homegrown Seas coax

[capslock]

Ok, I am starting a new thread to avoid further creep in this one:

https://www.htguide.com/forum/showthread.php?t=17418

[capslock]

My first question would be, how do you remove the phase plug?

Easy: remove the sticker on the back plate, use a crossed screwdriver (PZ2, not Philips!) to remove the screw. It goes all the way through the pole piece hole into a thread in the plug. The plug has a recess and sits on a copper tube of about 2 cm height that sits in a similar recess of the top plate. While the plug has is not really needed, I suggest putting the copper tube back, because it doubles as the upper shorting ring.

Are you sure that little Audax would work? The W18 with its higher allowable XO might be better suited for this.

All the little TW025 tweeters have amazingly low distortion, still, without loading, I'd not want to use them below 2-3 kHz. But as I said, building a more underhung motor is not rocket science. And then there are the new Seas Nd tweeters, the OD of which is still a mystery.

What makes you say the W18 has a higher allowable XO point? Both W22 and W18 have exactly the same break up mode at 4.8 kHz, the only difference being that the W22's is more pronounced.

Looking at CSD's from HobbyHifi, both are pistonic up to more than 4 kHz.

If you want to stick to 1.4 kHz with the W22 because of beaming, ok, I can understand that (but then the W18 won't take you much higher).

If it is because you want to avoid amplifying the 3rd harmonic, well, isn't that pretty arbitrary? First of all, the 4th and 5th harmonic may sound as nasty. Secondly, harmonic amplification may look ugly in the HD graph, but I doubt it sounds too bad because it does not intermodulate. Thirdly, harmonic generation is directly linked to cone excursion. The W22 has twice the cone area of the W18, so in order to reproduce the same level of midrange signal, it'll generate less harmonics, but these get amplified more because the 4.8 kHz FR peak is more pronounced, so the outcome may be the same.

Actually, and this might merit another separate thread, if one could further linearize the W22 or W18 (they share the same motor and spider), they might well be usable to 4 kHz. I was thinking along the lines of
a) adding Faraday shielding in the unshielded area of the T-pole
b) using a counterwound coil on the pole, which would have the same effect
c) using inductive feedback from a linear sensor attached to VC support or cone

You can look at what SEAS (KEF technology) did with the H1333, a small "horn" around the tweeter leading into the (cone)guide. Note also the inverted surround. If you were feeling brave after removing the phase plug, perhaps removing the rubber surround, inverting it, then gluing it to the back edge of the cone - may not seem too irresponsible :W . Then using a thin layered ring (felt?) to gradualy transition the cone (edge) out onto the frame/baffle to reduce reflection/diffraction.

Removing and reglueing the surround can be done, especially on Seas metal cones (the glue absorbs water and expands, becoming easier to detach, and the cones can withstand water). I suggest practising on an L18, though (which is identical with the exception of the cone material and profile and the absence of the shorting rings).

However, the surround usually has an assymetrical compliance which may have been partly compensated by prestretching it and the spider. So reversing the surround may give you a more assymetrical compliance profile.

But do you really think that the small (maybe 4 mm) hump of the surround makes a huge difference?

[capslock]

[QUOTE=AJINFLA]I swore I saw somewhere that Klang&Ton had tested it (H1333). It's fuzzy, but I recall seeing that it tested well. Perhaps better than the Thiel coax. Perhaps one of our European members can shed some light on this.
QUOTE]


Yes, they tested the T18RE, but it was the old version (H1144). They commented that the FR was much smoother than in the predecessor version with the old basket, and that the new tweeter had much lower distortion. However, distortion of the XPC/TPX cone driver was still worse than that of the L18, let alone something with shorting rings like the W18.

Madisound carriy three coax drivers with the TPX cone. The H1333 seems to be an improved version of the H1144, but other than the color of the surround, I find it hard to say what has been changed. FR of the cone looks slightly better to me.

The H1353 seems to be a custom model, no shielding but larger magnet. Not a bad idea...

[capslock]

I haven't read the the test, nor even seen the H1333, but I'm currently using a Seas coax with TPX cone and metal-dome tweeter. It's a custom made model for another Finnish speaker brand Chorus ( http://www.chorus.fi/ ). It's used in the Mondo series and also sold as a DIY kit. I've always liked coax-speakers, and this one is one of the nicest in addition to the Gradient models, and some Tannoys.

