Three year ago there was a newsflash that soon went silent: one hospital ER was quaranteened after two pneumonic plague patients seeking treatment came to BJ overnight from inner Mongolia province. I commented then, a bad turn and two million dead. A few weeks later -- Wuhan. During the ensuing on-again, off-again semi-lockdowns I picked up DIY to help distract the mind. A few of these, out of ~20 projects, I have already described in various random posts.
According to 'net estimates and models, BJ has just achieved peak daily increase and majority infection (indeed close to 2/3). This is truly an amazing feat for the 22-million population, well-innoculated but almost none of them ever exposed to live Covid, to all catch a 39C/102F fever within a short couple weeks. (My small VR software studio mavr3d.com, "home-working" for a month, confirms the 2/3 ratio including family members.) As late as 12/1, Dynamic-Zero was upgraded to require 48-hour (was 72) negative test to go out in public; then on 12/3 the entire policy was cancelled and most testing sites were quickly dismantled. The 8-month-long silent debate over "lying-down", and the "sit-up" excercise repeated over and over, finally culminated in a climactic "laid-down". With hindsight, it is clear that by 12/2 the city transits, elevators, offices, and apartment buildings had already saturated levels of Covid, beyond the daily tallies reported and followed diligently by everyone. It came as great relief, that one putative positive, in a 10-person mixed sample, will no longer cause your entire building to be "temporarily locked down" (happened to me twice, 11/16 then 19-20). I "escaped" to an unsafe suburb but had to go out 12/2 and 12/7. It was even greater relief, after mild symptoms 12/8-9 went away by 12/10, when high fever finally hit 12/11. Safe at last, statistically, from the current wave.
Here is the 17L 1.5m TLonken 8" Supravox that I hastily assembled 11/24-25, amidst Covid tests and mushrooming lockdowns just before my escape. The inexpensive ready-made 1.8mm MDF cab came with a 3-fold "line" ending in a slot. I added two verical plywood dividers (edged with gasket foam) plus a narrow slanted bridge to the existing labyrinth. (Egg-crate-foamed but not yet glued down.) The other unit left unchanged for comparison. A quick listen: got the bass I needed out of the TL; the netting diffused very high frequency a little but would not serve to protect the whizzer. Then I was out'a there!
According to 'net estimates and models, BJ has just achieved peak daily increase and majority infection (indeed close to 2/3). This is truly an amazing feat for the 22-million population, well-innoculated but almost none of them ever exposed to live Covid, to all catch a 39C/102F fever within a short couple weeks. (My small VR software studio mavr3d.com, "home-working" for a month, confirms the 2/3 ratio including family members.) As late as 12/1, Dynamic-Zero was upgraded to require 48-hour (was 72) negative test to go out in public; then on 12/3 the entire policy was cancelled and most testing sites were quickly dismantled. The 8-month-long silent debate over "lying-down", and the "sit-up" excercise repeated over and over, finally culminated in a climactic "laid-down". With hindsight, it is clear that by 12/2 the city transits, elevators, offices, and apartment buildings had already saturated levels of Covid, beyond the daily tallies reported and followed diligently by everyone. It came as great relief, that one putative positive, in a 10-person mixed sample, will no longer cause your entire building to be "temporarily locked down" (happened to me twice, 11/16 then 19-20). I "escaped" to an unsafe suburb but had to go out 12/2 and 12/7. It was even greater relief, after mild symptoms 12/8-9 went away by 12/10, when high fever finally hit 12/11. Safe at last, statistically, from the current wave.
Here is the 17L 1.5m TLonken 8" Supravox that I hastily assembled 11/24-25, amidst Covid tests and mushrooming lockdowns just before my escape. The inexpensive ready-made 1.8mm MDF cab came with a 3-fold "line" ending in a slot. I added two verical plywood dividers (edged with gasket foam) plus a narrow slanted bridge to the existing labyrinth. (Egg-crate-foamed but not yet glued down.) The other unit left unchanged for comparison. A quick listen: got the bass I needed out of the TL; the netting diffused very high frequency a little but would not serve to protect the whizzer. Then I was out'a there!
