[hum4god]

hi

i am ready for  a new project . finally i will try some ob's .

as a starting point :
i already have a boxed bass (the bd design oris bass ) with a 15" driver , around 98db efficient that can cover easily the range up to 300hz and possibly higher .
a photo can be seen below in the next thread ,( i finally figured out how to post photos , yippee)

the bass is relatively tall and the front is flared which makes it a little more complicated to add an ob . the two ideas i came up with are :

a linkwitz orion frame that would go around the bass bin . photo below.
or
hanging the baffle from the ceiling as seen below but with the bass box placed underneath:
see photo next thread .

my questions are in regard to baffle construction and crossover:

the baffle width is restricted in width to 18"  max or less. and i was wondering if there is a formula with which i can calculate rolloff of the driver in a particular size baffle ?
the distributor told me that the fertin 20ex will roll off at around 197hz with a baffle width of 18"  which seems to be high , but actually not a problem .

a few general questions for building a baffle:

how do you best decouple the driver from the baffle?
is it important to have the driver off centered ?  (why?) i have seen many designs that do not seem to care .
what is the best way to mount the driver ?  from the back , or flush , or ???

next thing would be to crossover between the bass and the widerange . options are actively crossing over or using the natural rolloff from the built baffle so i would not need a crossover on the fertin .
i would just need to  crossover the bass active or passive which is not so critical in the lower freq.
so what i would prefer is building a baffle that fits my spacial restrictions (18" width max. or less) and preferably also gives me a natural rolloff for the fertin driver in the 150-300hz region.
is that possible to achieve?

how could i measure freq. rolloff once i built a baffle with the fertin to choose appropriate crossover freq./slope for the bass  ?

specs for the fertin 20ex is afaik ( there is a new model coming out with hopefully the same specs .):

20EX
HAUT-PARLEUR LARGE BANDE 21CM
Impedance (ohms) 8
power capacity (W rms) 50
sensitivity (dB) 100
flux density (Tesla) 1.4
total flux (Mx) 200000
moving mass (g) 18
mecanical compliance (Cms m/s) 0.00085
resonance frequency (Fs Hz) 42
mechanical efficiency(Qms average) 15.813
electrical efficiency factor (Qe) 0.428
total efficiency factor (Qt) 0.416
BL factor (0) 8.26
inductance (mH) 0.25
equivalent air volume (L) 48
displacement of diaphragm (Xmax mm) +/- 3
voice coil resistance (Re ohms) 6.4
effective diaphragm area (Sd m2) 0.02
overall diameter (mm) 230/210
bolt circle diameter (mm) 216
depth (mm) 150
net weight (Kg) 8
power supply (V/A) 12,5/1,5
consumption (W) 20
GAP height (mm) 12
voice coil height (mm) 6
surround neoprene
cone paper
voice coil material cop

any help is much appreciated
malcolm

[hum4god]

here a few photos for illustration:
the bass:


the diy orion i have seen somewhere:




the hanging baffles:

[D OB G]

Hi Malcolm,

Funny you should mention Linkwitz, because not only does he use the Orion baffle now, but in the 70s he suspended Kef B110s in boxes, just as you show in your photo.
Three companies at least, Naim, Jamo and B&W isolate their tweeters from the box, and Linkwitz magnet mounts the midrange in the Orion, having it only resting against the baffle, with a gasket.

I have used the magnet mounting method to good effect for both bass and midrange, but I don't think you would want to go too low in frequency with isolated tweeters (Newton's Law and all that), and it would be interesting to know if others have had success with suspended midranges/full-rangers, such as the Fertin (hanging baffles will lose the advantage of floor coupling, and the reduction in size of baffle that needs to be used anyway).  The Fertin, it seems, will be going down into bass regions, so I don't think hanging them would be a good idea.  It really would be good to know how dedicated midranges would perform this way.

Gilbert Briggs, of Wharfedale fame, compared about six mounting configuarations.  Mounting on the front was the secound best, and mounting from the rear came down the list. (The best was mounting in the midldle thickness of the baffle , on a small thinned sub-baffle, itself mounted in the middle thickness of the baffle.  He took diffraction effects into account).