How much is the diy kit? Or can you get the driver separately from them? I went to their homepage, but Finnish is the one European language I find really impossible to read (although I also find it hard to grasp more than the general subject from Polish or other slavic texts).

[Landroval]

How much is the diy kit? Or can you get the driver separately from them? I went to their homepage, but Finnish is the one European language I find really impossible to read (although I also find it hard to grasp more than the general subject from Polish or other slavic texts).

The DIY-kit is 135 euro and it includes the driver, crossover and binding posts. The driver can be bought separately for 120 euro and it goes under the name T17RECOAX/TVAG-EH. There is also a smaller version with the same metal-dome tweeter and poly cone, 100 euro EP14RECOAX/AG-EH.

They can be bought here:

http://www.hifitalo.fi/

Rakennussarjat (DIY-kits) --> Audiokit --> AW-10 Coax
Kaiutinelementit (Drivers) --> Audiokit by SEAS

[capslock]

The DIY-kit is 135 euro and it includes the driver, crossover and binding posts. The driver can be bought separately for 120 euro and it goes under the name T17RECOAX/TVAG-EH. There is also a smaller version with the same metal-dome tweeter and poly cone, 100 euro EP14RECOAX/AG-EH.

They can be bought here:

http://www.hifitalo.fi/

Rakennussarjat (DIY-kits) --> Audiokit --> AW-10 Coax
Kaiutinelementit (Drivers) --> Audiokit by SEAS

Thanks! So it would be the old 14 and 17 cm baskets. Wonder if the tweeter is already of the new design used int he T18RE.

[Landroval]

Wonder if the tweeter is already of the new design used int he T18RE.

I dont know what you mean by new design, but it definitely isn't the same tweeter. The retail models all have coated fabric-domes, while these are metal.

[capslock]

The fabric tweeter of the T18 has much lower distortion that that of the T17. The comment in Klang & Ton at the time was that the motor design had been improved. Seas ususally make one motor for different domes, so your tweeter could use the old or the new motor.

[AJINFLA]

Eric,

thanks for starting this new thread! We were really drifting asea on the old one Thanks to our kind hosts on that one.

Easy: remove the sticker on the back plate

Nothing's ever that easy for me :roll: . I epoxyed pucks onto the back of mine for bracket mounting a long time ago.

Image not available
Time for some drilling . I'll be very careful of course.

What makes you say the W18 has a higher allowable XO point?

Off axis. Not the breakup mode. I'm a huge believer in smoothness in the (way) off-axis. Matching directivities of the drivers at XO. Tis why I have used waveguides in my (non-coax) designs. The next logical step for me in my journey was coaxial or even coincident driver arrangement. I've been accruing coaxials for the past several years from many maufacturers to see what they were thinking. JBL, Tannoy, KEF, P-Audio, etc.

Actually, and this might merit another separate thread, if one could further linearize the W22 or W18 (they share the same motor and spider), they might well be usable to 4 kHz. I was thinking along the lines of
a) adding Faraday shielding in the unshielded area of the T-pole
b) using a counterwound coil on the pole, which would have the same effect
c) using inductive feedback from a linear sensor attached to VC support or cone

That might be slightly above my neanderthal head. For sure I would destroy the driver there 8O . By all means, feel free to do so if you can.

Removing and reglueing the surround can be done, especially on Seas metal cones (the glue absorbs water and expands, becoming easier to detach, and the cones can withstand water). I suggest practising on an L18, though (which is identical with the exception of the cone material and profile and the absence of the shorting rings).

The Dayton RS180 would be my frankenstein there. $32 .

But do you really think that the small (maybe 4 mm) hump of the surround makes a huge difference?

Huge? No. Difference? Yes. Or more importantly, better. Have you seen the new KEF reference https://web.archive.org/web/20071209..._uniq/wave.htm or the TAD?



Think Andrew Jones.

Like I said, I love it when others spend a lot of research $ for me :T . The TAD driver represents the ultimate IMHO. Mounted open baffle.... ;x( .

Cheers,

AJ

[AJINFLA]

Madisound carriy three coax drivers with the TPX cone. The H1333 seems to be an improved version of the H1144, but other than the color of the surround, I find it hard to say what has been changed. FR of the cone looks slightly better to me.

