What you really need is the cone area, and I am doubtful, perhaps wrongly so - that you can get a more accurate number than the one Altec provided.
There should be some information at Lansing Heritage that should help you get a handle on Sd...
Petite onken
John,
I played briefly with the calculator using the t/s data I could muster online for the 414z. Same confusing results.
I'll be following your progress with interest!
MJ
John,
I played briefly with the calculator using the t/s data I could muster online for the 414z. Same confusing results.
I'll be following your progress with interest!
MJ
I was only able to work backwards from the value given for Vas in the Altec data and that minor correction just made things worse. The only way for me to get a reasonable slot dimension for 2 x 414C using that spreadsheet is by limiting the port area to 80% Sd. Designing a reasonable Onken enclosure for a single 414C from that spreadsheet is impossible.
John
Last edited:
If it doesn't, the SS is corrupted since increasing Rg will dramatically increase box size.
Ideally, you measure the T/S specs and either the program or you can calculate it more accurately than you can measure it, which BTW will vary depending on the equipment, accuracy of the measurement, so for the 414 it appears to be somewhere in the range of 505-511 cm^2, so call it 508 cm^2 on average if you don't measure BL, Cms, Mms, etc. which are all larger than Altec's 78"^2 (503.225 cm^2).
As I've posted many times though, there's no Onken 'magic' per se and following its guidelines exactly isn't required to get excellent results when using low Fs, Qts, high Vas drivers. As Vas shrinks though, the vents get too long due to being grossly over sized, so at this point it needs to be designed as a tapered TL (TQWT) to keep the huge vent area.
GM
Hello, my first post in three year of membership 
About two drivers in one box –
Simulated different configurations myself, as well with 2 drivers. Only meaningful thing to do so is to increase efficiency of the system, right? Then drivers should have parallel connection. Had no idea how Qms, Qes and Qts do change in case, so simulated the behaviour of parameters in another available software.
What turned out is (in order of their appearance in OC)-
Fs does not change
Rs divides by the number of drivers
Qms, Qes, Qts does not change
Moving mass multiplies by the number of drivers
Sd multiplies by the number of drivers
And system gets extra 3dB boost
While all changes are logical, I was quite surprised that values of Q stayed intact in every configuration I managed to simulate.
To give an idea how much Rg alter the dimensions -
Let's say, Rg is originally 0,3 Ohm in the spreadsheet, using all other needed data about my driver of use, box volume turned out to be 198,8 l.
Now doubling the Rg to 0,6 Ohm increased volume about 7,7 litres ie ca 3,9%
If instead multiplying Rg ten times to 3 Ohms, volume increased to 272,5 litres, ie increasement is ca 37%. Are those reasonably dramatical changes? I do not know.
About two drivers in one box –
Simulated different configurations myself, as well with 2 drivers. Only meaningful thing to do so is to increase efficiency of the system, right? Then drivers should have parallel connection. Had no idea how Qms, Qes and Qts do change in case, so simulated the behaviour of parameters in another available software.
What turned out is (in order of their appearance in OC)-
Fs does not change
Rs divides by the number of drivers
Qms, Qes, Qts does not change
Moving mass multiplies by the number of drivers
Sd multiplies by the number of drivers
And system gets extra 3dB boost
While all changes are logical, I was quite surprised that values of Q stayed intact in every configuration I managed to simulate.
To give an idea how much Rg alter the dimensions -
Let's say, Rg is originally 0,3 Ohm in the spreadsheet, using all other needed data about my driver of use, box volume turned out to be 198,8 l.
Now doubling the Rg to 0,6 Ohm increased volume about 7,7 litres ie ca 3,9%
If instead multiplying Rg ten times to 3 Ohms, volume increased to 272,5 litres, ie increasement is ca 37%. Are those reasonably dramatical changes? I do not know.
Yes, changing Rg does affect box size. I was more concerned by vent area and length, which are unaffected. The reason I started this discussion is I've seen single and double 414s in Onken boxes and but now I wonder how they were designed. Maybe they know something about the Petite Onken that I don't.
John
Last edited:
Greets!
Vas doubles also for a given alignment, ergo if you design for one driver, then doubling net Vb and number of vents yields the same response except +3 dB over the BW that's < ~WL/pi of the driver's spacing, so butting them together is a good plan. If you wire them in parallel, then you get another +3 dB due to halving system resistance (ignoring any series resistance).
