Seas W22 3-way Question

[Brian Walter]

Seeing as how quite of few people are saying that the midrange of the Seas Excel W22 is pretty much tops, what would you recomend using for an upper midrange, if you wanted to use a ribbon tweeter? A lot of the designs around here are using really low crossover points which limits you considerably in regards to tweeters. I would really like to go with a ribbon tweeter, and not many that I know of will go low enough to cross directly to the W22.

I'm not sure I want to go this route, I'm just fishing so to speak at the moment.

Brian Walter

[JonMarsh]

Depends on how deep your pockets are. W15CH is one to consider; I'd suggest the Accuton C44, as the W22 has very low distortion in the mids up to 1200 Hz or so, which would be a very easy crossover point for the C44 2" dome mid. It would easily handle scaling up to the crossover point of any ribbon- say, 4-6 kHz. Higher is better for the ribbon, but there are driver spacing issues which are a function of wavelength, which would make me suggest lower.

[Jed]

Seeing as how quite of few people are saying that the midrange of the Seas Excel W22 is pretty much tops, what would you recomend using for an upper midrange, if you wanted to use a ribbon tweeter? A lot of the designs around here are using really low crossover points which limits you considerably in regards to tweeters. I would really like to go with a ribbon tweeter, and not many that I know of will go low enough to cross directly to the W22.

I'm not sure I want to go this route, I'm just fishing so to speak at the moment.

Brian Walter

Well, I just think you can do better for a woofer than the W22, and I will know this for sure when I replace my W22s with some RS225s in my system which also uses the SEAS W15CH and OW1 tweeter. Note: I have heard the RS225 in a similar 3-way, and IMHO they were better under 300HZ than the W22. I'll do the swap this week in my speaker since the RS225s are a perfect drop in replacement with my current X-over topology!!! I feel the best X-over point for the W15CH is below 2.5K because disortion spikes up just over 2K for the W15. I'm using a Hiquphon OW1 with my W15CH and it sounds amazing now that I've got the power response worked out better. Crossed over around 2.2K LR4 MT and 325ish LR2 between WM. Roman B helped me with the crossover design, so if you are interested you can email him. He hangs out on the madisound board a lot, so you can contact him that way too.

Now, if you are set on using the W22, how about choosing a different tweeter like a millennium and use the W22 as a midrange? A better woofer might be those RS subwoofers or for aesthetics, maybe a W26 in a sealed enclosure with Q.7 - about 50L and then use an active 15" subwoofer for the bottom end below 40HZ? I'm seeing an oversized sophia clone type speaker for the mains. Sorry, I got off track again. Yeah- do what Jon said :T

Jed

[Brian Walter]

Jed, I was originally thinking of running the W22 a little bit higher as I have a RS10HF sub that I intend to use on the bottom. I was thinking that the RS10HF should be good to around 300 hz and then I'd cross to the W22 and run it up to around 1200 hz. I guess there are all kinds of ways to skin a cat, so to speak. Keep us posted as to how your woofer swap or exchange works out. I had considered using the W15 and Hiquiphon quite some time ago, but I guess I had somewhat forgotten about that.

My original intent was to use the RS10HF with the RS180 and either the Seas H1212 or the RS28A tweeter, but I was being urged or pushed to go with the dual RS180's in an MTM and then add another RS10HF to the setup. This is starting to get a little bit larger than I wanted, so I've been looking into other combinations. I will be using and active crossover between the RS10HF and whatever is above that, so sensitivity isn't an issue.

Thanks for your input.

Brian Walter

[JonMarsh]

Dual RS180s and a single W22 will require about the same amount of enclosure volume for the same LF tuning. IF you only use them as midrange (say, 145-1500), then this isn't too critical, except I think that you sould err on the side of enclosure depth, to help absorb/control the backwave.

~Jon

[Brian Walter]

Jon, right now I've been investigating building something more like David Gatti's Delta speaker. I don't remember it we are allowed to post links on this site, so I won't. But basically he has a 10" scanspeak, a 7" scanspeak, 4-5" Skanning and a Morel 110 Supreme tweeter. I was thinking along the lines of using the 10" RSHF, Seas W22, ? and possibly a AC 2Gsi or similar ribbon tweeter. I will admit that the Acuton is probably out of my price range. But at the same time I have been considering adding the RS10HF to the Skanning 7" and LCY tweeter combination that E-Speakers sells. This wouldn't be very cheap either, but might be rather nice sounding and easier for me to design.

At the moment I'm just investigating a bunch of options hoping that one of them will jump out at me as the one I would like to build. Thanks for spending the time to point out various options.

Brian Walter

[JonMarsh]

Those sound like intersting options to consider, but most ribbons don't like being crossed lower than 4 kHz, and most midwoofers aren't clean enough to run up that high, which makes for some transistion problems. A compromise is choosing a mid that will make it up to 3 kHz, and settling for that with the ribbon.

