Martin King's MLTL-48 for jx92

[keladrin]

Hi,

I am impressed with this design, having built it last year and I am seriously thinkimg about building a similar one but wish to include 2 drivers instead of 1 for a far-field application. What changes are recommended for this configuration? Would doubling the internal volume suffice?

 

[Scottmoose]

You mean GMs? http://www.ejjordan.co.uk/diy/index.html

Yes, doubling the CSA of the line should suffice, no other mods needed. However, if you're going to do this, I'd suggest running bipole. Two FR drivers on the same baffle is begging for trouble.

 

[keladrin]

Hi Scottmoose,

Please explain - is the touble you refer to to do with the back-pressure on the drivers?

 

[Scottmoose]

Nope, it's the inevitable comb-filtering & attenuation that bothers me.

OK, if you place two drivers producing the same frequency on the same baffle, unpleasent things happen. The outputs will interact with each other, starting at a frequency determined by the distance the two driver-centres are apart (so, if you mounted 2 JX92s units as close together as possible, this would be 2.45KHz). When this happens, everything above this frequency will suffer from progressive attentuation. Worse, while you can Eq that flat, there's nothing you can do about the severe comb-filter effect / lobing that occurs (basically, the response will become increasingly ragged).

 

[GM]

Yes, doubling the CSA of the line should suffice, no other mods needed. However, if you're going to do this, I'd suggest running bipole. Two FR drivers on the same baffle is begging for trouble.

Greets!

Not quite , for a given alignment you have to also either double the number of vents or use an equivalent area single one which will be a little shorter due to being more mechanically efficient (less friction), otherwise it will be tuned lower and the vent may audibly 'chuff' at higher power.

Maybe not , it depends....... Consider we angle the drivers to focus on some point in the distance. The outputs will sum to create a well defined, but relatively small 'sweet spot' at this point in space/time, so if the focal point is far enough away that the angle is sufficiently shallow there's no obvious comb filtering. How far away this point is where dual drivers on a flat baffle sounds as one over an acceptably wide arc depends on the line's height, early reflections (if any) in our acute hearing BW and of course the individual's hearing acuity.

Another thing to consider is acoustic energy rolls off with increasing distance and frequency, so the further away you are the more tonally balanced a rising response speaker is. This means in the 92's case it may not need to be toe'd in as much if we're talking a really big room, which allows for some unconventional baffle layouts to be considered.

Regardless, bipole is great for typical listening distances if the rear driver is rolled off to just being an acoustic baffle step filter, but in large areas/distances with wide BW point drivers you need all the acoustic energy focussed towards the audience for max tuning flexibility.

GM

 

[Scottmoose]

Blast. Forgot to mention the vent. Thanks Greg.

I'm obviously not having a good day today. I'd forgotton the rising response of the Jordan too. Good point re forward energy for the nearfield. I did think about possibly focusing the drivers (I was looking at foucsed arrays only yesterday, which is kind of ironic), but with only two the options are a bit more limited. Still, worth exploring anyway if it's done right.

 

[keladrin]

So what we are saying is that at a listening distance of say 3m + and the drivers a few cm apart (the distance between drivers small compared with the difference between the distances of each driver from the ears - the drivers are either side of ear height), comb filtering will not really be noticeable by many people, unless they move closer to the speakers (assuming flat baffle), while at closer distance there is some attenuation noticeable and nodes apparent as one moves up/down.

Bearing in mind in stereo mode there will be a certain degree of comb filtering anyway if the treble is coming from 2 separated sources and you move off-centre, and this is very apparent even at a few metres, because of the wide separation.

Then what about in single driver mode when the driver is toed in as recommended - are you not listening through two ears, each at a different distance from the driver so there will be comb fitering and attenuation created in the listener?

 

[Scottmoose]

No, but with careful positioning & angling of the drivers (not necessarily a flat baffle) it can be reduced or eliminated.

In stereo? Perhaps in the centre image. I think you might have meant mono? That certainly has an issue with combing, & why some 78-fiends have reverted to mono setups. OB with a co-ax is particularty favoured I believe.

 

[planet10]

Then what about in single driver mode when the driver is toed in as recommended - are you not listening through two ears, each at a different distance from the driver so there will be comb fitering and attenuation created in the listener?

This is a natural situation for the ear/brain and is how we can tell where a sound is coming from.

dave

 

[GM]

So what we are saying is that at a listening distance of say 3m + and the drivers a few cm apart........

Bearing in mind in stereo mode there will be a certain degree of comb filtering anyway.........

Correct, though per my caveats, I don't know if 3m is far enough away for most folks, I never did much in the way of experimenting with arrays except 'infinite'. About all I can say is that two stacked Altec 511 horns (10.62" c-t-c spacing) was unlistenable to me back in my youth above a few kHz at ~17.5 ft IIRC and still a bit 'phasey' sounding ~46 ft away at the other end of the room, but these 'throw' the sound a ways, extending its nearfield, and are more phase coherent in their passband than a wide BW point source driver, so the distance can't easily be scaled AFAIK.

Bozak did a vertical tweeter array using four 2" cone units, but I didn't like it as much as his earlier virtual multi-cell tweeter horn, which used eight 2" cone units, even though from an acoustic phase POV the array had to be far superior, so go figure. Bottom line is neither were very coherent to me when used as stereo pairs, though the virtual horn was better IMO for mono, which it had been designed/optimized for, so no surprise there.

