Speaker Building Advice

Hello all; I'm looking for some friendly advice for my speaker building project...

My current and long-time “hi-end” speaker system consist of 17 year old Audio Physic Virgo III’s, which offer more inner detail and resolution, transparency, and pinpoint imaging compared to my previously owned mid-1980's Apogee Duetta’s. And while the Apogee’s could produce prodigious amounts of bass, it was not as tight as I desired (especially after hearing the Entec SW-1’s). I’m not a loud-volume listener, but because the Virgo’s have limited bass extension, I chose to run them in parallel with a pair of line-level late 1990's Velodyne 12 inch servo subwoofers which I can blend well with the main speakers in my relatively large listening room.

Years ago I thought I would pursue an improved DIY speaker project that would offer even more of the above qualities, using what were considered then to be some of the best speaker driver units available:

Scan Speak D2509/9900 Revelator tweeter
Scan Speak 18W/8545-00 mid-woofer
NHT 1259 subwoofer

Since I lacked the requisite measuring equipment, knowledge, and experience to design my own crossovers, I was going to use the 2-way Scan Speak Reference Monitor, Brennwald's ScanSpeaker, or Madisound designed crossover that had been employed for the Scan-Speak drivers, in conjunction with a pair of sealed subwoofers for the NHT 1259's powered by plate amplifiers. The general design would follow the Tony Gee Andromeda speaker, which I find quite aesthetically pleasing.

Now approaching retirement, I have been bitten by the idea of building these speakers once again. Pulling the drivers out of storage, I discovered they are pristine. But I have become aware that after 15+ years, significant advancements with speaker technology have been made. I therefore have the choice to:

1) pursue the project as originally envisioned, or

2) sell one, two, or all of these driver pairs and invest in an updated design, such as those from Troels Gravesen, who seems to have a very good reputation and obviously is both very knowledgeable and experienced.

However, I find the potential number of possibilities bewildering. Troels refuses to make any recommendations, and I don’t have the time or inclination to become a serial speaker builder. Other than frequency response and distortion graphs, and without the ability to audition a particular speaker, let alone hear it in the personal home listening environment, how does one decide between – for example – the Ekta Grande, Jensen 1071, Illuminator Monitor, Ellam FLEX paired with a suitable subwoofer, Ellipticor-1 with subwoofer, Ellipticor-3, or the ATELL-3?

Are the “classic” drivers I have still good enough that I wouldn’t hear a significant difference compared to the newer designs? Or given a specific budget, would it be best to pursue the most cutting edge drivers and designs one can afford, hoping for a significant and worthy improvement over what the classic drive units can provide? Given the few and far between reviews of DIY speaker projects by an independent reviewer (e.g. not the builder or kit seller), which direction should one move towards? I’m looking for suggestions or a methodology by which – ear-unheard – one can make any kind of informed decision.

Much thanks 🙂
 

Attachments

  • Scan Classics & NHT1259 sm.jpg
    Scan Classics & NHT1259 sm.jpg
    484.1 KB · Views: 660
  • Andromeda a.jpg
    Andromeda a.jpg
    49.1 KB · Views: 744
  • Scan-Speak refmon.pdf
    Scan-Speak refmon.pdf
    1.8 MB · Views: 359
  • NHT 1259.pdf
    NHT 1259.pdf
    209.4 KB · Views: 161
  • 18w-8545-00.pdf
    18w-8545-00.pdf
    679.3 KB · Views: 78
  • D2905-990000.pdf
    D2905-990000.pdf
    227.9 KB · Views: 83
  • Madisound 18W8545K-D29059900.pdf
    Madisound 18W8545K-D29059900.pdf
    344 KB · Views: 131
  • ScanSpeaker INTERESSANTE PROGETTO.pdf
    ScanSpeaker INTERESSANTE PROGETTO.pdf
    266.1 KB · Views: 208
"I wouldn’t hear a significant difference compared to the newer designs" - most likely
"ear-unheard" - part of learning about speaker design is training your ears... untrained, you're at the mercy of other's opinions
I'd say, just build them....
 
