Quick & Dirty 4430-Inspired Two-Ways Part I

[Mr. Widget]

- You know , it seems nobody worldwide gave 2 hoots about the lowly L200 until you threatened to cut the baffle board. "Fair-Weather Fans"

I was not joking when I offered to build a pair of replacement cabinets in exchange for those mint L200s.

Zilch, I guarantee mine will be Sub1500 proof as well and you can have them in walnut, maple, cherry, or any other wood veneer that is in that price range.

Widget

[Zilch]

... and you can have them in walnut, maple, cherry, ...

We'll talk.

[It'll require compound mitres, tho....]

[Zilch]

RTA in "Average" mode, GEQ flat, 2.5 dB per dotted line.

First, N3134 controls set to "0," i.e., HF at 4:00, MF at 12:00, with closeup.

Then, all controls set to max, and closeup.

[Dunno WHAT's up with the LF there. Somethin' hummin', probably.]

Does this make sense? Yes. Too much boost above 7 kHz.

Can't dial it down without losing VHF at 20 kHz, too....

[Guido]

I played a bit with my spice program.

I didn't come any further with modifiying the 3134 network.

My recommendation is to take the LF part of the 3134 for the 2235H and to take the HF part from the AM6200 but bring in an 8 Ohm L-Pad BEHIND the parallel RLC!
Maybe we have to shift this RLC a bit but we will see...

Green is 3134 set to "0"
Yellow is AM 6200 network with L-Pad behind RLC set to app. "3:00" or Rs=2 Ohms and Rp=24 Ohms

This should work well.

Please try it.

[Earl K]

I was not joking when I offered to build a pair of replacement cabinets in exchange for those mint L200s.

Zilch, I guarantee mine will be Sub1500 proof as well and you can have them in walnut, maple, cherry, or any other wood veneer that is in that price range.

[Zilch]

A minor glitch. Factory crossovers use 2.5 ohms for R1, not 4.0 Ohms. DUH!

C2 remains 6.0 uF

Value of L4 is unknown, and no LCR meter here. Hopefully Widget has a way to read it at the audition tomorrow....

[He'll be busy hosting, tho.... ]

By "Behind the LRC," I assume you mean between the LC and the LRC. No other location makes sense, actually, unless you're just doing attenuation on the output.

[I'd call that "Ahead," going from in to out. ]

[Mr. Widget]

Bring your coils and check 'em out... we offer full service here at the Works.

Widget

[Zilch]

I feel the same way about the dispersion pattern of the 2344A, even though I listen in a room that would charitably be called "difficult". I would think that with my 7'2' celings, the narrower dispersion horns would be better, but I miss the more open sound field I used to have.

Why they work for us is here, I believe:

http://www.jblpro.com/pub/technote/tn_v3n01.pdf

We WANT that combination of direct and reflected sound. Others want to minimize the reflected part. Both are perfectly valid preferences....

[Zilch]

N3134 (4430) crossover schematic says Zobel is retained in the LF circuit in external (biamp) mode. N3135 (4435) crossover drops it, apparently.

I'm running 2235H's full range to fill the notch between 1 kHz LF and 1.1 kHz HF crossover frequencies revealed in initial tests with AM6212 HF crossover. Would adding in the Zobel help the situation?

First RTA plot in "average" mode shows response running full range.

Second is with Zobel in circuit. Apparently, JBL 6260 amp running the 2235's doesn't give a whit about 2235 impedance.

Third is N3134 low filter kicked in. Ooops! Notch City. It MATTERS. We'll have to move one or the other of the crossovers' frequencies to close the gap if we're going to retain N3134 low filter as Guido suggests. Whether 2431 will play lower remains to be determined. JBL don't do it anywhere I've seen.

Last two plots show full range vs with filter at lower resolution. The notch is 12 dB deep. We can't ignore it; it's quite audible when switched in and out.

[Guido]

Last two plots show full range vs with filter at lower resolution. The notch is 12 dB deep. We can't ignore it; it's quite audible when switched in and out.

Might be a polarity issue?

2426 (4430 original) is negative, 2431H most probably positive.

In any case I suggest to keep the LF part conencted to the 3134 network as it is designed for the 2235.

I'll rework my curves for the AM 6200 modification and post shortly. Will correct the R1 value

[Guido]

A minor glitch. Factory crossovers use 2.5 ohms for R1, not 4.0 Ohms. DUH!

By "Behind the LRC," I assume you mean between the LC and the LRC. No other location makes sense, actually, unless you're just doing attenuation on the output.

[I'd call that "Ahead," going from in to out. ]

R1 2.5 Ohms works even better. See pic. Blue=4 Ohms, Yellow=2.5 Ohms

NOPE, I meant just attenuation between the LRC and the driver! We need to attenuate the 2431 a bit so try a L-pad first.

[Guido]

To close the 1kHz gap please -in addition to check polarity issue- play with the value of L1 in the AM network.

4dB attenuation is already in that voltage drive!

Green=stock
Yellow=1.0mH
Blue=1,5mH

[Guido]

This 1kHz gap is weird.

Voltage drive of the 3134 and AM networks shows even more boost in the 1 kHz region for the AM. See pic.
Green=3134
Blue=AM

Should the 2431 be so much weaker in the 1kHz region?
Do we have measures of the 2426 and 2431 w/o any network to see their response?

[Zilch]

L4 spec = 0.08 mH. Actual units read 0.041, 0.041, so it was changed as well as the resistor. Call it 0.040, a likely selection.

L1 spec = 0.6 mH. Actual units read 0.583, 0.596. It was not altered, apparently.

Most of the 1 kHz notch is a phase issue, as you suggested. I'll post curves as soon as I get these crossovers back together tonight.

We don't have raw driver curves on 2431 or 2435 yet. Widget's doing Project May work, so it'll be a week or so before he has any time available, most likely. I wouldn't think 2431 would be weaker than 2426 at 1 kHz. Lower, like 800 Hz, maybe. We'll know soon....

[Zilch]

Left out of phase
Left in phase
Right out of phase
Right in phase

2.5 dB per dotted line.

I was playin' outta phase, because AM6212 did. Adding N3134 low filter reversed it. N3134 HF disconnected at C11; there's no interaction.

Note 2431 driver variability. It could be the horns, but I doubt it.
Edit: It's the drivers. I swapped waveguides.

Something in the room is eating 500 Hz pretty good....
Seems to be other speakers. I short them out or move them out, 500 Hz comes back.

I agree with Guido now. N3134 low filter is very important. It has built-in room correction, apparently. The bass comes to life.

[3 dB/octave seems a little overwrought, tho.... ]