MLTL driver/port placement, floor reinforcement etc...

[GM]

If you stuff, do it loosely and normally only about half way down its length, which means you have to keep it from falling down. I prefer to line one side, back, and top with 1" acoustic fiberglass insulation and if I need more damping I add it at the top since it has the most effect over the widest bandwidth (BW). The main thing is that normally you do not want much behind the driver, only just enough to damp standing waves between the driver and back wall.

Anyway, unless the speaker is near a wall or corner it should sound 'lean' even when turned in as Jordan recommends, so a baffle step compensation (BSC) filter will be required.

GM

[Colin]

Ah, the lean, mean 48 MLTL. After waving my arms about for months saying it didn't need BSC, I finally added one last year and then booked a restaurant to eat my words. The BSC helps a lot. Male voices in particular sound more natural.

My triangles have a front face 7.75" wide, sloping back to 9.25" wide and the compensation circuit was a 4.7Ω 25w resistor and 3.2mH inductor in parallel with each other and in series with the speaker positive lead. These were measured and showed the mid (500-2kHz) is brought down and the extreme HF goes up slightly.

Jim Griffin’s version of the MLTL (7.5“ wide baffle) uses a BSC of 4Ω + 1.5mH (3dB baffle step).

Hope this helps.

[Doug Kim]

Thanks GM and Colin for suggesting BSC.

I'll experiment with BSC and also with Zobel, but the problem is Jordan website didn't specify what the nominal impedance and voice coil inductance are. Any help please?

Doug

[GM]

Zoebel? Calculating off a couple of different impedance plots I get a <0.18 mH/~8 kHz based on a ~5.1 ohms Re, so you'll probably lose more SQ through the added components than any perceived HF improvement.

GM

[Nardis]

There's a thread started by EC8010 who measured the JX92S a couple of years ago. Unless the parameters of current models have been changed.....

Jordan JX92S impedance

[Colin]

According to Ted, they haven't.

[GM]

Well, I didn't understand Svante's response and for the most part I don't agree with the rest for one reason or another, so FWIW here's the solution for reactance annulling Re that at least one major electrical apparatus manufacturer uses: measure Re and find the frequency where Re has doubled, then:

R = Re*1.25

C = 10^6/(2*pi*f*Re)

It's worked well for me, but as always though, YMMV.

GM