W18EX001 vs Scan-Speak 18M/4631T00

Status
This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.
Back at it again with another question.

For my 3-way build I was planning to use a Seas Excel W18EX001 as mid to cross with my Fountek NeoX3.0 at 2 kHz but I recently saw nearly everyone using a NoeX3.0 saying they cross it at at least 2,5 to 3 kHz because it should sound way better. So I got in some trouble with the mid woofer. On the frequency response graphs the Seas Excel W18EX001 seems to just make it to 3 kHz on- and 30º off-axis but when I was researching if it can actually do that because 3 kHz is pretty high for a 6,5" I saw many reviews saying it has cone break up at 1,7 kHz and the highest possible crossover frequency should be at 1,6 kHz but when I looked up the source from where the people who said that got that information, Zaph, it was indeed the magnesium cone W18EX001 but it seemed to be an older version. It was reviewed in 2006 and has a way smaller motor on the picture. And at the same time I saw people using it to about 2,5 - near 3 kHz so as you could imagine, I got confused.
So can I use a Seas Excel W18EX001, brand new, latest version which is on the website of Seas right now, to between 2,5 - 3 kHz?

I've already picked replacements in case it can't play to above 2,5 kHz: the Scan-Speak 18M/4631T00 or it's little brother the 15M/4531K00. Which one purely depends in aesthetics. I will just pick which one depending on which one I find to look better in my design. Shout at me for compromising sound quality but I don't care because they both sound amazing. The 15M has a lower moving mass and better dispersion but so slightly it's practically irrelevant, the bump starting at 1 kHz is way bigger and it doesn't get as loud. The 18M still has a amazingly low moving mass, the bump is smaller and it gets to 110,2 dB which is insane. I aim at 110 dB, with baffle step if a certain woofer doesn't get to 110 dB at it's own (which is nearly always the case). The normal W versions of these woofer are already known to sound amazing anyway so these M versions should sound even better for my midrange application.

But if it can play to at least 2,5 kHz without any problems I will certainly prefer and take the Seas Excel because it also has amazing specs. Because of it's magnesium cone. Because of it's large, powerful motor it has an amazing impuls response, even better than mid-level Accuton and in my opinion it looks amazing. And I initially picked it because of those characteristics. If it works, I love it.
And all of these nice features have got it even better reviews than the Scan-Speak. It should sound more neutral and accurate.

BTW, when doing research for if this would work or not, I noticed the crossover makes all the difference here. So to inform you of my capabilities, I use a MiniDSP 2x4 to control it and do all the corrections. Separation is being done by a passive filter, but just the separation. Everything else is handled by the DSP making a sort of hybrid crossover. I do it this way because I also need two outputs of my DSP to control the subwoofers (they are just the woofers in the towers but subwoofer drivers so they go very low) and I didn't want to have to use two MiniDSP 2x4's and something with more channels in one unit didn't fit my budget.
 
I think it's more than just crossing at this freq. or that freq. It depends on how it is crossed and using which order filter. You may also have to use a notch filter for the Seas. The Seas will have a lot more resolution vs. the Scan Speak. I have used the Scan Speak Revelator and Illuminator and although there are very good drivers, I think at the end, both being paper, and no matter how good they are, there is a limitation on how much resolution the paper material can muster. I personally would go with the Seas if it means having to work harder on the xover. The Seas will have a lot more resolution especially on the low frequencies.
As for the tweeter, I personally don't like the ribbon. I would go with a good dome and cross at 2khz.
 
I was planning to use 2nd order because that's the recommended but if you think something else would be better, tell me. I have the ribbon because no dome tweeter of my around €150,- budget has the bandwidth and smoothness I want. And anyway, if you don't like ribbons, that doesn't mean I have to dislike them as wel.
 
18M is a midrange with foam suspension (confirmed by Scanspeak, I asked them by e-mail as some websites list it as SBR), high Qts and shortened voice coil. You should be really comparing the Seas with the 18W/8531.

My 2c on this - the motor of the Scanspeaks is much superior to the Seas but that is less important if you limit their excursion. The Seas would need to be used really not above 1.6Khz - the cone resonanse peak is centered at what looks like 4.8Khz, 3rd harmonic distortion will be spiking at around 1.6Khz. That means you`d have to pair it with a large waveguide or use 24db/slope as few tweeters can do that low. The Scanspeak also has higher cone area - 152sqcm vs 126 for the Seas so it will move less for the same sound level. The breakup, if used without a crossover, sounds like someone is blowing in an aluminium pipe and it can be excited by tones corresponding to the 1.6Khz frequency as it will produce its own harmonics. But these metal cones do produce bass really better than paper in my opinion.

