Multiple drivers - space them apart by how much?

[bhg41088]

hey screten, do you have any instant messengers? I would love to chat with you.

[BillFitzmaurice]

You can't place the drivers too close together so edge to edge isn't a bad idea. You can place the crossover at the wavelength frequency but lower is also OK.

Wavelengths are calculated by dividing the speed of sound (approximately 1130 feet per second) by the frequency.

[johninCR]

Keep in mind there's some debate as to how to measure the actual spacing distance. Edge to edge is not 0 spacing and it's not cone edge to cone edge spacing nor is it center to center. Theoretically it's the spacing between where the sounds originate and what that is exactly is a matter of some debate.

[Mudge]

Originally posted by johninCR
Keep in mind there's some debate as to how to measure the actual spacing distance. Edge to edge is not 0 spacing and it's not cone edge to cone edge spacing nor is it center to center. Theoretically it's the spacing between where the sounds originate and what that is exactly is a matter of some debate.

If only it was that simple. The latest thinking is that it should be measured from the edge of the pressure zone where the impulse vectors (the direction the air is compressed/expanded in) are fairly similar. Nobody agrees on what constitutes similar though.
For a large rounded dustcap, this region is quite wide, and with a phase-plug it could be all over the place

Stick with centre to centre, it seems to work

Unless dealing with planars or flat pistons

[chipco3434]

JohninCR

I'm on my 6th line array project and I can tell you that the effects of lobing and comb filtering are vastly over stated.

Agreed! I love the sound and haven't listened to box speakers in three months. I can't go back.

Let's see some pictures, talk drivers, hear about some crossovers. Gimme somthin'.

[johninCR]

Don't have digital pics of the best one so far that was built for a bar. Each channel consisted of:

16 BK drivers- 4ohm, 20watt, 4", $1.50ea, Fs 95, on OB they roll
off steeply below 150hz and above 9800hz and are reasonably
flat in between except for a gradual rise between 150 and 300
I let them run full range. Very close spacing.

1 Pyle Pro PDBT28- 1" Titanium Super Bullet Tweeter, 300watt,
104db, harshness tweaked out with 6 pencil eraser sized
pieces of foam rubber stuffed in the horn around the bullet plug.
Used the included capacitor and an Lpad. It's mounted in the
middle of the 16 driver array.

2 Nippon 12" woofers - 235watt, 6 coats of PVA glue to bring
the Fs from 31hz down to 28hz. They're set up in a W baffle
baffle at the bottom with push/pull alignment and a 1st order
roll-off above 125hz.

The baffle has a narrow front profile, 5" at the top and 9" at the
bottom. Baffle wings are swept back at 55 degrees. One is 12"
at the top and 22" at the bottom and the other is 4" shallower
to avoid cavity resonances. I have 3 2"x2" cross braces in the
back to minimize baffle vibrations and make nice handles for
moving. It's made out of 3/4" plywood over which I glued a
textured vinyl. I made a rounded grill out of thin expanded
steel that runs the length of the array painted with black epoxy
appliance spray paint, which fits into pre-routed grooves. After
putting on the vinyl, I ran an exacto knife the length of the
grooves and the vinyl holds the grill in place plus a few black
screws for security. I always put a layer of fiber fill and 1" foam
rubber in back of the drivers to damp the highs a little and cover
the ugly backside.

Hints- It's a fairly easy 2 day build. Cut and assemble the baffles
the first day. Wings at angles are tricky since you can't use
clamps, so get a helper to hold them and use screws. Solder
driver wires before mounting because it's a pain afterward.
Vinyl or another material can make an attractive easy and quick
finish, so you aren't tempted to leave it unfinished for months
or years. Do a test run with the drivers mounted before covering
just in case you have a cavity resonance and need to cut a little
off of the shorter wing.

[jesper]

Hi,
In order to avoid lobing at crossover frequency you need to space the drivers much closer than 1 wavelength. At a half wavelength you'll get the first null at 90 degrees, to totally avoid lobing you have to have even less spacing. This means you have to cross very low for any practical 2-way, below 1khz.
So, lobing can't be avoided in an ordianary multiway design. Going for MTM or TMM will worsen the lobing (MTM beeing worse)
You should make sure that you have no lobe tilt by time alingneing the drivers (execpt for MTM which don't have lobetilt no matter what).
My point is, choose crossover frequency according to what is pracical for the drivers at hand and then place the drivers as close together as possible. Don't bother so much about the actual wavelength spacing at crossover frequency.

/Jesper

[BillFitzmaurice]

Actually lobing is not a problem with line arrays. It can be problematic with MTMs if the presence of the tweeter forces wide spacing of the outboard elements, which results in they're acting as separate point sources. With closely spaced multiple drivers, however, what is critical is the space between their radiating planes. ie, their frames. At frequencies where that space is less than a wavelength the multiple point sources will merge into a single coherent cylindrical wavefront and lobing does not occur.