The Skanning drivers should be easier to work with, with their more damped cones. Good luck with your project which ever way you decide, and keep us posted of your progress.

BTW, I like the RS265HF a lot myself, and may be using it in a new sealed cabinet project. I know, I know, what's Mr. Dipole doing building another sealed cabinet speaker, when I swore the Modula series would be the last ones!

~Jon

[Brian Walter]

Jon, have you had any luck at all with ribbon tweeters or have you found that because they need to be crossed over so high that they just doesn't work very well with most midranges. Another option might be to go with a dome midrange. I wish the RS dome midrange would become available. It would be really nice to see some response curves for it. Maybe it isn't worth waiting for. Overall, I haven't been too impressed with dome midranges, the ones I've heard seemed to lack impact. Now maybe that was more of a problem with the particular implementation, I don't know.

[JonMarsh]

It's not easy to do a good crossover with dome mids, and pick a workable crossover point- often folks err on crossing too high, which makes for problems in the crossover region due to the physical separation of drivers. I've used the MB Quart 2" with success in several designs, as well as the Accuton C44. Big soft dome mids don't do much for me personally- BTDT.

I'm not against ribbons by any means- I have a box of 16 big Fouteks waiting for me to free up some time. They'll be crossed, 8 to a side, with a BG RD50 planar driver (sometimes called a ribbon, but misnamed that way), at about, well, you guessed it, 4 kHz. The BG goes up to ~18 kHz, so the crossover frequency is more dictated by avoiding the cavity resonance and reduction in disperion of the BG RD50, as well as the capabilities of the larger Fountek, where they work best (just look at CSD plots to get an idea).

Ribbons that are big enough to have decent power handling dont' have much vertical dispersion, and the only way around that is to either put your head in a clamp (not so desirable) or to use line arrays. Line arrays are expensive, unfortunately. I've yet to hear a moderate cost ribbon that beats a Millenium Excel in top to bottom naturalness, especially when you factor in dispersion. But they're fun to play with...

~Jon

[Jed]

Jon, have you had any luck at all with ribbon tweeters or have you found that because they need to be crossed over so high that they just doesn't work very well with most midranges. Another option might be to go with a dome midrange. I wish the RS dome midrange would become available. It would be really nice to see some response curves for it. Maybe it isn't worth waiting for. Overall, I haven't been too impressed with dome midranges, the ones I've heard seemed to lack impact. Now maybe that was more of a problem with the particular implementation, I don't know.

Not a ribbon, but a nice sounding pseudo ribbon is the Visaton MHT-12. It has pretty high 2nd order distortion when played loud/ over 90DB, but it sure is detailed and it probably has less distortion than most true ribbons. I saw a Hobby HIFI test of the tweeter and they must be used over 3K and preferably around 4K on up. I used them in a speaker with Eton 4-300 midrange and 7-360 woofer. I gave that speaker to my mother and she is very happy with the sound. Detailed, but very easy to listen too- if not a touch too warm for my taste. The Visaton, is detailed and airy, and might be just the ticket for a 4 way speaker that you are proposing. RSSub to 70-150HZ, W22 in about 25L sealed Q.7, Accuton 2" dome, and Visaton MHT-12.

Wait a minute--- maybe I should give that a try :W
You see, until I finalize my RS225, W15CH and OW1 I can dream all I want. I've got a pair of W22 that will need a home in a new speaker too- 4 ways are just way too complicated.

Jed

[Jed]

Well, I just ordered some RS225s. I'll let you all know how they sound in place of the W22s below 300HZ in my 3-way speaker. The best thing is they only cost 1/4 the price of the W22s and are a PERFECT drop in replacement for this application. I think I've used the free shipping from parts express about 4 times in the last month or so!

Jed

[JonMarsh]

I'll be really surprised if you don't like them pretty dang well below 300 Hz. I've worked with 7" and 8" Scanspeaks, too, and I think the RS225 would be fairly primo even if they cost twice what they go for.

[Jed]

Jon or anyone,

How low can the C44 be crossed over LR2? Will they go below 1K and still have less distortion than a W22? Jon, you mentioned you have used the C44, would you mind sharing more details about that design?

[Jed]

Well, I got my RS225s today and the frame is too big for my holes that I have recessed for the W22s. Unfortunately this will take me a lot more time to work out than I had originally planned. The specs for the frame size from partsexpress are inaccurate, so that threw me off--- by an 1/8".

Any quick tips on how to enlarge my recessed openings to allow the RS225s to fit nicely? I've seen people put a temporary piece of wood in the cutout opening and then use the jasper circle jig. That's probably what I'll end up doing, even if it will be a pain.