The problem I have with the 18M is its way too efficient so will need padding or more SPL support from below. Look up a thread I started on it vs the 18W - its somewhere in Multiway - hificompass has measurements of the driver and it looks very very good.
 
I don't think Sd is to important. Moving more air is just louder right? Yes the 18M can get wayyy louder dan the Seas but not just the Seas but it the 18W which you say is a fairer comparison doesn't get as loud as the Seas. The SPL depends on more than just Sd and the effect of moving more air in terms of better bass response also doesn't matter to me at all since I will use it as mid. You often mention bass but I don't care about the 18W being able to produce better bass. These will be my mid in my 3-way towers. These are the drawings (still with Seas). The driver we are discussing now is the mid on top. At the bottom are two 8" SB-Acoustics SB23MFCL45-4 subwoofers they will play to between 200 and 250 Hz (I'll play with that with the DSP) so the mid has to play from that to 2,5 - 3 kHz.

It seems pretty obvious to me that the 18M is better for my application because it is actually designed to be used as a mid, has a better frequency response in the intended range, lower moving mass and so better impuls respons. What should make the 18W better for me as a midrange?
 

Attachments

  • Schermafbeelding 2018-06-28 om 12.27.35.png
    Schermafbeelding 2018-06-28 om 12.27.35.png
    538.8 KB · Views: 1,625
Oh and another probably pretty important thing. In the other thread about these Scan-Speak you were all talking about the 0,707 Qtc in a sealed enclosure. Why, is this some miracle number? And do I need to have it as well for my midrange enclosure? What happens when it is not 0,707?
 
The W18EX001 is really better suited as a woofer because the peak in the frequency response boosts up the non-linear distortion components to unfavourable levels at much lower frequencies than it occurs at. See:
http://www.zaphaudio.com/6.5test/W18EX001-HD.gif

To avoid the peak in 3rd order harmonic distortion at ~1.7kHz you'd want to cross around 1kHz, to ensure that there is already significant attenuation of the fundamental at 1.7kHz and therefore significant attenuation of the 3rd order harmonic which lands on the 5kHz cone breakup peak. The benefit of metal cone drivers is that there is a region about an octave wide over which there is less breakup than an equivalent paper cone driver - for the W18EX001 this is perhaps 800Hz-1.5kHz, which translates into very low harmonic distortion from about 200-500Hz making it an excellent woofer.

The 18W/8631 doesn't have this problem due to its benign cone breakup behaviour, so you can cross as high as you want.

Oh and another probably pretty important thing. In the other thread about these Scan-Speak you were all talking about the 0,707 Qtc in a sealed enclosure. Why, is this some miracle number? And do I need to have it as well for my midrange enclosure? What happens when it is not 0,707?
in box response of Q=0.707 somewhat simplifies the crossover design as it is a perfect 2nd order butterworth response. Therefore when you combine it with a 2nd order butterworth electrical it creates a perfect 4th order linkwitz riley acoustical response. If you plan on crossing at a different frequency than the box Fb, sizing the box to achieve a lower Q (make the box bigger) will make designing the crossover easier as the acoustical slope that you get from the box will be more gradual and won't cause a big 'hump' in the response where you don't want it.
 
Last edited:
So long story short. I can't use the Seas as my midrange from 200 - 250 Hz to 2,5 - 3 kHz so we'll scrape that one from the list., it's done for.
Then the Scan-Speak argument. 18W vs 18M. I think it's pretty obvious the 18M is better as a midRANGE than the 18W but bring your arguments. Again I DO NOT care about it's performance as woofer but only in it as a MIDRANGE playing MIDRANGE frequencies from 200 - 250 Hz to 2,5 -3 kHz. I do not care about bass performance.
 
I've already considered those for a previous concept but moved away from them because I know they are good but Scan-Speak are the best. So now you're all of the sudden saying I can't use any midrange of that size? You just said about the 18W I can cross it as high as I want.
Do you think I should take a better (dome) tweeter? Domes with the same kind of frequency response are expensive and I already have the NeoX3.0's so it will be a pain in the *** but maybe I can do a little trading.
 
18W measured here:
http://www.zaphaudio.com/6.5test/18W8531G-HD.gif
http://www.zaphaudio.com/6.5test/18W8531G-FR.gif

18M here:
ScanSpeak 18M/4631T00 | HiFiCompass

These measurements are from different sources so we have to compare them with a grain of salt, however I trust that the measurements are accurate from these guys. We also have to assume that 8ohm and 4ohm drivers measure similarly in frequency response smoothness and harmonic distortion (they usually do).