The tweeters low end is also slightly more extended, perhaps allowing a slightly lower XO. Might be a bit smoother also. The overall improvements might be marginal, but without testing/listening that is purely speculation on my part.

Hi Landroval,

I recalled running into the Chorus site many moons ago when searching for waveguides. The site has changed quite a bit, but I see that they are thinking along the same lines as me. They have a coax now also. The prices you quoted seem quite reasonable

The DIY-kit is 135 euro and it includes the driver, crossover and binding posts

.
Do you know if the Gradient driver can be bought seperately?
Maybe SEAS will relent and give us an aluminum, or even magnesium cone and nice motor to go with the Alu tweeter? That would be very nice indeed.
The new SEAS neo tweets might be ideal for our type of DIY. I've been keeping an eye on this one also: http://www.diymobileaudio.com/forum/...ead.php?t=1666
Check it out Eric. Looks promising.

Cheers,

AJ

[capslock]

Epoxy softens when exposed to Acetone. However, wouldn't it be a shame to dismount those beautiful open baffle speakers? And new W22's are at a lot more affordable in the US than they are here.

OK, directivity is a valid argument. With feedback control of the cone motion, one can also go to smaller cone diameters without generating too much extra distortion.

Limiting factors though:
- motion of the waveguide unless XO point to sub is raised -- what kind of an amplitude is allowable?
- will be hard to place a tweeter inside the VC of an W15 or W12 (however, one might go to a 19 mm dome, considering the higher breakup node of the W12/W15 cone


By the way: while the mag cones have substantially cleaner CSD than their L-series counterparts or the RS drivers for 22, 18 and 15 cm, if one really wanted to go to the 12 cm category, the RS125 is every bid as good as the W12 CSD- and distortionwise, and the L12 would also be after the addition of shorting rings.


On the surround surgery: if we only need about +/- 0.5 mm excursion (due to large cone and say 300 Hz XO), a flat surround cut from a sheet of thin rubber or foam is a viable alternative.


Greetings

Eric

[TacoD]

I have seen a thread on madisound, where a Seas designer explained why you do not want to mount in a Alu/ Mg chasis. If you mount a tweeter in a metal cone, the tweeter can trigger some resonances of the metal cone. Main thing is, the designers at Seas had a very good reason for not using the W18.

[capslock]

I have seen a thread on madisound, where a Seas designer explained why you do not want to mount in a Alu/ Mg chasis. If you mount a tweeter in a metal cone, the tweeter can trigger some resonances of the metal cone. Main thing is, the designers at Seas had a very good reason for not using the W18.

Now that you mention this, yes, Björn said something like this, but I find it hard to believe, especially for Mg which has only one resonance (as opposed to Al).

There is a German high end manufacturer (one of the older and smaller ones) who uses a Mag cone with a tweeter inside the VC. Saw it at the high end audio fair in Neu Isenburg two years ago, just can't recall the name.

Also, things can be done about this single resonance. If it is a concentric mode (which it most likely is), gluing linear segments of rubber near the circumference of the cone will probably do the trick. This is similar to the linear segments Vifa cuts out of their cone edges. The idea is that you have a circular wave travelling outwards that gets partially reflected because of the mismatch of cone and surround. With linear segments, some parts of the wavefront get reflected earlier than others, so that the wave travelling back is not one circular wavefront.

[capslock]

The only two non-metal cones that I am aware of that have low distortion and good CSD are Peerless sandwich cones (as in the phase plugged HDS164) and the Usher 8137 (a Karbon / Aramide weave with pretty hard resin). There may be others around, but if so, the cone signature was swamped by an average motor.

[Landroval]

I recalled running into the Chorus site many moons ago when searching for waveguides. The site has changed quite a bit, but I see that they are thinking along the same lines as me. They have a coax now also. The prices you quoted seem quite reasonable .
Do you know if the Gradient driver can be bought seperately?

Now that I think of it, every succesfull Finnish speakerbrand is using mostly waveguides and coaxial drivers in their designs.

I believe Gradient might be selling some version of the coax they're using. Here's one guy's Prelude-like project using the G17RECOAX/TVA-GR:

http://www.ele.tut.fi/~artoko/audio/speakers/coaxial.html

[bidland]

Now that you mention this, yes, Björn said something like this, but I find it hard to believe, especially for Mg which has only one resonance (as opposed to Al).