Matching impedance's = driver Q doesn't change since they are the same at each measuring point and if they are different you get an averaged set of Qs WRT to what the box compliance 'feels'. The latter can be used to good effect with a high Q woofer and low Q mid: http://www.fortunecity.com/rivendell/xentar/1179/theory/dddllqd/dddllqd.html
Qes is the driver's relative motor strength and since it dominates Qts, it dominates box design, ergo the 'weaker' the motor (increasing Qes/Qts), the more compliant the box's air mass 'spring' (net Vb) must be to 'support' it which in turn means it must be tuned lower to get a maximally flat response.
GM
Vas doubles also for a given alignment, ergo if you design for one driver, then doubling net Vb and number of vents yields the same response except +3 dB over the BW that's < ~WL/pi of the driver's spacing, so butting them together is a good plan. If you wire them in parallel, then you get another +3 dB due to halving system resistance (ignoring any series resistance).
Matching impedance's = driver Q doesn't change since they are the same at each measuring point and if they are different you get an averaged set of Qs WRT to what the box compliance 'feels'. The latter can be used to good effect with a high Q woofer and low Q mid: http://www.fortunecity.com/rivendell/xentar/1179/theory/dddllqd/dddllqd.html
Qes is the driver's relative motor strength and since it dominates Qts, it dominates box design, ergo the 'weaker' the motor (increasing Qes/Qts), the more compliant the box's air mass 'spring' (net Vb) must be to 'support' it which in turn means it must be tuned lower to get a maximally flat response.
GM
As of petit or not, may be wrong, but my understanding is that just a name addition to differentiate onken box calculated for Altec's 414 driver, which turned out to be considerably smaller than original box that was meant to house 416 driver. While both were somewhat standard solutions, 416-onken with his 360l volume is normal onken, and 414-onken is just Onken Petit or Little Onken. There is no difference in calculator's logic.
Thing I dont get though is that several pics of double (416) onkens one can find in the net expose only two little port holes in front ...
Noting that this is only true with a true voltage source such as a solid state amplifier with feedback where halving the load impedance results in a doubling of output power. In the case of a tube amplifier where the amplifier is matched to the load by changing taps, and taking the example of two 8 ohm drivers in parallel you would then use the 4 ohm tap, and taking the square root of the impedance ratio (this being a transformer) you end up with the 0.707 X the voltage present on the 8 ohm tap and hence the same exact power as you had on the 8 ohm tap with a single driver. Just an FYI that most speaker designers don't seem to get. You still get the +3dB acoustical gain however..
Tube amplifier output power is relatively constant into any of the load impedances it is designed to match. A 60W tube amplifier does not become a 120W amplifier because you connected a 4 ohm speaker to it instead of an 8 ohm speaker. You change the transformer tap used, and power capability is relatively constant assuming an ideal load.
Last edited:
That is the Onken W, nothing to do with the Jensen Ultraflex inspired box that is so well known.
A wonderful bass box, BTW. Better than the Ultraflex version, but also much bigger and heavier. (I think it was double 515s, but will check.
..
I admit I was unaware of the Onken per se until joining the on-line audio community in '96, but had long since built a variety of Jensen Ultra-flex alignments which were based on the original 'Onken', A.L. Thuras's original reflex patent # (US) 1869178.
General design details were posted on the old basslist, so having a 416-8C lying around and it being a bit different from the way I knew how to design such cabs, I cobbled one together and wasn't too impressed compared to my TL loaded tower vented alignments (now MLTL, MLTQWT) only to find out later that I was missing some key formulas that to my way of thinking would have made it perform even worse.
By then though, conflicting info WRT design details had been posted, so made a basic Excel SS with some questions and beseeched Jean-Michel Le Cleac'h to critique/correct it since he seemed the most knowledgeable on the subject, which he graciously did. I'd made some some wrong assumptions plus there was yet a few more details he added that AFAIK hadn't been posted, but by the time I'd just about finished a ~user friendly SS to post though, Marc posted his, so for a variety of reasons never did. The main difference being I changed it to input Vas in lieu of Mms, Cms and Sd in cm^2 since these are the published norm. They aren't always accurate though, so a good plan to input more complete T/S specs when available into Hornresp, WinISD Pro or similar and let it calculate more correct Vas, Fs, Qts values.
One thing that Manfred Huber noted is that Onken (the Japanese Co.) that allegedly started all this used in-house custom made woofers to suit and in another post Jean-Michel noted that these speakers designed by 'Mahul' (the owner?) were small monitors requiring either stands or built as tall/thin columns to get them up to ear height, so have no real correlation to either the Onken 360 or Petite.
Regardless, many (most?) posted Onken builds, especially older ones that I've checked, either aren't built per the formulas as best I can tell or any measured specs used were quite a bit different than what I had even after factoring in series resistance, but manipulated to get a desired result. Remember, when all this was worked out, T/S alignment design routines either didn't exist or were at least widely unknown, especially to the typical DIYer. I only knew of them well before most folks because I was fortunate enough to be on Altec's internal tech letter mailing list.
If you want to use the SS to get acceptable vent lengths, then you can adjust Fb via changing it's 0.39 factor to a higher value to shorten them. Note that other parameters may need to be changed to get the desired frequency response, so I recommend designing the alignment in a T/S program and then adjusting this factor and 'n' as required to get it to calculate the multiple vents probably a bit more accurately than WinISD or similar. Otherwise, if the vents are too long and don't exceed 13,560"/(20*Fb), then my attitude is.........
Oh well, it will just make them even more 'colorful'/damped/TL like in nature as whoever made the drawing below erroneously believed.
GM
Attachments
The 'W', a simple reflex with technically too small a vent area for the alignment, the antithesis of what Onkens are all about and never seen any reasoning for it, only how great it sounds (presumably due to its sand damped double wall construction).
GM
Yes, but working backwards with the spreadsheet using the Petite Onken's published vent area and known accurate Altec data for the 414C, one ends up with a vent length approximately half again that of the Petite Onken.
And regarding Rg, how does one reconcile the large increase in box volume resulting from a small increase in Rg when using different amplifiers?
John
Last edited:
After two days of plugging numbers into the Debien spreadsheet and endless fudging I have to conclude that Altec drivers, new and old, are unsuitable for Onken enclosures. I would like to try this enclosure but the spreadsheet and the way it can be manipulated leaves me without enough confidence to want to put the time and effort into building such a box. If the published Onken box works well for the Altec 515 then I'll go ahead and build it without regard to theory and spreadsheet designs. It seems popular enough, but I've only heard one once and not for very long. Can anyone with any long term experience with this enclosure recommend it as a worthwhile project?
John
John
Yes, unreservedly.. I'm using the Iconic version of the late 515 as I have mentioned, and I am pretty happy with the end result after 4 yrs of continuous use. Think I said this all before..
Please PM me with an email addy and I will send you the relevant spreadsheet as I mentioned in an earlier post. My design is similar but not identical to the published design.
On the mids I have JBL 2440 drivers on 2311 horns with 2308 diffusers - in my room this provides the best compromise in directivity and very good transient response. (I listened to Smith horns, tractrix and exponentials as well) On the top I have JBL 2402, and also tried the 2405.
Attachments
Last edited:
Previous post where I mentioned my modified version of the debien spreadsheet. Note that the measured performance turned out to be somewhat better than predicted by the spreadsheet so it could be that the original design is a bit more on the forgiving side than it might appear. The original Jensen Ultraflex article if you do not have it may be useful. Reprints are cheaply available on eBay and IIRC I believe there is a least one copy on the web.
The GPA 515 I believe is virtually identical to the Iconic 165-8 so my numbers might be of interest to you.
My slots are several cm shorter than 35cm which is supposed to help ever so slightly with the combing issues. Slots area is just under 86%. My boxes are pretty large with a combined box and vent volume ~ 11.4 cu ft.
I've placed my slightly modified version of the debien calculator here.. Something to note is that in my case most of Rg is actually from the SE power amplifier which has about 2.5 ohms of source impedance. Obviously relating the efficiency measurements to the voltage on the amplifier terminals results in much better efficiency than the calculator reports based on an ideal voltage source with an external resistor to get the required Rg. The balance of Rg is from the choke in the X-O and its wiring. (The speaker cables are less than 0.05 ohms)
Last edited:
I have no idea. Never heard them used for P.A., just hifi. Biggest amps I heard them with were a 50 watt P-P EL 34 (passive x-over) and a 40W class-a solid state (active). So nowhere near the max rating. The Onken W with the 40W Keneda class-a was amazing! Never thought a speaker could do that.
John, I've heard the Onken with the 416, the petit Onken, the mini Onken with the Focal 10" and a few of Dave D's Fonkens. They were all very nice and very worthwhile. But bigger was always better, IMO.
..
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.