Right, stereo is nothing more than mass quantities of complementary comb filtering and why it doesn't take much to audibly degrade it.

 

[keladrin]

No, but with careful positioning & angling of the drivers (not necessarily a flat baffle) it can be reduced or eliminated.

With just 2 drivers perhaps the baffle can be flat because they can be equi-distant from the ears (one above, one below listening line), while with more than 2 drivers there will always be some at different distances (in theory)?

 

[Nardis]

Interesting idea. I like the idea of moving a bit more air without adding a different driver and more xover. I have added a Fountek ribbon to my 48inch MLTL using a xover based on Jim Griffin's

I use pentagonal cabinets and a quick crunch on my spreadsheet says that the cabinet dimensions will not have to be unacceptably greater to double the internal CSA. while keeping to a narrow baffle and with a good WAF.

Would I be able to use the 2 Jordans + ribbon in a D'Appolito arrangement?
Presumably there would be an efficiency gain?
What would happen to bass extension?

Nardis

 

[keladrin]

I assume bass extension would not be significantly better but there would be a large efficiency gain (and probably no need for baffle-step correction). Also better dynamics and punch I assume, and higher power handling capability. Come to think of it, apart from the horrendous expense (nearly double) it seems a good mod for a serious system.

 

[Colin]

You could roll off one of the JX92s above 250Hz or so if both facing forward.

If running them parallel, a 3 ohm resistor in series is recommended (by the Aurousal.co.uk site, which has a short section on using two of their JX92 speakers per side). The JX92S are 4.5 ohm.

Question for Scott or GM - presumably, if using two drivers, you centre them about the recommended distance from the top of the line?

 

[Scottmoose]

Yep. Centre taken from the point where the two drames touch.

 

[GM]

Interesting idea. I like the idea of moving a bit more air without adding a different driver and more xover. I have added a Fountek ribbon to my 48inch MLTL using a xover based on Jim Griffin's

I use pentagonal cabinets and a quick crunch on my spreadsheet says that the cabinet dimensions will not have to be unacceptably greater to double the internal CSA. while keeping to a narrow baffle and with a good WAF.

Would I be able to use the 2 Jordans + ribbon in a D'Appolito arrangement?
Presumably there would be an efficiency gain?
What would happen to bass extension?

Nardis

Greets!

Right, 2*CSA = ~W*1.4142, ~D*1.4142, but I don't know how to figure a simple calc for a pentagon (assuming it can be done).

Anyway, this is a very popular layout, but like any D'Appolito alignment it will have more off-axis issues than just a two-way, so must be either XO'd low/steep enough to not have to toe them in so much for best performance or use a fairly complex XO to the desired FR without stressing the ribbon.

Bass extension is what you make of it. Obviously, using the same tuning as a single driver yields the same tonal balance except theoretically no need for any BSC filtering. Tune it lower and you trade some LF 'weight' for a lower cut-off, which is what I did with my dual driver cabs since they are backed into the corners.

Acoustic efficiency should be +3 dB down low, but due to the need to keep the nominal impedance high enough for most amps it will be mostly negated by a lower voltage sensitivity, so the net gain will probably be minimal for relatively high extra cost.

GM

 

[GM]

Ran across some math today that defines the transition from near to far field of point source drivers. It doesn't jive with my ears, so as always YMMV:

l = d^2/(c/f)

where:

l = distance from speaker
d = center-to-center (ctc) spacing
c = speed of sound (SoS)
f = frequency of interest

Since the distance increases with increasing ctc and/or frequency, 20 kHz is the practical limit, or only ~45" for the JX92S.

 

[Nardis]

Greets!

Right, 2*CSA = ~W*1.4142, ~D*1.4142, but I don't know how to figure a simple calc for a pentagon (assuming it can be done).

GM

The answer is there is no simple calc that I've found - hence the need to write a spreadsheet which divides the pantagon into triangles and rectangles.

Here's a PDF of the cross section in 18mm of a pentagon for twin JX92s.
I'm wondering if I want to do it all again with the second driver. I hadn't figured the need to cancel out the gain in efficiency by the need to beef up the impedance. But at least this time I know what the assembly process mistakes are liable to be.

Your original CSA translated into metric at 19381 mm2.
In inches the twin design works out at exterior measurements
Baffle 6.1" Sides 8.4" Rear panels 6.3"
This keeps the narrowest possible baffle. Its widest point is about 12" and it's about 10.5 deep.

It needs a good table saw.

 

[keladrin]

Greets!

Not quite , for a given alignment you have to also either double the number of vents or use an equivalent area single one which will be a little shorter due to being more mechanically efficient (less friction), otherwise it will be tuned lower and the vent may audibly 'chuff' at higher power.

GM

Hi GM,

Sorry for the delay - just to clarify, do you mean a doubling of port area will suffice (same length) or do you have to double length also. My tuning prog says you have to double both for same tuning (with doube volume) but this is for BR so may not apply?

 

[GM]

Greets!

Your program is wrong if it doubles the vent length every time you double the drivers, Vb and number of equal area vents for any vented alignment. In this case the vent length stays the same. Think about it, what's the difference between this and bolting two identical vented speakers together?

Indeed, there's a good argument for doing it this way since two drivers typically have different specs, so won't electrically equally share the available power causing one to excurse more than the other.

GM