The two scanspeak drivers are still in production and very well thought of. They are used in some very spendy commercial speakers as well as some highly regarded DIY kits. I am not sure I would buy them new today, but that is a function of their price now being so high that they compete with some other very high end drivers. Since you already own them, by all means use them. I wish I had them.

The NHT 1259 was considered an excellent woofer in its day. When it went out of production, many folks regretted not getting one while they could. Even by todays standards, it is an excellent 12 inch driver for a sealed box design.

I recommend proceeding as you originally intended. Even if you don't want to use an existing kit design, there are numerous people on this site who will gladly help you with a new crossover.
 
Those tweeters are still among the very best available, even among Scanspeak's newer offerings. The midwoofers are excellent by today's standards, but there are better available in the Scan lineup as well as other brands (Satori, CSS, etc.). The NHT driver was great in its day, but I think you can do much better now. At the same time, if you don't plan on driving the NHT to high excursion levels, it should perform admirably.
 
Thank you gentlemen. I appreciate the input.

The consensus so far seems to be the existing units are all very good.

DiscoJones: I would be interested in hearing what better available midwoofer(s) you would be recommend.

Troels uses the ScanSpeak 18W/8531-G00 in the Jensen 1071, and the Audio Technology 15H52-06-13-SDKAM in the Prelude - either of which would eliminate the need to design a new MT crossover and are quite similar to the Andromeda. How good are these drivers compared to the 8545? Any other midwoofer recommendations for the critical midrange area that would work well with the 9900?

Thanks again.
 

Attachments

I think the ScanSpeak 18W/8531-G00 and the Audio Technology 15H52-06-13-SDKAM would both be upgrades from the 8545, as well as similar sized offerings from the Scan Illuminator line.. The Satori MW16, MW19 and CSS LDW7 would likewise be upgrades. I'm sure there are others, but these are the few that stand out at the moment. Again, the 8545 is no slouch, but I think the above mentioned drivers will take your performance up another notch.
 
I was under the impression that your 18W/8545 was one of the "sliced cone" drivers from scan speak. Now I see that it is not. I agree with DiscoJones: there are some better options available. To add to his list of excellent drivers, I would include the SB17CAC35 ceramic cone midwoofer (which is probably the least expensive driver under consideration here).
 
What people should be saying is:

1. It will be very difficult to improve upon what you already have.

2. You can't just pick a "particular looking loudspeaker" and insert drivers other than those designed for that loudspeaker and expect a decent result.

3. IF you don't want to become a loudspeaker designer, then either stick to what you have OR select a proven design and use those drivers.

4. IF you absolutely have to use your drivers and want the Andromeda aesthetic then you will need to be somewhat of a designer and have enough channels of amplification. This method entails using a digital crossover as mentioned here (in and "active" configuration):

Digital Crossover basics


..and well, there is "5" as well: you could pay someone to do the work for you (..which is probably what you would want anyway for your Andromeda cabinets).

As for the driver's themselves..

The tweeter is still excellent (though you can achieve greater clarity with other designs - particularly something like the Viawave SRT7), the midbass is OK (..and really it's more of a low-output bass driver and less of a midrange driver), the woofer is likely past it's prime. Even then, you would have to make sure that the tweeter's still have nearly identical performance (..and it's something you'd have to test), if you wanted to use them in a "passive" crossover design.


Sometimes my realist perspective is a bit of a "buzz-kill". 😱
 
Last edited:
Very appreciative of the information coming forth, so thanks again to all for your time and interest. I do support the "realist perspective," but I also have to be pragmatic regarding my priorities, knowledge, and related constraints. No buzz-kill - just mandates more exploration, and of course more expense ;-)

It seems a newer mid-woofer is in order, and if I move forward with the Jensen 1071 design the specified 18W/8531-G00 would seem to be it.