Last edited:
Very nice indeed. 
If it isn't a silly question though -what's a 'TLOnken'? A transmission line is a quarter-wave (or a half-wave) enclosure. An Onken is a specific type of vented box alignment (a descendent of the old Jensen Ultraflex bass reflex design) which uses a massive vent system arranged up each vertical side of the front baffle, with a total vent CSA normally somewhere between 85% - 100% of driver Sd. Eigenmodes are not part of an Onken's intended alignment: in fact it's usually advised to keep duct length down to avoid / reduce the buildup of unwanted 1/2 wave modes. Am I missing something, because there doesn't seem to be any functional connection between the two in the physics, and none at all in the names?
If it isn't a silly question though -what's a 'TLOnken'? A transmission line is a quarter-wave (or a half-wave) enclosure. An Onken is a specific type of vented box alignment (a descendent of the old Jensen Ultraflex bass reflex design) which uses a massive vent system arranged up each vertical side of the front baffle, with a total vent CSA normally somewhere between 85% - 100% of driver Sd. Eigenmodes are not part of an Onken's intended alignment: in fact it's usually advised to keep duct length down to avoid / reduce the buildup of unwanted 1/2 wave modes. Am I missing something, because there doesn't seem to be any functional connection between the two in the physics, and none at all in the names?
FYI/FWIW when the late, lamented JMMLC was deciphering the various French articles for me including the so called Onken ones he noted that the actual Onken brand speakers were small inexpensive 'FR' drivers in 'bookshelf' BR cabs using an Altec 8 inspired high aspect ratio vent, even sent me a photo that looks very much like the OP's, though like so much of my 'knowledge' it's currently? lost in two seriously damaged HDs.
In short, I think we do 'Onken' the brand a disservice for what is clearly just a Jensen Ultraflex variant by any other name.
Last edited:
Very interesting indeed; I stand corrected & many thanks for the background. I'll have to see if I can find any of the original French articles too (not that my French is up to much). That would also move Dave's many resistive-vent alignments closer to them than the Ultraflex type that Onkens are usually taken as being.
I'm also wondering if one of the various data recovery companies could do something for your HDDs. Worth us looking into to see if we can help you get at least some of your saved data back.
I'm also wondering if one of the various data recovery companies could do something for your HDDs. Worth us looking into to see if we can help you get at least some of your saved data back.
At the time I had a bone fide 'guru' that picked through both and the computer one was literally fused together and the BU he managed to retrieve most of the data from the added drives? (sections), but 'C' drive where all the MS, added programs, misc. software and any of its 'byproducts' such as MathCad and associated WSs and other programs like HR, etc., just flat wasn't there.
That said, fairly recently someone mentioned that if the discs were still intact and a matching case could be found that swapping them out sometimes completely recovered all data on so-called 'blank' drives, but haven't tried yet and being so old now.......
All those French article pdfs, my pms with him are on it too.
That said, fairly recently someone mentioned that if the discs were still intact and a matching case could be found that swapping them out sometimes completely recovered all data on so-called 'blank' drives, but haven't tried yet and being so old now.......
All those French article pdfs, my pms with him are on it too.
Thanks everyone for the "Onken" discussion -- please continue. When I worked on my first original DIY (HeilEve 7L 1m TL) I did peruse the online archive of historical French "Onken", as well as use the Onken online calculator to design/reconcile/justify the folded, long single-slot "labyrinth" cab that is ubiquitous in China. The origin of Onken seems obscure and convoluted. Although first documented in the West in the French designs, the word itself is Japanese for "audio research (lab)" and presumably referred to a specific well-respected research center. In any event, the single loaded slot design has always been more popular in Europe (e.g. JMReynaud, ELAC) than the US, except for subwoofers.
