jbl 2426h problem!!!

[vince]

Hello everybody,
i have just found this excellent forum,
i hope that somebody may help, as i have a problem with my passive networks!
actually i have 2426h drivers, mounted on 2391 horns(with lens) for the mid frequencies, and a jbl E110 8ohm for low medium
i have a 2405h as super tweeter...
the extreme low frequencies are dirved bi amped with 2 15" woofers...
all of this is installed in an original Jbl 4350 cabinet...
my main worrie is that the crossovers i make never sounds good!
actually i always have the same probvlem with the 2426, it has an annoynig resonnance around 2.5k to 3khz...
i used a new 18db octave filter with a compensation for the top octave (with 43micro farad capacitor) but nothing works!
i still have an irritating sound...i really can't manage to make those 2426 to sound correct...
please could anybody help me find a way to equalize them ? is that a common problem with the 2426h??
thanks !
Vince

[boputnam]

Originally posted by vince
...the 2426, it has an annoynig resonnance around 2.5k to 3khz... :wave: Hey, vince...

I'll take a first kick at the can: That E110 is my first thought. I wonder of you've got some overlap/doubling with the 2426. That E-series are pretty "bright", being a different application than say the 2122 or 2123. Have you tried swapping out the MF to see if that's it? All this assumes, of course, that your crossover copied JBL design... :confused:

[Alex Lancaster]

If You have the 4350 cabs, get some 2202H´s (right Giskard?), 2440/41´s with 2311 horns, 3107's from the JBLpro tent sale, and You´ll have Great sound and pretty original too.

Alex.

[vince]

Hello,
well what do you mean by swapping out the e110?
well i didn't copied any jbl schematic, as i never found any...i mean not one using a 2426h...
well the E110 sounds very good, my main problem is the 2426, actually i used already some other low med speakers instead of the e110 , but the 2426 still was the worrie...
do you have any idea of schematic i should use with the 2426?
thanks!
Vince

[vince]

well sure i was thinking of getting 2202 and 2440 speakers, but i wanted to try firts to do something good myself before! not easy!
the 2426 should be a good speaker ! i really hope i will manage to make it sound correct!
Vince

[vince]

actually i also have the original 4350 speakers crossover network, i first put all my speakers on it, but had the same problem the 2426 was sounding too agressive around 2khz to 3khz...
i had at that time an other speaker than the E 110 ...so i wouldn't incriminate the E110 even if it's a bit too bright...
hope someone can let me know if he has experiencen such problem using the 2426H...maybe someone has a schematic with compensation?
Vince

[JonFairhurst]

Sounds like a job for a parametric EQ. You can then tone down the 2426, and you can also adjust for any issues in your listening space.

Otherwise you will want an L, C & R connected in parallel, and that circuit is put in series with the driver. It works like this: At high frequencies, the energy passes easily through the C. At low freqs they pass through the L. At the resonant frequency of the L and C (say 2.5 kHz) the frequencies are blocked and must pass through the R only, which reduces the output.

The LC equation is L*C=1/(2*pi*freq)**2.

For 2,750 Hz L=1/(C*298,556,929).

The rub is that this is a relationship. Not an absolute. As you grow L, you shrink C. This sets the Q. I can't remember which direction it goes, but as you push the L (or is it the C) up, the Q goes up, and you cut a narrow band of frequencies. As the Q goes down the frequencies cut are wider.

Finally there is the R. A high value will cut the resonant point (2,750 Hz) completely. A low R will only cut it slightly.

But who knows how this will interact with your crossover? And who knows the Q of your resonance?

I'd start looking for a parametric EQ. Then you can adjust ot taste.

[vince]

Thanks for the rlc tip,
actually i aleready made one(it's called a rlc knock filter) i used a 16mf 0.22mh and a 15ohm resistor) it's working at the right frequency, but it only manage to lower 2db not more...

the parametric Eq sounds right to me, but i would like to know first place opinions from people using the 2426, and know what type of schematics they use...can the 2426 sound flat without parametric EQ?
Vince

[vince]

well i was using a rlc filter with 16mf, 0.22mh and 15ohm, but as what you explained me, to reduce the resonnant peak i have to get a bigger resistor value...i am trying it right now, i put a 50 ohm instead of the 15ohm...let's listen!
vince
PS if anyone has a tip or a schematic let me know!
Vince

[vince]

hello Johnfairhurst,
well i tried a 50ohm resistor and i wouldn't like to speak too quickly, but looks like it sounds incredible! the resonnance has disappeared totally on the 5 cd's i just tried...thanks so much for the help!
Vince

[Guido]

Hi Vince!

I posted a reply to your second thread regarding this problem in the general section.

There are some links to original JBL networks for your driver.

[vince]

Thanks so much Guido,
i'll see that right now!
great image, the 4343 right? i love them! i falled in love with JBL when i discovered them ! i wasn't rich enough to get a pair, so i had to make my own!
Vince

[vince]

20 minutes on this forum and my speakers already sound much better, thanks everybody, defenately a great place to get advices! thanks so much!
Vince

[Guido]

You are welcome Vince!

Yep, not only the image of the 4343 is nice. These speakers are perfect (IMO).