There is a German high end manufacturer (one of the older and smaller ones) who uses a Mag cone with a tweeter inside the VC. Saw it at the high end audio fair in Neu Isenburg two years ago, just can't recall the name.

You are thinking about Audiodata?
They use a W18 coax. http://www.audiodata-hifi.de/produkte/partout.html

I have not seen the effects you describe here, but I will not dismiss this theory without investigating it. I would imagine that the benefits in improved polar response is more important anyway.

Bjorn

[capslock]

Well, when Taco mentioned the theory, it rang a bell (no pun intended). I thought this is what you said when I discussed my idea of converting an L18 to coax about a year ago, but it may well have been somebody else.

Anyway, I was indeed thinking of Audiodata, just couldn't recall the name.


Cheers,

Eric

PS: any info on the new tweeters to be presented at CES?

[Dennis H]

Welcome to HTGuide, Bjorn!

A question about the production H1333 coax. What has changed from the older version? I've seen the crossover you designed for the older version. Will it work equally well for the newer one?

Another question about the H1333's on-axis ripple in the top octave. Could the ripple be damped with felt or foam, either around the tweeter or around the outside of the woofer? I realize that's something you wouldn't want to do at the factory because it would look ugly but we DIYers don't care about such things.

[bidland]

The only difference from the H1144 to the H1333 is the surround material in the woofer. The H1144 use a very lossy material that is good for midrange use, but not so good for the bass performance. It's properties also change quite a bit with temperature changes. The H1333 use a newly developed surround material that has frequency dependant damping. Good damping in the midrange (but not quite as lossy as in the H1144) and less damping in the bass. The crossover is compatible.

The main cause of the axial ripple in the tweeter reponse is the woofer surround. Using a surround with a small inverted roll, like the TAD unit, will improve the tweeter response.

There is also some improvement to be made by applying some felt in the right place, but unfortunately I cannot disclose that at this point.

The new 27neo tweeters is made in a new chassis. Dual-chamber magnetsystem and a 4ohm v.c. with a bit more Xmax.

There will also be shown some other interesting drivers that will be available during the next months. Las Vegas is the place to be....

Me? I'm far from Vegas in -5degC, snow and 18 hours of darkness every day: Norway

Bjorn

[AJINFLA]

Greetings Bjorn,

thank you for your replies and welcome to our humble forum. Your input here is quite valuable. It would seem that one either has to decide whether the driver is to be used above say 300hz, where excursion is limited, or as a midwoofer (below 300hz) with its excursion demands. Each has its benefits of course. I will try both - since I like building loudspeakers.
I hate to tell you this, but here in Florida it is very bright and 62 degrees. Of course, we have hurricanes also....

Cheers,

AJ

P.S. love the (Mag) Excel drivers. Hopefully more to come :W

Image not available

[bidland]

For those of you willing to experiment I can give you some hints on how to make your own W18 coax. I can not really recommend doing this, because it will not be covered by any warranty.

Use a tweeter from a H1333 or H1144 and a W18E (or EX).
De-solder the tweeter lead-outs from the terminal before loosening the bolt holding the tweeter. Hold the tweeter carefully when loosening the bolt. It must not rotate. If it does, and the lead-out wires gets attached to anything, the v.c. wire to lead-out wire joint will break.
Remove the phaseplug as described by Eric. Turn down the Cu ring to 17mm to align the height of the tweeter to the cone neck.
The tweeter will fit into the Cu-ring, but don't assemble it until you have made a passage for the v.c. wire.

The v.c. wire is glued to the outside of the magnetsystem and soldered to the lead out wires inside the plastic "foot". The critical point is the edge where the Cu ring meets the magnetsystem. The wires will be cut if the Cu ring is pressed against this edge during assembly. So you make two small openings, 2mm wide and 3mm deep, to avoid contact between the Cu ring and the v.c. wire. Glue the Cu ring to the magnetsystem with 3 drops of superglue. This will ease assembly and make sure the grooves is aligned to the v.c. wire.

The next problem is the foot and the hole in the polepiece. A standard W18 has a 7mm hole in the pole, but the coax needs a 12mm hole to make room for the foot. Drilling a hole in a magnetized system is impossible, so forget that. However it is possible to cut off the sylindrical part of the foot, and that will solve the problem. The Cu ring will center the tweeter anyway.
But be very, very careful! If the flat, remaining part of the foot that is close to the magnetsystem comes loose, you will most probably break the v.c. wire inside the foot. Maybe a hot knife is good. You must not use any sideway force on the foot.