Jed

[cjd]

Template is more accurate - especially if you have veneered speaker faces, I would go this route if you can.

Also, just 'cause I'm curious: Would the B-G Neo8 serve as another midrange alternative in the original scenario?

C

[JonMarsh]

Jon or anyone,

How low can the C44 be crossed over LR2? Will they go below 1K and still have less distortion than a W22? Jon, you mentioned you have used the C44, would you mind sharing more details about that design?

800 Hz LR2 would be the limit, IMO, but I'd recommend B3, or even LR4 (well, modified, a Bullock style- slight axial bump, flatter power response). I used B3 in one case, LR4 in another. 800 Hz for both.

[Jed]

Brian, Have you made a decision yet about your driver selection?

One thing that I am curious about is whether or not it is worth having a dedicated midrange for the range from 900-4K. This to me at least, seems like it should complicate things. But on the other hand, if there is no crossover in the critical 2K-5K range, I'd imagine the speaker would sound better. Which begs the question, should one then just use a robust tweeter and cross over around 1.5K to a larger midrange/woofer. I'm starting to see the light here, Jon. The only disadvantage I can see with the robust Millennium tweeter is off axis response etc. I wonder how important that is above 10K anyway.

[JonMarsh]

Brian, Have you made a decision yet about your driver selection?

One thing that I am curious about is whether or not it is worth having a dedicated midrange for the range from 900-4K. This to me at least, seems like it should complicate things. But on the other hand, if there is no crossover in the critical 2K-5K range, I'd imagine the speaker would sound better. Which begs the question, should one then just use a robust tweeter and cross over around 1.5K to a larger midrange/woofer. I'm starting to see the light here, Jon. The only disadvantage I can see with the robust Millennium tweeter is off axis response etc. I wonder how important that is above 10K anyway.

An interesting point. Depends a little on your hearing and your room.

-Question- have you ever heard a Millenium Excel in a system? It's probably the easiest tweeter from Seas to voice and get to sound quite good, doesn't take much of any tuning or tweaking- about as easy as a ScanSpeak SS98000. It's probably my favorite soft dome tweeter, though I'm not a major fan of that class as a whole. But then I'm not a major fan of many metal dome tweeters, either.

The one I really want to do some testing and auditioning with soon is the H1283. That might answer the "off axis" thingie, but then I have to have a very good midwoofer or midrange to use with it. They're already on order, too.

~Jon

[Jed]

An interesting point. Depends a little on your hearing and your room.

-Question- have you ever heard a Millenium Excel in a system? It's probably the easiest tweeter from Seas to voice and get to sound quite good, doesn't take much of any tuning or tweaking- about as easy as a ScanSpeak SS98000.
~Jon

Nope- haven't tried the Millennium yet. It was one that I was considering, though. I've heard the cheaper 1" seas tweeters- not to say that they are in the same league. I wouldn't want to offend anyone.

I just got back from listening to the MHT12 tweeters in an 3-way Eton setup. The etons are really mellow sounding- not nearly as detailed as the excel mag., but man the MHT12 is a detailed tweeter with a lot of space around the instruments. However, the MHT12 has a LOT of 2nd order distortion at 90DB according to Klang and Ton. It's pretty much under .3% from 3.5K-20K, but there is a huge spike at around 5-6K (1.8%) when played loud. I wasn't listening to levels that would excite the peak, but my question is, say I used a C44 crossed over to an MHT12 tweeter at around 5-6K, would it solve the distortion issue? Of course then the driver spacing CTR to CTR distances might become an issue too. Does the slope of the crossover effect the CTR to CTR distances that are "allowable". Since the MHT12 is a membrane, what constitutes the center?

[TacoD]

Interesting considerations, I think you can use almost every tweeter if you cross that high. Why not use a 19mm dome? Those new wide surround tweeters of Seas looks interesting. [oops Jon is recommending the same tweeter..., I didn't look up the number at Jon's post... ]

And the radiation patern will be better. Those MHT12 are bundeling at high frequencies.



I think center of the drive unit is already a bit rough, because the acoustic centre of the units are not laying on the baffle itself most of the time.

[Jed]

Interesting considerations, I think you can use almost every tweeter if you cross that high. Why not use a 19mm dome? Those new wide surround tweeters of Seas looks interesting. .

Good points Taco. I like the Hiquphon quite a bit and use it in my system currently. However, the MHT12 has so much delicacy and ease of reproducing the treble that I wonder if any dome can match it. When I first looked at the "clustering" of frequencies, I was a first disappointed since it does not compare to the even dispersion of a 19mm dome. With the MHT12, the dispersion seems better horizontally than it does vertically, which is in alignment with most tweeters constructed like the Visaton MHT-12 (ex. most ribbons). Does that mean it should not be considered in a system? Perhaps this is too great a compromise for some and there are better true ribbon tweeters out there.

[JonMarsh]

Nope- haven't tried the Millennium yet. It was one that I was considering, though. I've heard the cheaper 1" seas tweeters- not to say that they are in the same league. I wouldn't want to offend anyone.

I just got back from listening to the MHT12 tweeters in an 3-way Eton setup. The etons are really mellow sounding- not nearly as detailed as the excel mag., but man the MHT12 is a detailed tweeter with a lot of space around the instruments. However, the MHT12 has a LOT of 2nd order distortion at 90DB according to Klang and Ton. It's pretty much under .3% from 3.5K-20K, but there is a huge spike at around 5-6K (1.8%) when played loud. I wasn't listening to levels that would excite the peak, but my question is, say I used a C44 crossed over to an MHT12 tweeter at around 5-6K, would it solve the distortion issue? Of course then the driver spacing CTR to CTR distances might become an issue too. Does the slope of the crossover effect the CTR to CTR distances that are "allowable". Since the MHT12 is a membrane, what constitutes the center?

I'd be curious what the dispersion versus frequency looks like on that- and how uniform the surface excitation is. When I see a spike in distortion, my first suspicion is that you've hit a frequency that is at 1/3 the first diaphram breakup mode, and are exciting that, and seeing the "amplification factor" of the resonance. The more critical thing might be whether that causes an increase or broadening of the IM distortion sidebands- that's probably more audible.

There's some effects open to debate about spacing- if you don't have line sources, then it may be arguable that the diaphram radius should be added to the distance- what we're concerned here with is the radiating size of the combined drivers in the crossover region. For small drivers, center to center describes things well enough for a first order approxmation. If it's a planar driver, then things are more complicated.

Me, I have high hopes that the H1283 is a real "sleeper" in the fold, if the THD looks good from 2.5 kHz and up- say, at least as good as an H1212. Combine that with the excellent dispersion, no phase diffuser, and breakup mode at 35 kHz, and we could have a winner for the right matching midrange.

I'd like to do a smallish speaker that plays "out of class" in performance, and I'm leaning more and more to the idea of a compact three way similar to a Gradient Revolution, but unfortunately not conincident drivers. Dual RS315 on bass. C90/6 on midrange, and possibly H1283 on treble. Maybe an additional bass driver/filler with the RS315s (possibly an RS265HF), as I don't want to take the RS315HF up too high, and I'd like the tweeter/mid assembly to be closer to ear level. I have several weird ideas about crossover, but don't to talk about them until I have a chance to play around in LspCAD. Think something like the Natalie P only a bit weirder.

This will be a passive crossover, probably with a passive in line equalizer with 6 dB gain loss. (balanced and unbalanced; it's already designed, or rather, stolen from another project )

~Jon

[Landroval]

I'd like to do a smallish speaker that plays "out of class" in performance, and I'm leaning more and more to the idea of a compact three way similar to a Gradient Revolution, but unfortunately not conincident drivers. Dual RS315 on bass. C90/6 on midrange, and possibly H1283 on treble.

Is your idea to go with dipole bass and acoustic resistance enclosure for the mid, or did you have something else in mind?

[Jed]

C90/6 on midrange, and possibly H1283 on treble.
~Jon

I guess I'm trying to justify the use of a C44. They seem like really interesting drivers, but the C79 seems a bit more practical. I've found that practicallity doesn't always equal good sound though.

Some configurations I'm "thinking" about are:
RS225s to 900hZ, C44 to 4-6K, MHT-12, or H1283
RS225s to 300hz, W22EX to 1K, C44 to 4-6K, MHT-12, or H1283
RS225s to 400-600hz, C79 to 3K, H1283 or C13 or OW1

I'll see if I can dig up the info for the MHT12.

Jed

[JonMarsh]

The C79 has a wider range. The C44 is much easier to work with, and less expensive.

Hey, challenges are fun, right?

[JonMarsh]

Is your idea to go with dipole bass and acoustic resistance enclosure for the mid, or did you have something else in mind?

Dipole bass, and I'll probably experiment with both an acoustic resistance/cardiod and a full dipole midrange. Because of the intended crossover point, and my experience going full dipole to ~1 kHz, I'm a little reluctant to commit to a cardiod style midrange, as Gradient uses. But if I find that works the best...

OTOH, I mistrust instinctively combining different dispersion patterns at different frequencies- like a cardiod bass in a U baffle combined with a dipole midrange. I understand the reasons folks do that to cut the work load on the woofer, but IMO it removes a lot of the gains, also. I'd prefer to see the off axis response out to 45-60 degrees track well from 100 Hz to 10 kHz.

Just my 0.02.