I do hope someone will take the trouble to simulate or measure a series of cabs extending from the single loaded slot without fold, then adding folds, to encompass approximately tapered lines terminating in the loaded slot port (i.e. TLonken). I lack the tools and experience for doing so, except of course I listen and compare as the line grows incrementally. (I did find a site with a large systematic variety of CAD cabs just like this but had trouble downloading the associated spreadsheets.)
As for nomenclature, all the ones in use seem confusing or easily confused with each other. And in China, "labyrinth" the catch-all term used to refer to complex folded horn (such as my Fidelio Lowther PM2A) but now cover slot-loaded too, sometimes (very inconsistently) labelled N, M, Z-or-7.
May I ask, at what aspect ratio does a BR become MLTL instead of "reflex" (technical definition of phase-inverting resonance)? And when does cavity become a "line"? A tapered line joining a loaded slot port is what, technically?
I do hope someone will take the trouble to simulate or measure a series of cabs extending from the single loaded slot without fold, then adding folds, to encompass approximately tapered lines terminating in the loaded slot port (i.e. TLonken). I lack the tools and experience for doing so, except of course I listen and compare as the line grows incrementally. (I did find a site with a large systematic variety of CAD cabs just like this but had trouble downloading the associated spreadsheets.)
As for nomenclature, all the ones in use seem confusing or easily confused with each other. And in China, "labyrinth" the catch-all term used to refer to complex folded horn (such as my Fidelio Lowther PM2A) but now cover slot-loaded too, sometimes (very inconsistently) labelled N, M, Z-or-7.
May I ask, at what aspect ratio does a BR become MLTL instead of "reflex" (technical definition of phase-inverting resonance)? And when does cavity become a "line"? A tapered line joining a loaded slot port is what, technically?
Last edited:
A tapered line joining a slot vent is a mass-loaded line. You should be able to do away with the slot for much the same results by increasing the taper ratio.
There isn't a hard & fast rule as Dave says; technically you can define the transition from a vented box to an MLQW as the point at which the eigenmodes alter tuning (either Fb or the alignment shape) away from what you would obtain under pure cavity resonance assumptions. In outright terms, most MLTLs fall into that mid-ground rather than having tuning purely defined by standing waves.
There isn't a hard & fast rule as Dave says; technically you can define the transition from a vented box to an MLQW as the point at which the eigenmodes alter tuning (either Fb or the alignment shape) away from what you would obtain under pure cavity resonance assumptions. In outright terms, most MLTLs fall into that mid-ground rather than having tuning purely defined by standing waves.
Appreciated, I'll check once I'm reunited with them post-fever/quarantine. The seller may have added glue to the base of cone?
Attachments
Didn't know this, thanks! Again, according to JMMLC it was a small shop that made inexpensive bookshelf size speakers with single drivers presumably (on my part) primarily for the strong foreign local presence, tourists of the '50s - '60s.
IME and later 'proven' using MJK's software the point where its axial length began shortening the vent (not audible initially) was at 36" with the vent at the extreme bottom, driver offset at 0.21 and barely audible at 42"/0.349 offset and 56" for what I call a true MLTL since it allows a larger vent for a given length (more vent gain BW) with a 0.349 offset = the most used 38" seated ear height.
addendum:
http://mh-audio.nl/Calculators/Onken.html
http://dbdynamixaudio.com/labyrinth-vent-bass-reflex-enclosure-calculator/
onken is written like this
http://mh-audio.nl/Calculators/Onken.html
http://dbdynamixaudio.com/labyrinth-vent-bass-reflex-enclosure-calculator/
onken is written like this
音研 (オンケン)
(It was not easy for me to "undo" the romanization; in China the double-slot Onken is named after a Taiwan company 志平 pronounced zhiping... Actually, onken.co.jp/aboutus and tokyo-onken.co.jp are valid URL, the latter being Tokyo audio communication research center.)
