Quick & Dirty 4430-Inspired Two-Ways Part II

[johnaec]

My personal preference would be to follow the front trapezoidal shape all the way front to back...

John

[Zilch]

Giskard:

0) 2.15 uF, 0.60 mH
1) 2.15 uF, 0.76 mH
2) Full range

[Guido]

Sorry for late cooperation!
I was on a business trip.

Hope Giskard don't mind if I jump in with the voltage drives.

Green: original
Yellow: My suggestion 1.5uF, 0.8mH
Pink: My suggestion 1.0uF, 1.0mH
light blue: Giskards suggestion 2.1uF, 0.6mH
dark blue: Giskards suggestion 2.1uF, 0.76mH

We're coming closer Slowly

[Guido]

Green: original

Yellow: 2,15uF, 1,0mH

[4313B]

Hope Giskard don't mind if I jump in with the voltage drives.

That isn't what I get. You must be using a different load. I never got the impedance curve from Zilch so I just used 8 ohms. My suggestion was based on 8 ohms and the Q was kept identical to the original, again, based on 8 ohms. You guys figure it out from here.

Input adjusted for gain to show actual curves.

[Zilch]

The RTA curves are on my best driver mounted in L200, which I'm pulling today for Widget's testing tomorrow, so I'll be able to get the complete WT2 measurements completed and posted this evening. I'm wanting to keep the variables of different drivers out of it for now, though I may measure and report at least two. We'll have CLIO results as a cross-check on WT2 in a few days, looks like.

From those curves, it seems the entire correction is less than 15 dB. It is very encouraging that we're not having to push these as hard as earlier designs to get the requisite HF boost....

[Zilch]

Giskard:

BBP and WOO files as requested. The impedance sweeps are in the .woo files, presumably, Arb1, 40 pts. If you need higher resolution or want them in ASCII log files, let me know.

XLe (1 kHz) is 5.9740 on #1, 6.2528 on #2

Note: The second filter in the circuit was changed early on in conformity with actual factory crossovers received:

C2 = 6.0 uF (not changed)
R1= 2.5 Ohms
L4= 0.04 mH, .225 Ohms

[Guido]

That isn't what I get. You must be using a different load. I never got the impedance curve from Zilch so I just used 8 ohms. My suggestion was based on 8 ohms and the Q was kept identical to the original, again, based on 8 ohms. You guys figure it out from here.

Input adjusted for gain to show actual curves.

Giskard, I used the real values given from Zilch.

2.5 Ohm for R1
0.04 mH for L4

The load I used was 8 Ohms

I think this is the reason for the other results.

[Mr. Widget]

Here is the Clio impedance curve from one of Zilch's LE14H-3s. The red curve is with the woofer on a carpet covered concrete floor sitting on blocks to avoid blocking the vent. The Blue curve is in free air. The Zipped data file is from the Blue curve.

Widget

[Zilch]

Mr. Widget's latest horn/driver studies for this project are now posted here:

http://audioheritage.org/vbulletin/s...ead.php?t=6050

[Zilch]

Have you made a comparison between the WT2 and Clio on that LE14H-3?

Not yet, alas. Too much data, too little time.

I'll get to it tho. I promise!

My suggestion was based on 8 ohms and the Q was kept identical to the original, again, based on 8 ohms. You guys figure it out from here.

Assuming 8 Ohms impedance, the Q is 0.425. I don't know why JBL's calling that a 1.1 kHz crossover frequency, 'cause it's down 10 dB there. 2.15 uF with 0.76 mH shoulda done it, at least at that frequency.

I was expecting the entire knee to shift leftward, but constant Q seems to be the right approach, as the response is nowhere near flat in that region. Part of what it's doing is attenuating a 5 dB peak there. Am I starting to get this?

What tool(s) do I need to get on the same page and raise my level of incompetency here?

[Guido should be gettin' his drivers soon....]

[Zilch]

Kinda 60's moderne esthetic:

[johnaec]

Kinda 60's moderne esthetic:

I think a walnut finish would look better...

John

[Guido]

Assuming 8 Ohms impedance, the Q is 0.425. I don't know why JBL's calling that a 1.1 kHz crossover frequency, 'cause it's down 10 dB there. 2.15 uF with 0.76 mH shoulda done it, at least at that frequency.

I was expecting the entire knee to shift leftward, but constant Q seems to be the right approach, as the response is nowhere near flat in that region. Part of what it's doing is attenuating a 5 dB peak there. Am I starting to get this?

What tool(s) do I need to get on the same page and raise my level of incompetency here?

[Guido should be gettin' his drivers soon....]

Zilch, I didn't get this. Maybe the language barrier

How exactly can I assist you here?