Tang Band W6-2144 vs W8-2145 for Onken enclosure

I'm intrigued by hopefully building some speakers with 'Onken' style enclosures)(I'm currently leaning towards 'classic Onken' rather than the Planet10 alternate 'Fonken' -although I have the highest regard for Dave and his work).

I'd be open to either Multi-way or Full-range, but with a definite gladness to go full range if I find suitable drivers within my reach. And I think I have found two potentially-excellent candidate drivers for this: the Tang Band W6-2144 or W8-2145.

I've been using the 'Onken calculator' at HiFi Loudspeaker Design

The W6-2144 looks like it performs very, very nearly as well at lower frequencies as the W8, and the W6 manages to do so in a much more modestly sized enclosure. PDFs of calculations (from the above-mentioned onken calculator website) attached. The results I am getting from the online 'onken calculator' show the W6 an f3 of 33.8hz in a 60.8 liter enclosure vs the W8 with 32.5hz in a 100.7 liter enclosure.

Difference in price between the drivers $91.10 for the W6 vs $96.90 for the W9 is practically a rounding error, so prices of the drivers aren't a reason for me to lean towards one vs. the other.

My questions:

1) Can I reasonably rely on this online 'onken calculator?'

2) I don't yet have real depth of understanding (at least not nearly what I'd like) of how the multiple characteristics of drivers interact with different size and configuration enclosures - so- am I missing the boat here somewhere in what the calculator is telling me, both in relation to:
a) overall results?
b) the relative lack of benefit of going with the W8 in a bigger enclosure compared to the W6 in a more modest-sized enclosure?
3) Aside from "graphed frequency response" and a slight gain in db per watt efficiency in the W8 over the W6, is there any other reason I should consider going with the W8 over the W6?

4) similar to "3" but more narrow: if I use a single-ended triode tube amplifier with no negative feedback, would the W8 in a large enclosure present a better load and match to the amp than the W6 in a smaller enclosure?

I realized that I stacked a lot of 'ask' into this question, so thanks for plowing through it. and I appreciate and look forward to input. Thanks in advance!
 

Attachments

  • tang band w6-2144 design.pdf
    1.1 MB · Views: 177
  • tang band w8-2145 calculations.pdf
    1.1 MB · Views: 145
Technically, both drivers are unsuited for Onken loading due to too low a Qms [note the 2-16 limit in red].

All else is ~equal, so the higher Vas one [can] yield greater resolution [AKA 'inner detail'], which IME with high Vas vintage drivers Vs modern driver's generally much lower Vas makes it the tie breaker for me, misalignment or not. ;)

GM
 
Thank you very much GM for your input.

Most 'modern' drivers seem to model poorly in the Onken calculator, either flunking entirely or having disappointingly 'un-low' f3. These TBs were a refreshing exception/ The Visaton BG20 8" fullrange actually models very nicely _except_ that the cabinet volumes are huge, in the 160-180L range

With the two Tang Band (W6 and W8) that interest me, the QMS is below the recommended lower limit of 2 (for either of these speakers, with the W8 being 1.65 and the W6 being 1.73) but I don't have the experience or insight to really know how much that "beneath the lower limit" really does or doesn't really throw things severely out of whack? Your thoughts?
 
I like this box design
After you use the cal. How do you work out the inside of the port lengths of all of them?
Is there a page that goes over the side of the box?


What I have been doing is picking and trying different widths and heights and numbers of vents, and putting them into the calculator; it rejects unacceptable ones, and then, if acceptable, tells you the inside length to achieve resonance
 
You're welcome!

Right, this 'Onken' alignment is based loosely [mostly its looks] on the '50s Jensen Ultraflex series, but its cab alignment is Snyder's Vb/Vas*Qts^2 = 6.34, so as Vas shrinks, vents increase in area/length same as BR and with so many it aggravates the problem, so better to use just one big round one for peak vent efficiency.

Well, here's what all Qms represents, but the takeaway is "A higher impedance peak at resonance will translate in a higher Qms", i.e. takes plenty of electro-mechanical-acoustical power to drive a big vent system [scroll down to 'Qes, Qms and Qts']: Thiele Small parameters equations - How each one affects the others

No experience, but the implication is that when driven hard the motor may overheat around/at Fb, resulting in poor bass performance.

Onkens are popular among us vintage types, though I seemed to be the only naysayer other than its industrial 'look', so got a chuckle earlier today while browsing some new [to me] Western Electric [W.E.] sites and came across this: ElectraVolt: Western Electric TA-7331-A Baffle

I was sceptical of this [W.E.] design as it breaks with some of my expected notions. After assembly and before any finishing it was impossible not to have a quick listen. There were thoughts of the fire pit outside to which the several cabinets have been disposed of after testing, Onken and Ultraflex to name a few. Nothing like a glass of wine and a fire to celebrate the difference between a myth and a legend. It is amazing how large a cabinet this little pit can support while it burns and the Limousine rains clean up most of the mess... if the weather is inclement it can all be watched through six panels of glass door. Chin, chin.
 