Last edited:
Cool! A decal of it to put on original concept Onken builds would make a nice finishing touch!
Good to know, thanks! Couldn't remember if JMMLC had listed a link or not.
You're welcome!
Yes, IIRC it changes based on the taper ratio and whether it's expanding or contracting, though never bothered to figure it out since all were 'voiced' in its intended app. I imagine a bit of research using HR will get the desired results and for those folks with the math skills can derive some (not so?) simple equations, but frankly, until the taper gets pretty extreme (10:1 in my case) the proven MLTL ones work just fine IME.
A member of China's bbs.hifidiy.net made a pair of labyrinths out of dresser cabinets, using the design below he found online. Another member reported that the drawing actually came from a book that he once owned, published in 1958 translated from Russian. Coincidentally, my 1.3m "remodel-in-place" TLonken follows the exact same folds.
库里科夫斯基，业余无线电新技术，人民邮电出版社 1958 (Khurikovski? Amateur radio new techniques, People's Telegraph Pub.)
库里科夫斯基，业余无线电新技术，人民邮电出版社 1958 (Khurikovski? Amateur radio new techniques, People's Telegraph Pub.)
Just read the revived old thread that covered my questions in great detail, thanks for the info and pointers therein. But the amateur-scientist in me is still not 100% satisfied. First a specific technical question: (at the listening chair) does the slot energy output and dispersion benefit significantly from its geometry -- does the slot act as a (short) line-source and focus its output at the listener's narrow profile ceiling-to-floor, "1-pi space" so-to-speak, whereas the driver cone is projecting bass in 4-pi space all around the speaker? Very close to the speaker (TLonken) this effect is obvious, slot bass dominates direct speaker cone, more so the lower the frequency (so not just due to mouth-area < Sd).
Then something more basic and perplexing. When I "computed" my first TL two years ago, I simply figured bass backwave was time-delayed while midrange was absorbed by the line. Simple vector-sum of partially out-of-phase bass (after line losses) with front direct bass, yielding a net-positive boost for wavelength 1.2X to 6X the line-length (for example 2X wave is 180deg delay of backwave so now in-phase; 1.2X and 6X wave respectively -/+60deg delay of backwave). Using GM's favourite 56"~1.4m line, this means 40-200hz will be boosted (120hz maximally, 60-180hz usefully). Then I read Martin's "Anatomy" and discovered the "quarterwave" (really 3.8-something not 4X) half-open-pipe resonance, in the case of 56" at 60hz and integer-multiples. How are these two phenomena related or mathematically correlated, or not??
Then something more basic and perplexing. When I "computed" my first TL two years ago, I simply figured bass backwave was time-delayed while midrange was absorbed by the line. Simple vector-sum of partially out-of-phase bass (after line losses) with front direct bass, yielding a net-positive boost for wavelength 1.2X to 6X the line-length (for example 2X wave is 180deg delay of backwave so now in-phase; 1.2X and 6X wave respectively -/+60deg delay of backwave). Using GM's favourite 56"~1.4m line, this means 40-200hz will be boosted (120hz maximally, 60-180hz usefully). Then I read Martin's "Anatomy" and discovered the "quarterwave" (really 3.8-something not 4X) half-open-pipe resonance, in the case of 56" at 60hz and integer-multiples. How are these two phenomena related or mathematically correlated, or not??
Belated drawing of the thread-title TLonken, 17L 1.5m-line converted from a very inexpensive but solidly-built (18mm MDF), ready-made labyrinth cabinet. Measurements are inexact and certainly non-optimal. The TLonken has served as a testing platform for "high-powered" (low Qts) 8-inch fullrange drivers including: Supravox, Lowther PM6A, Isred/MKHIFI drum-paper, and Fostex F200A....