While I appreciate an attractive speaker because it is also de facto living room furniture, I'm not married to the Andromeda design per se. But I am married to a woman who finds the Troels 1071 version (first attached pic) unacceptable unless it is in the garage - where I probably will end up living if it is built as he shows. Fortunately there are a number of other modified 1071 cabinet builds Troels has linked to which have a better WAF (attached pics).

I'm intrigued by the reference to the Viawave ribbon tweeter, and have had some experience with a ribbon's smoothness via the Duetta's. Troel's ATiRi build (picture attached) uses one, along with an AT 18H52 mid-woofer and passive radiator. Judging by his description and exuberance, this design seems very worth exploring. Unfortunately it looks like there are no crossover component values that accompany the design 😕 and the Viawave tweeter(s) look to have limited availability, so it seems here in the USA I might have to get the whole-shebang kit overseas from Jentzen.

As another alternative when considering a new mid-woofer, looking at the wooffer Troels uses for the ATiRi, I stumbled on to the Jeff Bagby (RIP) Spirit Wind (picture attached). This speaker uses the D2509/9900 Revelator I have, and it has a good WAF, therefore another viable possibility. But again no crossover details and a kit purchase is apparently required to get one.

And it seems no matter what "6-1" two-way I may choose, I will end up needing an independently powered pair of sub-woofers for the bottom octave - so potentially more audio-nervosa regarding the 'past it's prime' 1259 may set in... But that appears a topic for another forum.
 

Attachments

  • ATiRi-1.jpg
    ATiRi-1.jpg
    275.3 KB · Views: 278
  • jensen1071-ziggy-4.jpg
    jensen1071-ziggy-4.jpg
    29 KB · Views: 208
  • jensen1071-yetkin-5.jpg
    jensen1071-yetkin-5.jpg
    53.8 KB · Views: 470
  • jensen-1071-miguel-15.jpg
    jensen-1071-miguel-15.jpg
    72.1 KB · Views: 594
  • jensen_B02 troels .jpg
    jensen_B02 troels .jpg
    81.6 KB · Views: 827
  • SpiritWind2.jpg
    SpiritWind2.jpg
    151.2 KB · Views: 207
  • give-him-a-sedagive.jpg
    give-him-a-sedagive.jpg
    104 KB · Views: 207
I stand by my enthusiasm for the NHT1259. If it were offered today at a price of $250, it would be considered past its prime. If it were offered at $140 however, it would sell nicely.

A 20 hz Fs subwoofer with an Xmax of 13 mm 0-pk, an Sd of 491 cm^2, and a sensitivity of 89 dB/2.83V... what is not to like?
 
"Dammit Jim I'm a Paramedic - not an engineer!" 🙂

Really - thanks Jim, that's reassuring. Interestingly, the 1259 sold for about $150 in 1996 - which today is about $255 USD. The 1259 was used in the NHT 3.3 had a lot of great press when it came out and was generously made available to the DIY community, so it seemed like a great driver and a good value.

I'm curious, however. At a minimum it appears I may be getting a better mid-woofer to go with the Scan Speak 9900 Revelator (if not also replacing it with a Viawave ribbon tweeter and the new mid woofer driver design). Would the sealed 1259 with a Qtc of 0.7 match well with what I'm now considering? What would a newer $250 (or greater) subwoofer driver provide that the 1259 doesn't?

The older Velodyne servo subs used with my Virgo's are OK, but nothing I'm aware of beats my memory (long ago!) of the Entec SW-1's. It's three 10 inch servo drivers were incredibly fast, deep, and tightly controlled, and hideously expensive. Unfortunately these are impossible to find in working order, and apparently there's no DIY subs that I know of that try to emulate them. I've thought about checking out the Rythmic servo sub kits to attempt a clone...

Ultimately I'm hoping to find something that has close to the qualities of the SW-1's and can blend well with the M-WT speakers. But like I said, this probably needs to be on the Subwoofer forum after I nail down the rest of the system.
 

