Printable View
Continued from Quick & Dirty 4430-Inspired Two-Ways Part I
Giskard:
1) 1.7 uF, 0.6 mH
2) 2.15 uF, 0.6 mH
3) Full Range
Boosted everything about 3 dB, no frequency shift. (SOUNDS good tho....)
I assumed you meant stay with original inductor value, not Guido's suggested 1.0 mH?
Ok... should have been ~ 1.7 dB but whatever. How are you on coil values? Can you go in increments of 1/100 mH? With the 2.15 uF you have change the inductor to 0.76 mH. That should give you a 4 dB increase at 1 kHz and a 2 dB increase overall in the high pass.Quote:
Originally Posted by Zilch
I have four Jantzen 20 Ga 1.0 mH coils, and a B&K LCR meter that reads to 0.01 mH, yes. WT2 confirms the readings are correct, as required.Quote:
Originally Posted by Giskard
I'll unwind one to 0.76 mH and try it. I can also crank the RTA resolution up to 2.5 dB per major division, now that we're close. Can you illustrate graphically what's going on with this portion of the circuit for us please? I'm surprised the values require this level of precision.
I should invest in a filter simulator myself, probably. One thing this thread has made abundantly clear to me is that it's about the crossovers as much as the drivers, and, surprisingly, within fairly broad limits, the boxes themselves are the least of getting it all to work together....
Yeah, I can post what's going on with the voltage drive in awhile. Let me know how the 0.76 mH works first. Post the bb6p file for the horn/cd combo complete with Q, fs, etc. I'll also need an impedance curve from 500 Hz on up.
"One thing this thread has made abundantly clear to me is that it's about the crossovers as much as the drivers"
We've talked about this at length before. Back in the day when JBL was cranking out such stellar circuits as the N100, other companies were taking junk drivers and making them sound listenable with proper application of filter design theory.
Bigger than a breadbox, but decidedly smaller than L200 (rear baffles aligned here), Z1 has good WAF potential; no matter the viewing angle, you can never see all of the sides at once, thus masking its real volume. Trompe l'oeil confers delicious asymmetry (think L250,) in all but direct head-on view.
In comparison to real 4430, note that it's the same width (22") and 1" deeper at the base (ignoring 4430 horn, which sticks out an additional 3-1/4"). No matter, the difference in perceived mass is apparent.
O.K., standing-wave and reflections gurus, does this design really (or theoretically, even) SUCK?
[Excuse the clutter, please. I DO have a life here.... :p ]
My personal preference would be to follow the front trapezoidal shape all the way front to back... :)
John
Giskard:
0) 2.15 uF, 0.60 mH
1) 2.15 uF, 0.76 mH
2) Full range
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:(
Green: original
Yellow: 2,15uF, 1,0mH
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.Quote:
Originally Posted by Guido
Input adjusted for gain to show actual curves.
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....
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
Giskard, I used the real values given from Zilch.Quote:
Originally Posted by Giskard
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.
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
Mr. Widget's latest horn/driver studies for this project are now posted here:
http://audioheritage.org/vbulletin/s...ead.php?t=6050