During the Chinese New Year holiday I was "idle-fishing" and saw bargain-priced F200A and W300Aii, both models of course NLA/rarely show up/usually very expensive. Turned out to have suffered damage but still functional (a piece of F200A aluminum faceplate broken off and missing; one W300Aii dustcap reglued). So I spent my "2023 budget" and here's a brief summary of the fun I had.
I ran the W300Aii (Vas170L,Fs25hz,Qts0.28,93dB,sharp drop>2.5khz) straight-through (following johnk), over a steel can, paired with a vintage alnico bigbutt paper-cone 4" (15uF at 2.7khz). Power & passion; best piano -2 A /A /A+ (named John'k for now).
I notched the F200A (Vas100L,Fs30hz,Qts0.32,89dB,high shelf 3.4-4.5khz, dip 6khz) with 0.68mH|3uF at 3.6k as a compromise. Precision & dynamics; best orchestral -1 A /A+/A+ (named Fate as in F8).
Lowther PM6A (15ohm) was a real pain-in-and-between-the-ears. I had to double-notch (1.5mH|2.7uF|15ohm at 2.5k) -> (0.47mH|0.68uF at 8.7k) to get close to realistic instrumental voices. Of couse still a dry sound, but very clear and even relaxing (!); best plucked attack -3 A+/A /A-.
Isred drum-paper 8" with whizzer, not best at/for anything except saving labor & money. Maybe after more time (or effort) -3 A /A /A .
Supravox in the same TLonken for comparison: Realism & palpability; best voice -2 A+/A+/A-.
And the Eazy FE108eΣ sealed/baffle-less and aggressivly notched/shelf-filterd at 1.1khz (1.5mH|2.7uF|6ohm->1ohm), supported below by one relaxed-sounding vintage Grundig 12" (9mH 1.85ohm). Subtlety & authenticity; best for solo or small ensemble -2 A+/A+/A-.
(That's my demerit-based rating system for "realism" hi/mid/lo; perfect -0.)
During the Chinese New Year holiday I was "idle-fishing" and saw bargain-priced F200A and W300Aii, both models of course NLA/rarely show up/usually very expensive. Turned out to have suffered damage but still functional (a piece of F200A aluminum faceplate broken off and missing; one W300Aii dustcap reglued). So I spent my "2023 budget" and here's a brief summary of the fun I had.
I ran the W300Aii (Vas170L,Fs25hz,Qts0.28,93dB,sharp drop>2.5khz) straight-through (following johnk), over a steel can, paired with a vintage alnico bigbutt paper-cone 4" (15uF at 2.7khz). Power & passion; best piano -2 A /A /A+ (named John'k for now).
I notched the F200A (Vas100L,Fs30hz,Qts0.32,89dB,high shelf 3.4-4.5khz, dip 6khz) with 0.68mH|3uF at 3.6k as a compromise. Precision & dynamics; best orchestral -1 A /A+/A+ (named Fate as in F8).
Lowther PM6A (15ohm) was a real pain-in-and-between-the-ears. I had to double-notch (1.5mH|2.7uF|15ohm at 2.5k) -> (0.47mH|0.68uF at 8.7k) to get close to realistic instrumental voices. Of couse still a dry sound, but very clear and even relaxing (!); best plucked attack -3 A+/A /A-.
Isred drum-paper 8" with whizzer, not best at/for anything except saving labor & money. Maybe after more time (or effort) -3 A /A /A .
Supravox in the same TLonken for comparison: Realism & palpability; best voice -2 A+/A+/A-.
And the Eazy FE108eΣ sealed/baffle-less and aggressivly notched/shelf-filterd at 1.1khz (1.5mH|2.7uF|6ohm->1ohm), supported below by one relaxed-sounding vintage Grundig 12" (9mH 1.85ohm). Subtlety & authenticity; best for solo or small ensemble -2 A+/A+/A-.
(That's my demerit-based rating system for "realism" hi/mid/lo; perfect -0.)
Last edited: