Olson/ Nagoaka Manifold Horn for Audio nirvana?

Corvus corax

Member
2005-08-31 11:59 pm
I've been living with the AN Classic 12 ferrites for about a week now, and given the success I've had with them in a big TL enclosure, I'm thinking about slotting them into a huge manifold horn. The driver has been sounding quite good, and I'm now thinking about its "forever home", and I love the lines of the big Sachiko horns.

The Qts of this guy is currently measuring about .43, and 43 Hz Fs, but that's going to go down as the suspension continues to break in- the specs for the driver state .34, with an Fs of 31Hz, and I think it'll end up at least close. Its pretty loose already, but the Fs hasn't changed at all with about 20 hours of exercise so far.
 
Providing the horn is designed for the driver / system and is sufficiently large, there's no technical reason why it can't be done. I would be a trifle surprised if Q dropped from 0.43 to 0.34, but YMMV of course.

The main issue you will need to contend with is size. It will need to be extremely large due to the high Vas of the AN 12in drivers. Ideally, it will need to be approached from the POV of the actual system Q, i.e. account for any increase due to amplifier output impedance and / or series R from wire etc. in the chain. Providing you've got the space for the titanic enclosures, a speaker designed along those lines should work fine.
 

Corvus corax

Member
2005-08-31 11:59 pm
Yeah, driver Q is one of the seldom accurate published specs- I usually assume and add at least 15% to the published Q, and why I go through the trouble of measuring each driver before I build enclosures, hence the use of the old failed TL as a test box…

Once I have the parameters sussed out, how do I begin? I've never built a manifold before- these guys are a bit mysterious to me. What are the steps?
 
That would take a while, since it's essentially needing to distill the best part of 15 years worth of research on my part into what has been done before in the West and in Japan, and from that, how I develop my own variations, since they differ somewhat from Olson's and Nagaoka in conception, while sharing some features. It depends too on the specific type of horn, since there are sub-variations with different objectives in mind.

At a very simple level, design suitable horn, approximate the expansion with a number of straight manifolds & decide how you can best fold it up. That tends to be over-simplistic in practice, since it gets a lot more complicated very quickly if you want to go further, since manifold expansion technically allows the most efficient use of a given OA cabinet size when it's designed with that in mind. And in many cases, you're actually dealing with resonant rather than full horn loading since the flare rate has been compromised to get the box size managable. Then you start to get into the chamber design & throat location (which has quite a signifiant impact on behaviour) and especially the dimensional ratios of the individual manifolds, their positions, the folds & edges between them etc., which I tend to use as part of the low-pass filter; that last varies depending on who designed it, some include it, some don't. That then leads you into possibly adapting the flare ratio you're targeting -back to the compromises & the balance. About the best advice I can give is, assuming the horn will be a compromised type, remember you are dealing with resonance. I'll have a look myself re the AN units, but if memory serves you probably won't like the size the horn will need to be, and that's only ever going to go one way if system Q rises, as it almost certainly will. ;)
 
Well, its doing really well with the TL (that didn't work) I made for the B&C 15 inch driver. Its just a shade under 7 cubic feet, and the bass is really about all I could ask for, with the possible exception of the mid bass or lower midrange, which seems a bit depressed. Voices are robbed of body, currently. Perhaps bass reflex is the only way these will work.
 
Not necessarily. It depends on the design. As noted, a horn variation will be very large, but a Karlson type may do OK. As far as a BR is concerned, working purely off the published data a ~185 litre box tuned to the low 30s should be reasonable, or if you want some 'safety' for changes in Q, amplifier output impedance etc., then the trad. approach of Vb = Vas, Fb = Fs should sort that.

A TL / QW variation will be larger for a given tuning, but again, should do OK. The current lack of body you note in the midbass may be due to insufficient suppression of F3 or possibly F5 (more likely the former).
 
I'm going to try the 5.6 cu. ft boxes he recommended for the drivers, with the recommended port size- see where that gets me. The driver sounds as if it has a lot of promise, just needs a little more output between 100Hz and 1K. I may try a hybrid FLH/ bass reflex cabinet like the Tannoy Westminster with it…
 
The Audio Nirvana cabinets are often not very optimal.

For the Classic 12 (Ferrite) 5.6ft^3/158L is actually pretty good, but the factory tuning is not very good.

attachment.php


It also looks like it will work pretty well in 80-90L sealed (add subs below ~50 Hz for full extension).

If i were to do a monster miniOnken it would likely be 100L.

A smaller cabinet is easier to keep from ringing.

dave
 

Attachments

  • AN-classic12-sims.gif
    AN-classic12-sims.gif
    73.2 KB · Views: 471
So I've finished running in the drivers as far as I'm going to, and I've gotten the following results:
Re- 6.9 Ohms
Fs- 44.41 Hz
Qts- .412
Qes- .466
Qms- 3.562
Le- .149 mH
Mms- 31.1g
Vas- 218 L

So, a bit off from the published specs, which is not a surprise, but the two drivers measure exactly the same for all intents and purposes, so at least they are consistent. The results for a generic bass reflex box call for a 8.25 cu ft box, which I'm not really keen on…
 
Since it's a trad. style driver, a straightforward reflex following the old Vb = Vas, Fb = F0 alignment will be decent enough, once lined appropriately. You could drop that down to about 165 litres, maintaining the same tuning which will be a slightly more damped alignment than the near-maximally flat that you'd get with the larger box above. You can lower tuning a bit in both these if desired to get more LF damping.

Of course, if you wanted to go all-out for a trad. style box, you could crank it to ~18.5ft^3, again tuned to F0, and damp the box via the old click test method until you're happy it's properly damped for your requirements.
 

freddi

Member
Paid Member
2005-08-16 4:21 pm
I'd think it would behave alright in the original Karlson that is 7.3 cubic foot external bulk, but go lower in a Karlsonator 12 (about 6 cubic foot external) or similar Karlsonator Since you've stable parameters, you might ask xrk971 to run a Karlsonator sim.