Last edited:
Thanks again GM - some of the things I have been finding and reading about Onkens seem to indicate that all of the "modern onkens" (starting with Hiraga's work that has so strongly influenced nearly everything since) had been designed and used with the onken-cabinet-ed speaker being powered by a solid state amplifier (with high intra-amplifier damping) and then the higher frequency drivers had been powered by tube amplifiers (with less intra-amplifier damping). That may partly (maybe largely) explain why people are having disappointing results with the Hiraga-influenced recent decades' design- if they're powering them from the inherently less-well damped tube amplifiers. And I much prefer tube power amps so if that's a failing of the current Onken design assumptions, that's a total fail for my desired uses.

I wonder whether it could be possible to reverse-engineer the alignments and proportions of the original Jensen Ultraflex that is an ancestor to the Onken, just to get some additional insights; since the Ultraflex dates from the '50s, it'd definitely be pre-solid state.

I guess I also need to do considerably more learning about cabinet alignments, generally, something that I have only recently begun to read up on; I only relatively recently fully and clearly realized that there was anything other than a one/only alignment for a bass-reflex.
 
You're welcome!

Right, SS or ~1/8 ohm PP tubes, so with ~0.5 ohm wiring it's still a high enough damping factor [DF] to not obviously alter speaker response and of course if you use a bit lower Qts driver, then it 'washes' out. Yes, there's a certain synergy with SET tubes/Salmon family of horns.

Yes, you can reverse engineer any speaker given sufficient info, though TTBOMK the Ultraflex series were Vb = Vas/1.44 tuned to Fs, i.e. basically a prosound alignment that's the optimal trade-off between size Vs power handling with no consideration for how flat the response is since variable DF bass/treble tone controls were 'de riguer' regardless of the intended app.

Yeah, there's a bunch of possible speaker alignments since the range is from Fs to its upper mass corner [Fhm = 2*Fs/Qts'] and from Fs to its lower mass corner [Flc = Fs*Qts'/2] with the widest BW determined by the lowest practical Qts [strongest motor], which is around 0.1 AFAIK: Qts' = Qts + any added series resistance [Rs]: HiFi Loudspeaker Design

GM
 
Thanks to the guidance from GM, and some reading-up that I have continued to do building off from things I'm learning from GM, I've concluded that at least for any size cabinets or budget-for-driver that can 'fit' for me, trying to do a true 'classic Onken' is going to be chasing a mirage. The P10 'Fonkens' and their various variant progeny, seem to have great potential merit, but at least for the moment I've had my fill of onken-ish possibilities.

Parts Express did just recently have the TB W8-2145 8" fullrange drivers on sale for $75/ea, and I decided that (plus the free shipping) was something I've later kick myself for not seizing, so those are en route in my direction ...

Now trying to study up on other enclosure possibilities that might suit that TB driver. Liking the idea of something like a 'Pensil' (I like how it combines simplicity with sophistication of design/performance) _if_ that could be adapted to this driver instead of its 'native' Alpair ... or perhaps some other form of TL....
 
Last edited:
Now trying to study up on other enclosure possibilities that might suit that TB driver. Liking the idea of something like a 'Pensil' (I like how it combines simplicity with sophistication of design/performance) _if_ that could be adapted to this driver instead of its 'native' Alpair ... or perhaps some other form of TL....

You could if you knew Scott's design routine, but it's performance should in theory be [very] similar to this 36 Hz Fb MLTL I simmed back when these came to PE:

GM
 

Attachments

  • W8-2145_MLTL.txt
    1 KB · Views: 57
Technically, both drivers are unsuited for Onken loading due to too low a Qms [note the 2-16 limit in red].

All else is ~equal, so the higher Vas one [can] yield greater resolution [AKA 'inner detail'], which IME with high Vas vintage drivers Vs modern driver's generally much lower Vas makes it the tie breaker for me, misalignment or not. ;)

GM

Hi VAS gives detail? Never heard this before. Is this the same as transparency?
 
I’d always thought a high qms can show detail.

Basically a floppy suspension that doesn’t absorb motion or “stiction”.

Not as simple as that, I know.

What tracks a record better, an optimized tone arm/sprung cartridge or a heavy one with a nail? ;)

GM

I've never found t/s parameters all that helpful in telling how something will sound, at least at higher frequencies. Drivers do sound different even with a similar response. I'm not sure anyone knows why. A lot of nice tweeters have a weak "meh" sound, especially when crossed low in a two way. There might be something to the light cone big magnet cult. Going to a three way helps.
 
They're quite helpful, though only up to the driver's effective upper mass corner [Fhm] where T/S theory peters out, so no way it can predict any higher frequencies unless measuring open back mid and/or tweeters, otherwise you can only derive based on what its closed box impedance winds up being if you don't know its internal volume.

Fhm = 2*Fs/Qts'

Qts' = Qts + any added series resistance [Rs]: HiFi Loudspeaker Design

Lots of variables in driver design, but when BW limited to its [Fhm], they all sound near enough alike at low power if the full T/S specs are +/- 10%.

My mid '50s 15" Altec 515B [20-5 kHz] is known for having one of the most accurate, 'musical' performances [aka PRaT/pace, rhythm & timing], but even measured somewhat above its [Fhm] @ 500 Hz/2nd order sounds like a worn out washing machine of the era trying desperately to wash a very full load, i.e. a dull/flat sloshing sound, but add the mid/HF horn to complete the various frequencies and to many over the decades and still highly sought, it's audio nirvana in a H.E. system.

GM
 
Hi VAS gives detail? Never heard this before. Is this the same as transparency?

Correct as per my 'knee jerk' vinyl tracking comparison and 'PRaT'.

I view 'transparency' as top end air and around the instruments in proper proportion, which is all in the HF.

To wit, in 2k? at a Atlanta DIY Meet we auditioned a pair of TLs with Seas? drivers and not only were they a bit more 3D due to a very 'live' room, the mids, highs were so diaphanous/lacking in weight that I found it disconcerting to the point of leaving the room, though most around me found it a performance goal to 'die for', so could only assume they'd never been to a proper concert hall symphony.

GM
 
Lots of variables in driver design, but when BW limited to its [Fhm], they all sound near enough alike at low power if the full T/S specs are +/- 10%.

My mid '50s 15" Altec 515B [20-5 kHz] is known for having one of the most accurate, 'musical' performances [aka PRaT/pace, rhythm & timing], but even measured somewhat above its [Fhm] @ 500 Hz/2nd order sounds like a worn out washing machine of the era trying desperately to wash a very full load, i.e. a dull/flat sloshing sound, but add the mid/HF horn to complete the various frequencies and to many over the decades and still highly sought, it's audio nirvana in a H.E. system.

GM

Correct as per my 'knee jerk' vinyl tracking comparison and 'PRaT'.

I view 'transparency' as top end air and around the instruments in proper proportion, which is all in the HF.

To wit, in 2k? at a Atlanta DIY Meet we auditioned a pair of TLs with Seas? drivers and not only were they a bit more 3D due to a very 'live' room, the mids, highs were so diaphanous/lacking in weight that I found it disconcerting to the point of leaving the room, though most around me found it a performance goal to 'die for', so could only assume they'd never been to a proper concert hall symphony.

GM

I suspect a lot of modern drivers, tweeters in particular, are over damped for best frequency response and bandwidth. FR is still best objective measure. Some set ups like horns can have intensity to sound that is more real. Its like a very well braced cabinet won't measure better but will sound more life like and intense at low frequencies.
Dayton did a version of their 8" full range driver as a woofer. It has the powerful motor and light cone. Nice cast frame and begging to be crossed to a horn. But response would be a beast to tame probably because cone is less damped. Not sure if its selling well. But it definitely won't sound dead.

Dayton Audio PM220-8 8" Wideband Midbass Neo Driver
 
Last edited:
Cone/dome tweeters can vary from ~0.5 to at least the 2.21 Qts Peerless I used on some 'el cheapo' bookshelf speakers back in '77, though typically the more powerful/$$$ of course being down to ~0.4 Qts with the norm ~0.7-8 Qts.

Yes, properly done compression horn systems are more 'live' than almost any recording and why they've been my choice since age 6. ;)

Yeah, that Dayton looks like it 'rings like a ten penny nail struck with a ball peen hammer' ;) and unfortunately peaks/'snaps' right around its native XO point [~13543"/8" = ~1693 Hz].

FWIW, my Altec horn driver's response are near enough spot on to this plot [not surprising since both are of the same vintage ;)]: http://www.lansingheritage.org/images/altec/specs/components/515b/page02.jpg

It's designed for the 70 Hz FLH @ 500 Hz XO in my avatar. 210 dual driver cab response shows how well they blend. Note the ~150 Hz notch is due to the cab missing its baffle boards ['wings'] that pushes the [stronger] notch down to its ~35 Hz reflex tuning where it further damps the drivers [bottom response]: http://www.lansingheritage.org/images/altec/specs/components/515g/page3.jpg

GM