Attachments

  • Ready to go!.jpg
    Ready to go!.jpg
    301 KB · Views: 250
  • Entec SW-1 servo sub.jpg
    Entec SW-1 servo sub.jpg
    65.9 KB · Views: 285
And it seems no matter what "6-1" two-way I may choose, I will end up needing an independently powered pair of sub-woofers for the bottom octave - so potentially more audio-nervosa regarding the 'past it's prime' 1259 may set in... But that appears a topic for another forum.

Veleodyne are going to be considerably better than just about any woofer add-on you can think of.. (..unless they aren't working properly.)


I'm wondering why you are pushing for a 2-way *6" and 1" when you are trying to better the speakers you already have.. That's usually (almost always) a step backwards..

Really, you need to go all-out to do much better. (..though you can do different, with something much more efficient, OR perhaps a radial like the Pluto.)

In that "vein" ("all-out"), maybe look at Jon Marsh's ..err, Evil Twin's Ardent D w/ a Dueland-type crossover. The midbass's are hideously expensive and stick-out like a sore without some black grill cloths, but that should actually push past what you currently have:

Ardent D (aka Kurosawa Jr) Winter Camp - Page 7

(..and if you like this, better photoshop in the grill cloths before presenting the picture for WAF consideration.)


*btw, the new "darling" of the 2-way 6" world is the Purifi woofer.. (which looks a bit strange but is a very serious performer for it's size.)
 
Last edited:
I'm curious, however. At a minimum it appears I may be getting a better mid-woofer to go with the Scan Speak 9900 Revelator (if not also replacing it with a Viawave ribbon tweeter and the new mid woofer driver design). Would the sealed 1259 with a Qtc of 0.7 match well with what I'm now considering? What would a newer $250 (or greater) subwoofer driver provide that the 1259 doesn't?

A step up from the 1259 would be (in my opinion) the SB Acoustics 12 inch series.... such as the SB3475NRX-6. These are very low distortion drivers with 3 inch voice coils, plenty of power handling, lots of SPL capability. Very musical up to about 250 Hz or so. They would work great as a woofer or subwoofer.**

I use these as my woofers. In a 110 liter sealed box you get a Qtc of 0.7 and extension down past 25 Hz. I use them in a 70 liter box and utilize a Linkwitz Transform to get the equivalent of a 110 liter box (essentially trading excess electrical power for a smaller box).

Equally important for you will be the kind of processing and amplification you apply to the sub. If you go with the ScanSpeak 18W/8531-G00, you will probably want to cross to the sub somewhere between 90 and 150 Hz. I don't know much about subwoofer plate amps. If I were in your shoes, I would be tempted to use DSP and a class D amp to drive the woofers. Something like a Hypex FA501.

** I am talking about subwoofers for music. If you want a THX rated sub for movies which can recreate nuclear explosions and planets colliding, well that is a different kind of subwoofer than one which does Dvorak and Miles Davis.
 
Hi Scott,

Thank you for the link to the Ardent D! Looks to be a superb speaker, and with the help of a friend who is a professional cabinet builder, I believe it would be doable. Although it would be spending much more than I initially planned, this project will likely be a rest-of-my life speaker, so it’s not out of the question, and I will have to ponder this further...

I’m also intrigued by the Purifi woofer you mentioned. This woofer, and the Viawave tweeter(s), appear to have a flat response curves, low levels of distortion with smooth roll off’s. One of my audio ‘religious experiences’ was the first time I heard the Rogers LS3/5a BBC monitors. Forgive my naiveté, but from the little I know it seems that a very nice restricted LF 2-way “monitor” could be made using these drivers with relatively simple crossovers and suitable level matching. Thereafter, I could supplement these with a good LF driver and enclosure, or even repurpose the Velodyne subwoofer drivers and electronics into a new more suitable enclosure and possibly have a very good 3-way speaker system. Why would this 2-way 6 + 1 with a [servo] subwoofer be “almost always a step backwards?” This seems to be what several serious DIY speakers essentially are.


Hi Jim,

Thank you for the subwoofer information and crossover point information. I have become a little bit familiar with the Hypex amps from Troels designs, and I’m strongly considering these for 1259 or whatever LF driver I might select. These (sub) woofers would be for critical music listening only. I have a totally separate Infinity surround system with a single Velodyne 15 inch servo sub for home theater use.
 

Attachments

  • Annotation 2020-08-14 143626.png
    Annotation 2020-08-14 143626.png
    540.9 KB · Views: 222
Last edited:
FWIW, from my own testing the 18W/8545K00 (same as 18W/8545-00 and the original "18W/8545" as far as I know) is remarkably similar to other peoples testing of the 18W/8531G
http://feleppa.com.au/pics/speaker_imgs/plots/SS18W8545_RawHarm.png
vs
http://www.zaphaudio.com/6.5test/18W8531G-HD.gif

Note: Zaph's fundamental level being -10dB instead of 0dB, so all the harmonics appear 10dB lower on the axis than in my measurement.

8545 might even be considered to be better than the 8531G, as it is devoid of the resonance that the 8531G experiences at ~400Hz which causes a spike in distortion. 200Hz down the 8545 also appears a little cleaner.

Somewhat ironically, the Illuminator while still good is not in the same league as the above two drivers. Sometimes newer tech does not equal better than 20-30 year old belt and braces engineering.
Illuminator:
http://www.zaphaudio.com/6.5test/18WU8741T00-HD.gif

The 9900 tweeter is also still outstanding by today's standards.
 
Last edited:
Forgive my naiveté, but from the little I know it seems that a very nice restricted LF 2-way “monitor” could be made using these drivers with relatively simple crossovers and suitable level matching.

Thereafter, I could supplement these with a good LF driver and enclosure, or even repurpose the Velodyne subwoofer drivers and electronics into a new more suitable enclosure and possibly have a very good 3-way speaker system.

Why would this 2-way 6 + 1 with a [servo] subwoofer be “almost always a step backwards?” This seems to be what several serious DIY speakers essentially are.

As a 2-way it's still going to require baffle-step loss correction w/ a significantly lower spl.. The ribbon tweeter in particular will take a huge hit in efficiency. LOTS of resistance..

With a good active digital crossover you could probably out-do your current loudspeakers (..at least to a certain low freq. limit) with that driver combination (Purifi and Viawave). ..and you might get close to that with the driver's you have and it wouldn't cost anything extra to try it.

In any event, a good 3 or 4 or 5 way has certain advantages if done properly - mostly because you can use a driver that's better suited to its limited passband. Ex. the Morel midrange (as in the Ardent D) has a much lighter diaphragm (with typically better transient capability), while also having a smaller diameter that's better suited to off-axis at higher freq.s. (..which allows better integration with the tweeter - generally resulting in a more "spacious" result).
 
Last edited:
If you want to go full active, you might consider something similar to this

New active 3-Way, Hypex and SB

When I was designing this system, I strongly considered the 18W/8531G00 and the D2905-9700. I chose the drivers I did because I thought I could get equal performance at a lower cost, and this being my first speaker project in almost 20 years, I wanted to go in slow. My next build will with much higher end drivers.

I am enthusiastic about a project for you using the D2905-9900 and the 18W/8531G00. Combined with stereo sub cabinets, this could be a really outstanding system. It is true that the Purifi is the current darling, and the Satori Textreme will be the soon-to-be darling, but the 8531 is a world class, well proven (although expensive) mid driver.

If you have an amp you love, go passive on the upper crossover (hopefully 2 kHz or below with these drivers) but active on the subwoofer crossover. If you are ambivalent about your amp, consider going full active with Hypex Fusion Ncore amp
 
I stand by my enthusiasm for the NHT1259. If it were offered today at a price of $250, it would be considered past its prime. If it were offered at $140 however, it would sell nicely.

A 20 hz Fs subwoofer with an Xmax of 13 mm 0-pk, an Sd of 491 cm^2, and a sensitivity of 89 dB/2.83V... what is not to like?

From what I've read it's not really 89 db but more to 86 db. I too have a pair but have never put it to use.