The guys in this forum can help you with every JBL related problem. So they did with me.

[boputnam]

Originally posted by Alex Lancaster
If You have the 4350 cabs, get some 2202H´s (right Giskard?) Yea, I thought that too, but he said he "had" the E110, so I figured somehow his baffle was holding that 10-in, and wouldn't take the 2202. Anyway, it "sounds" like he and Jon sorted it out. :thmbsup:

[vince]

well actually i made a wood adaptator to put the E110 in place of the supposed 12" 2202...but the sounding peak of the E110 has just been solved, i just pot a Zobel circuit, with a 11mf capa, and 7.5ohm resistor...does miracle....
as for the 2426, it now sounds great...but i have to make sure i have no hole in the spectrum....and i am testing it with the scope right now...
Vince

[boputnam]

Originally posted by vince
well actually i made a wood adaptator to put the E110 in place of the supposed 12" 2202... Of course you realize the price for all this is PICTURES!! :yes: We feed on pictures of others gear - it nourishes our JBLust... :D

[JonFairhurst]

Originally posted by vince
hello Johnfairhurst,
well i tried a 50ohm resistor and i wouldn't like to speak too quickly, but looks like it sounds incredible! the resonnance has disappeared totally on the 5 cd's i just tried...thanks so much for the help!
Vince

Wonderful! Replacing a single resistor is certainly less expensive and quicker than buying a parametric EQ.

You might try a potentionmeter, tune it until the resonance just barely fades into the rest of the sound, measure the pot and replace it with a fixed R. It's possible that the 50 ohm R has poked too big a hole in the response, but nixing the resonance is such a sonic relief that it sounds better than it really is.

Anyway, I'm glad I posted. I nearly stopped mid way thinking "who would want to slog through all this filter jive". I now know the answer. Vince would. :)

[vince]

Thanks Jon, i will now be able to ajust all that system!
i am starting to check all the speakers with the scope and ajust ....sounds like something can be done ...but maybe a parametric EQ can really do miracles...
Thanks to you Guys!
Vince

[vince]

Hello Guys!
I have to give you justice! actually after having tested the compensation for the 2426(that is now sounding GREAT) the problem was the E110 much too loud around 1to2k, and very agressive sounding...i replaced it by a beyma 12" i had lying around there, and man it's crazy! smooth and precise!
i defenately have to get the original 2202h!
cheers!
Vince
for info, my RLC notch filter for the 2426h is 27ohm, 0.22mh, and 14.7mf it cuts the 2.7k band of aboutr 8 db...enough to make the 2426 smooth with a sencitivity still good , about 96db
Vince

[hightechhick]

Hello Guys!
I have to give you justice! actually after having tested the compensation for the 2426(that is now sounding GREAT) the problem was the E110 much too loud around 1to2k, and very agressive sounding...i replaced it by a beyma 12" i had lying around there, and man it's crazy! smooth and precise!
i defenately have to get the original 2202h!
cheers!
Vince
for info, my RLC notch filter for the 2426h is 27ohm, 0.22mh, and 14.7mf it cuts the 2.7k band of aboutr 8 db...enough to make the 2426 smooth with a sencitivity still good , about 96db
Vince

Hello everyone,
I have been reading along and I am also fine tuning my newly installed 2426J's behind 2344A horns. I am tri-amping using a recently purchased DBX DriveRack PA (this thing is incredible - 3 way active crossover, RTA, auto and manual EQ) which has a built in parametric EQ. I agree the place to add a notch is around 2.7k. I will try a deeper notch at 8db (I was using around 6 on my first try).

Anyway, does anyone know what "Q" to use on this? I am around 2.0 on my first try. The DBX DriveRack PA lets you change the Q, frequency, and db level on the parametric EQ.

Nice thread - very helpful. HTH

[jerv]

The peak is actually quite broad (from about 800 Hz to 5kHz), so you will need a much lower q-value - even lower than 1. The notch will therefore also interact with the crossover slope and frequency. Ideally, you will need (at least) two notches - acutal frequencies and q-values depending on your chosen crossover frequency and slope.

For a 1000 Hz crossover with 4-order slopes on my 2425's with 2344 horns, I use these values as a good starting point (for other crossover frequencies, the values will be different):

Crossover: 1780 Hz L-R 12
Notch1: 2520Hz, q=0.43, -8dB
Notch2: 4450 Hz, q=2, -5dB

This gives a nice 4th order rolloff at 1000 Hz, and a response of plus/minus 2 dB up to about 18 kHz.

[jerv]

The peak is actually quite broad (from about 800 Hz to 5kHz), so you will need a much lower q-value - even lower than 1. The notch will therefore also interact with the crossover slope and frequency. Ideally, you will need (at least) two notches - acutal frequencies and q-values depending on your chosen crossover frequency and slope.

For a 1000 Hz crossover with 4-order slopes on my 2425's with 2344 horns, I use these values as a good starting point:

Crossover: 1780 Hz L-R 12
Notch1: 2520Hz, q=0.43, -8dB
Notch2: 4450 Hz, q=2, -5dB

This gives a nice 4th order rolloff at 1000 Hz, and a response of plus/minus 2 dB up to about 18 kHz.