A program called The Edge can be downloaded for free, and it can be used to simulate the effects of baffle size and driver position.  There seems no question that it is absolutely soundly based in it's algorithms, but in some ways it doesn't seem to give entirely real world results, maybe because it doesn't take floor contact into account, and it doesn't take driver Q into account, which won't matter however in the frequency range you're talking about.

As far as as mounting off-centre goes, Linkwitz uses complex equalization to cater for such things as his central mounting, as well as floor bounce, baffle step, time delay of tweeter, and subjective contouring of the frequency response, to name just a few parameters.
He gets an optimum response this way, so I wouldn't say that an Orion clone with a simple crossover is going to work necessarily, yet others, such as the Hawthorne Silver Iris systems use symmetrical positioning, apparantly also to good effect, (but with lumpy frequency responses) without equalization.  I have seen a number of photos of people using Visaton B200s off-centre, and with a sloping baffle top

A common approach seems to favour mounting off-centre in every direction according to the golden ratio (1.61 to 1?), The Edge shows the peaks and troughs that occur with a symmetrical alignment.

You could go the whole hog and purchase, or borrow!!, test gear (I use Liberty Instrument's Praxis, but there are many others).  Obviously it depends on whether you can amortize the cost over other speakers.  Another option is to purchase a Radio Shack sound pressure level meter (relatively cheap, and relatively accurate), and use it in conjuction with, say, a computer based audio signal generator, being mindful that highish level pure sine waves can easily cook voice coils, and you won't be isolating room reflections, but it may be good enough to give the measurements you require.

As far as rolling the Fertin off goes, I would recommend taking the Xmax of +/- 3mm into account when it reaches down to bass frequencies.  It may be heresy in this forum, but you might want to consider a capacitor in series if you play your music loud.  Just watch  the driver when you play your music and see if it exceeds the 3mm excursion.  If not, then it is definitely best left alone.  Given the medium Q of the Fertin, the roll-off will occur at 6 dB per octave.  This a bit tricky to integrate with an active bass module, which usually rolls off at 12 dB per octave, but at least it is better than the other way around!!

Test gear with active amplification to both drivers solves your problems, but then we're talking money!

Hope this has been of some help, and not just confused the matter.

David

[hum4god]

thanks David

as i heard it seems that the golden ratio ( 2/3 =0.6666666 ) is something people take into account especially on bigger baffles . and its not so important on smaller ones  where it seems to suffice to put the driver off center .

i will probably try to play with the edge program if i can figure out how it works .
and then just build some prototype baffles to see how it sounds .

are there other methods of measuring response in room besides radio shack spl meter and how accurate is the radio shack spl?

you said that the fertin on the ob will roll off at 6db . is that generally the Case for drivers on ob ?

i am still waiting for some answers from Mr. fertin who will give some advice n that matter .

thanks

[D OB G]

Hi Malcolm,

I just checked the net, and the golden ratio/section/mean is 1.618 (the driver should be set within the baffle also at the golden ratio).
I'm just about to use DEQX to crossover and equalize some new speakers, and I want the smallest difference between peak and trough as possible, so that I need to lose the least efficiency.
Using the Edge (which is said not to be as accurate as Martin Kings Mathcad worksheet (requires Excel 2000 or above, and I don't have that on this computer!!!), I have found the golden ratio to be useful up to about 3 kHz.  Most alignments seem to be fairly flat above that.
I've also found that the flattest response is to have the driver a little closer to the edge of the baffle,  than the ratio would suggest, and to have one side angled
I think building some prototypes is a very, very good idea, but I am of the school where measurement doesn't necessarily mean good sound, (and open-baffles are notoriously difficult to measure, even with DEQX), but where it is easy to fool yourself about what is going on just by ear.
As far as I know, the Radio Shack meter is the cheapest way there, and if you take close enough measurements in a quiet room, might well be enough (it's considered to be reasonably flat to about 10 kHz, so the range is not an issue).
As far as I understand it, and I am open to correction, the acoustic roll-off of a driver with a Q of 0.7 will quickly become exactly 6 dB per octave,  but with a higher Q, the roll-off is not as steep, and with a lower Q, it is steeper.  This depends on the baffle, but is most significant below about 100 Hz (depends on just how much the Q differs from 0.7).

David

[D OB G]

I just re-read Linkwitz's site, and he still states that hanging a midrange and above baffle, from the ceiling, is a good option.

David