Inverting the W18 surround is not a good idea. The surround flange is moulded to match the cone angle, and inverting it will make a very odd shape and still have the HF problems.

I have not tried to make a W22 coax, but that would be interesting.
I did make a CA22 coax for a car last summer. It was very easy to filter. A single inductor for the woofer and only a cap and a resistor for the tweeter.

[capslock]

here's the little bugger

This is from a magazine review of the Audiodata Partout that uses a W18 cone and this little tweeter that looks very similar to the one used in the T18RE.

Maybe it can be obtained as a spare part? But I'd be even more interested in laying my hands on a pair with the very same motor design but an Al dome...

[bidland]

Eric,

We do already make a dual-chamber alu coax tweeter for some OEM customers.

I'd be glad to help you, but I hope you understand that I can not provide that service. If we should meet all requests from DIY'ers on forums like this and Madisound, I'm sure we could have enough work for two people full time. And just building two or three drivers at a time is not very rational, so it would be very expensive....

The Audiodata tweeter is without ferrofluid.


Bjorn

[capslock]

Yes, I understand. I had thought more about asking Madisound or Intertechnik for spares for the T18RE or Audiodata about their non-ferro version.

I have also been looking into magnet simulations for Neo tweeters. It really isn't rocket science! The center pole piece could be made from a sufficiently large mild steel washer, without much grinding or turning. Getting my hands on a suitable pot will be more difficult. I could turn it from a rod, but that would be a little hard on my small lathe. And good Audax and Seas replacment domes can be had from Madisound.

By the way, the field plot of the KEF coax that is on their homepage and was copied in the Mag vs. Nextel thread looks really suboptimal, for both tweeter and midrange gaps.

[TacoD]

Hmm it sounds difficult to make your own coax. Is it possible to use a Nextel driver. Or is the Nextel differently build?

Another question for Bjorn, is the Nextel comparable with the glasfiber units from the past? I liked the presence of those units ...

[AJINFLA]

Are you refering to the unit like what Gradient uses? I have never heard it myself.

Cheers,

AJ

[bidland]

Hmm it sounds difficult to make your own coax. Is it possible to use a Nextel driver. Or is the Nextel differently build?

Another question for Bjorn, is the Nextel comparable with the glasfiber units from the past? I liked the presence of those units ...

The W18NX use a deeper cone, but I think it will still be possible.

You could say they are comparable in the sense that they both are soft cones.
The Nextel coating makes the paper a bit stiffer, but it's a much more homogeneous material than glass fiber. The glassfiber cone has different properties in different directions due to the weave structure.

We have made a crossover for the W18NX and the Crescendo tweeter that came out really well. The sound character is very neutral and listenable, but not as transparant and revealing as with the magnesium cones. We just need some proof reading from our man in the US, and then the "Bifrost" plans will be released.

Bjorn

[Jonasz]

Would this THIS fit in a Seas coax-unit?

[AJINFLA]

Yes. A W26 or L26 coaxial . I've seen that driver before, looks interesting, but there would be issues for coincident use relating to the XT series itself. Unless its distortion profile (which I have not seen) is somehow radically different from its brethren. If I recall, its off axis was much better than the standard XT25, so I guess its possible. Although I doubt it.

Cheers,

AJ

[AJINFLA]

Someone else thinks all metal coaxials aren't a bad idea
From the new Thiel 3.7: the driver:

Image not available

Speaker:

Image not available

Cheers,

AJ

[RonS]

Also seen at CES:



HiVi cone mids with ribbon tweeters.

Apparently they will be available in different size cones.

[AJINFLA]

Thanks Ron, thats right up my alley. Any specifics? Link?

Cheers,

AJ

[RonS]

AJ, I grabbed the image off Stereophool or Soundstage, don't remember which. No specifics other than they have ribbon tweeters and will be available in different sizes. Maybe they'll come out with something in the 6.5-8" size al/mg cone, that would be cool for my dipole. No release date mentioned, I'll see if the HiVi site has anything.

[capslock]

The Thiel dome looks a lot like one of the Seas wide surround types, except it is not generously smeared with superglue like my 22TAF/G (H1283) that I received from Madisound last fall.