Comparing harmonic distortion:
Unfortunately we can't compare the 2nd order harmonic as the microphone used at Zaph audio has a 2nd harmonic noise floor of perhaps -50 or -60dB. This is a common problem with inexpensive measurement mics. The mic used at HifiCompass seems to be considerably better. Higher orders are unaffected.
The overall level of harmonic distortion seems to be very close between the 18M and 18M. The 18W peaks up in 3rd order distortion at 400Hz and the 18M peaks up with similar severity at around 1kHz. These are probably due to minor cone/suspension resonances. Take your pick - they are both very good performers.

Comparing Frequency Response:
The 18W rises about 4-5dB from 1-2kHz whereas the 18M only rises about 3dB. Perhaps the 18W might be a little bit smoother overall. This might just be indicative of the two different measurement setups though.

If I had to pick from these results, I'd go for the 18W. The T/S parameters also allow for a slightly smaller box without excessively high Q. If i recall correctly, HifiCompass prefers the 18M - maybe they can chime in with their opinion. Theoretically the 18M should be slightly lower distortion in the midrange because it has a shorter stroke motor (lower xmax) however I don't see any significant difference in these measurement results. Perhaps inter-modulation distortion testing is needed to reveal this advantage, however I would guess that if there is any advantage it is only slight since the harmonic distortion sweeps show such similar performance.

Both are very good drivers and I think you'll be happy with either.

Any midrange with a size of the 18M will start beaming at 2Khz or so, for your crossover point a Satori MR16 driver may be better suited.
An MR16 is barely any smaller in diaphragm size than a Scanspeak 18cm. If a SS 18cm has unacceptable beaming above 2kHz then an MR16 is unacceptable above 2.3kHz. Not a big difference. To significantly improve on the SS radiation pattern you need to go to a much smaller driver - say 3-4" and this won't be able to cross at the 250Hz needed.

edit: The on-axis performance of the 18W/18M dictates no hard limit on the crossover frequency because it is exceptionally well behaved all the way up - the frequency response is smooth and the harmonic distortion doesn't rise out of control. You do however have to decide how high to cross to maintain good power response (smooth off-axis response) due to the size of the drivers and centre-to-centre distances. I wouldn't cross any higher than 3kHz because of the size of the mid and the achievable centre-to-centre distance with the tweeter. Crossing higher than that will result in a considerable dip appearing in the system frequency response at a relatively shallow off-axis angle. As always, lower crossover frequency is always better for power response. For near-perfect power response you want to cross a 6.5" woofer at about 1kHz. Obviously you'll have to compromise because almost no tweeter will play that low. It's up to the speaker designer to pick a frequency which achieves acceptable power response and doesn't push the tweeter too hard (making it distort).

One solution is to put another driver of an intermediate size between a 6.5" driver and a tweeter - say a 2" dome mid then cross at say 700Hz and 3kHz. Another is to substitute the 6.5" driver with something smaller (~4.5") and have your bass driver play higher - perhaps 400Hz instead of 250Hz. This goes against the idea to get one driver playing as wide a frequency range as possible however. Everything is a compromise in one way or another - just gotta pick your battles :)
 
Last edited:
Member
Joined 2009
Paid Member
If qtc don t match with the target you have and you have sealed box for it you can use Linkwitz Transform to make Qtc and FS you want to achieve. As you don t target low end extension you may have benefits in term of Spl.

About the larger sd, distortion increase with movement of membrane of loudspeaker. Increased sd means louder yes, but it means lower movement so lower thd for a given spl.
That is one of the reason some prefer larger drivers for mid duties ( 12" or 15"). I m one of them.
There is an interesting paper about B&W engineers about that when they switched from one 15" to 2x10' in 803 models (iirc). Engineers thinked it sounded better with one 15, marketinf department it would sell better with 2x10".
Apparently they sell better. ;)
 
Two SB12s in MTM is going to have similar problems with off axis response as a single 18W because the entire assembly (M+T+M) is going to be similar in height to the larger M+T. edit: in the vertical axis at least, which is the main concern here since we are pushing the XO frequency higher than is optimal. Horizontal response will be improved, but that's not the axis that will experience a major problem in the first place.
 
Last edited:
If I can return my NeoX3.0's for around €200,- and earn another €125,- by selling other left-over drivers (which I was already selling anyway) and maybe ask one driver or some crossover components for my birthday I might be able to take new dome tweeters and fit it in the maximum time I want to spend raising money. Maybe (and that's a real strong maybe) I might even ask some components or a driver for Sinterklaas (it's like dutch Christmas) as well.
The Scan-Speak Illuminator D3004/6620.00 looks really good. I often see people cross those lower, right?
 
Status
This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.