View Full Version : Crossover advice for unique driver combo?

09-13-2004, 10:42 PM
OK. I'm building a high-efficiency center channel, to work in a VERY high output home theater system. Projection screen, large grilled "speaker bay" in the front of the room. Current left and right speakers are JBL "Perkins" horns loaded with K140s for woofers, 2370 flares with 2445 drivers for mids, and some supertweeter I don't remember (non-JBL, but similar in response to the 2405). Basically, a JBL-drivered "equivalent" to an Altec A7 VOTT horn-loaded system, with supertweeters. For sub bass, there's a pair of OAP W-horn sub cabinets loaded with 2241 18" drivers.

The main perogative here is to try and match the dynamic and tonal characteristics of the left and right speakers, but in a smaller cabinet (there's no room in the center for another Perkins-size bass horn). So, the idea came to mind, fueled by some available components I have here:

- Bass: Modified JBL 2022 12" woofer (3" voice coil, newer design) with a 15 or so gram mass ring and a softer spider than stock (gives an fs = ~55 Hz, Qts = ~.39, Vas = ~ 2 ft^3, sensitivity = ~ 97.9 dB/2.83V measured into the 4.6 ohm DCR), in a modified early-model L100 cabinet (ported at 50 Hz, f3 of about 54-55 Hz). This woofer should perform very similarly to a 2204... almost the same sensitivity, same F3 in the same box, only the midrange may be a bit more "extended", due to the 3" coil (vs. 4" on the 2204)

- Midrange: JBL 2410 horn driver on a Sammi Sound horn flare- looks for all the world, like a slightly shrunken JBL 2360 flare- has the rectangular throat, the same shape and flare curve, just only 14" wide and about 6" high, to the outsides of the flanges, with about an 8" deep flare length to the throat. Probably an 800 Hz flare, in all probability.

- Tweeter: JBL 2404H "baby cheeks".

The question is crossover topology and frequency points. My first guess was to maybe use existing JBL networks for stuff as similar as I could find... and I came up with the N200B network, combined with a N7000 as a start.

09-13-2004, 11:16 PM
Now, the questions:

-First off, since I'm using a 2404 supertweeter, would the compensation network in the N200B (red box with question mark) be necessary?

-And since it's extremly difficult to get a 16 ohm L-pad around here, how about changing the 5 ohm resistor (in the green box) with an 8 ohm, and replacing the L-pad with an 8 ohm version? That'd supply the input stage with what would ostensibly, be very close to the same impedence (roughly 6.5 ohms, including the 2.5 ohm series resistor in the green box) to the original values.

I know I'll have to change the Zobel on the LF section of the N200B to work with the 2022, but otherwise, I think with some tweaking of crossover component values (probably using a slightly bigger inductor), I can probably get it to balance out.

Now, the OTHER potential solution- using the modified 3133 network, as Giskard proposed for the L300/4333. Might this network work, mostly un-scathed, with the aforementioned retuning of the woofer section for the 2022H?

09-13-2004, 11:19 PM
Now, that still leaves the problem of not having the 16 ohm L-pad. How much would it probably muck up the midrange response, to use an 8 ohm L-pad in the modified 3133 network?

Otherwise, I will probably take the general topology here (crossover points, slopes, etc) and put it into LEAP/LMS... and generate a new network, with the drivers measured. It's just that it'll take a long time before I'll be able to get over to the measurement setup... and really, if one of the above methods will work, it'd certainly be OK with me...


Mr. Widget
09-14-2004, 12:37 AM
Originally posted by GordonW
Now, that still leaves the problem of not having the 16 ohm L-pad. How much would it probably muck up the midrange response, to use an 8 ohm L-pad in the modified 3133 network?

Check out Parts Express. They have several 16 ohm L-pads. The 100 watt 260-261 is only $5.25 retail.

As for whether or not one of those circuit topologies would work, I couldn't tell you as I am unfamiliar with many of your components.


09-14-2004, 04:37 AM
Hi Gordon!

I'm not sure if this really helps you but....

I was trying to build a 4425 type speaker with
a) partly JBL components
b) non JBL components
for a friend who doesn't have the money for the original components.

I started with the original 4425 network and spent 3 weeks with tweaking. Then I gave up!

I then started from scratch with the networks and after another week I got it. BTW a nice little speaker looking and sounding close to original :cool:

What I want to say is that IMO tweaking around is more work than starting from scratch.

09-14-2004, 06:01 AM
Hey Gordon

You are talking HT so more than 50% will come from the center channel. It's really the other way round. Your center channel is what will set the "tone" for the system. On the compression driver can a 2410 handle what you are asking?? Do you have another 2425/2426 arround or better a 2445?? Are you using passives on the rest or biamping?? How big is the space?? You aren't doing a perforated screen??


09-14-2004, 06:35 AM
2370 flares with 2445 drivers for mids:confused: Are you using 2"-1" adapters here?


09-14-2004, 05:36 PM
Oops, I meant 2382 flares on the 2445s... not 3270s. The little Sammi Sound flares I have, are basically "slightly shrunken" 2370s...

As for the drivers- now, the 2022 has pretty decent midrange bandwidth (light cone, 3" coil). I'm thinking I may just cross over the 2410 at like 1500 Hz, second or third order. Then, bring in the 2404 at about 8KHz, second or third order. The crossover in the cabinet will be all passive, and will run from the center channel output of a NAD T753 receiver, in "small" mode (highpass crossover at 80 Hz). That's probably a good, solid 90-100 watts into the 6 ohm impedence of the 2022. With a sensitivity of 98 dB (with 100 watts, that'd be a 118 dB max output fullrange, bi-amped it might just hit 120 dB on peaks), I'd think it'll be loud enough...

Also, I kinda have to use the 2410. A 2425 will not fit in the cabinet (the midrange horn flare is jammed up at the very top of the cabinet, where the outside of the 2410 almost already hits the top cabinet wal)l. I'm thinking, if I limit the bandwidth going to the 2410, the woofer and tweeter can "help" it handle power. Even if I have to go 2KHz between 2022 and pseudo-2370/2410, it should still voice reasonably well compared to the K140 and 2382/2245, I'd think...

Basically, the midrange horn is only going to have to "bridge the gap" between what the woofer can do and what the supertweeter can do.

The screen isn't perforated- the guy has a "speaker bay" below the screen, with a large grille-clothed panel over it. Worst think it may do, is require turning up the 2404 a bit higher than normal, I'd think...

The space is big, but not humongous- about 20 feet wide, 40 feet long, with a 16 or so foot ceiling (two floors worth of height). I'd think the 2410 should be able to cope, if I work with it as above...

Given I've gotta deal with a modified bass driver, I'm thinking more and more, that I may just use the general topologies (slopes) for the various drivers from the modified 3133 schematic, and just play with the frequencies and component values in LMS, starting with a W/M crossover of about 1500-1750 Hz (maybe higher if it measures and sounds good) and a M/T crossover of about 7000-8000 Hz. Probably just better to bite the bullet and make it from scratch, as Guido's experience suggests...


09-14-2004, 05:38 PM
Also, if the 2410 can't handle it, I've got a 2460 here. I KNOW that one will... I'll just have to get an aftermarket phenolic element for it, since JBL no longer has anything for those...