[Jonasz]

Here's another little gem I've never seen before, Scan-speak D2904-6100. It's outer diameter is only 36,5 mm. It's not a metaldome though.


Lousy info and a lousy pic

[TacoD]

Jon, you did a center with the D2904-6100???

[capslock]

In the new issue (2/2006) of Klang und Ton, they tested the T18RE /XFCTV2, i.e. the new version of the XP-membrane coax. The compared impedance curves to teh old T18RE, fs shifted from 42 to 30 Hz, and is also higher Q (now 45 Ohms instead of 24 Ohms). Was the old surround lossy enough to bring down Q_ms at resonance that much or are Seas now using a Kapton former (which would be a first, to the best of my knowlegde).

Tweeter is nice, 1100 Hz fs, medium Q (i.e. light ferrofluid), secondary impedance bump at 3 kHz, clearly visible in the waterfall.

2nd/3rd and 5th harmonic are -60 dB above 2 kHz at 85 db/1m. At 95 dB, 2nd is -50 dB whereas the other two remain at -60 dB.

[Dennis H]

Jon used the D2904-6000. They look about the same except the mounting flange of the 6000 increases the diameter to 52mm. That 6100 looks like a real winner for a DIY coax. I don't see it mentioned on the Tymphany page. Maybe hifisound bought out an OEM's stock of tweeters without the flange?

[Jonasz]

On the Tymphany homepage I can only find D2904/600001 and D2904/600000, so this must be a physically smaller variant of those?

[Dennis H]

Eric, how was the distortion of the woofer part of the T18RE /XFCTV2?

[capslock]

The 6100 can be obtained from Axel Oberhage who is the German distributor for Scan Speak. He has a couple of custom drivers that you can only get through him or dealers who buy from him. I don't remember seeing this driver on his pricelist the last time I looked (which may well be a year ago):

http://www.scanspeak.de/indexE.html

[capslock]

T18RE woofer:

85 dB:
2nd: -30 dB @ 100, -50 @ 200, - 60 300 - 5k
3rd: essentially -60 300 - 5k but -45 dB at 1 k with broad shoulders 400 Hz to 2 k (this is probabla Le modulation)
5th: -60 dB 100 Hz - 5k

95 dB:
2nd: -50 dB 300 - 5k but more wiggly than at 85 dB
3rd: -40 @ 1k, more like -60 at 300 and above 3 k
5th: -70 from 150 up with some wiggles

CSD: nice but not perfect

[AJINFLA]

I have seen a thread on madisound, where a Seas designer explained why you do not want to mount in a Alu/ Mg chasis. If you mount a tweeter in a metal cone, the tweeter can trigger some resonances of the metal cone. Main thing is, the designers at Seas had a very good reason for not using the W18.

Now that you mention this, yes, Björn said something like this, but I find it hard to believe, especially for Mg which has only one resonance (as opposed to Al).

I find that hard to believe also, since here is the new Pioneer (TAD design) speaker, Mag cone, Beryllium tweeter: http://www.pioneer.co.uk/uk/product_...oduct_id=11427



Eric, have you heard the Audio Physic Kronus? This is the non-coincident approach which can be used for a very wide variety of readily available drive units. This one does look interesting:



I still think that SEAS needs to oblige us with a Mag/Alum Excel coaxial, don't you? :W

Cheers,

AJ

[capslock]

First of all, to set the record strait, Björn didn't say that but neither Taco nor I can recall who it was.

Yes, I heard the Kronos two years ago at the High End Audio Fair (by the way, who's coming to Munich in June?). Wasn't impressed. Reviews (we have plenty of magazines than can be compared to Stereofool) were divided.

The midrange is a Seas L18, probably with an Excel motor attached to it. Even with their "active cone damping" (rubber O-ring round the circumference of the cone), it's hard to understand why they did not take a W18.

The tweeter is a 2904/7000. They have a felt ring around it, much as the other design I was describing, in order to reduce reflections off the cone. However, with the tweeter placement, they forego the potential time alignement as well as the loading effect. The 2904/7000 will produce plenty of second harmonic if stressed.

[capslock]

New issue of HobbyHifi is out. The German Vifa distributor is selling a coax made from a classic series M17 paper woofer minus the dust cap and a neo XT25SC. The Neo has a small horn-shaped faceplate. Its circumference is not circular but more egg-like, and it sits level with the surround. I'll try to scan and post a pic later.