Damaged 4430 driver?

[sfellini]

Hi all - I bought a DEQ2496 to check out the room
response with my recently acquired 4430s. There
was a broad dip in the mids and a peak at 6k, so I
thought I'd better look at the response of the individual
drivers.

Microphone placed a couple of inches from the bug
screen.

Looks like something's wrong with the RIGHT driver?
How do I check this out further?

Thanks,
Steve.

RIGHT and LEFT HF driver response:

[duaneage]

I would check out the L-pads. If the L-pad for the midrange is dirty or has a bad connection inside it will play havoc with the response.

[Zilch]

Lemme see if I can replicate that. What's the vertical scale, 5 dB per dotted line?

[sfellini]

Lemme see if I can replicate that. What's the vertical scale, 5 dB per dotted line?

Yes.

[Zilch]

Well, that's not where the midrange L-Pad operates. Crank the mid control down on the "good" one, and you'll see. It attenuates the very bottom of the range, not the middle.

At this point, I'd swap the horns with drivers between the two cabinets. If the problem follows with the driver, it'll need service....

If not, it's the crossover, and we may be able to figure out what's wrong....

[DEQ2496 ROCKS! :bouncy: ]

[Zilch]

a) At 2 inches in front of bugscreen. Mic tip is in the diffraction slot. (Poor microphone. :( Volume DOWN!)


b) At 1 Meter.

Controls at "0" both cases, RTA on "Average."

:hmm: Measure each at 1M, please....

[sfellini]

Measure each at 1M, please....

Hi Zilch et. al.,

This is at 1m., on-axis with the bug screen. Only one
channel getting signal at a time. The 2235 is obviously
there.

The LEFT is still smoother than the RIGHT in the treble.

What do you think?

Steve.

RIGHT and LEFT 1 m. response (sorry about the
wretched photography):

[Zilch]

Here's what happens if I turn HF boost full on with MF at "0."

That's the legendary 4430 HF response. I can dial it up 3 dB or so above flat (though I don't hear any difference when I do that :p ).

I'm not seein' good response in yours above 10 kHz. You're down 10 and 15 dB at the top.

Try that. If you have control of both mid and high with the L-Pads, but can't dial them flat, at least, I'd be suspecting the HF drivers need to go in for test and service. Both of them.

If your controls don't DO anything, there's trouble in the crossovers.

Having the 2235H's running is not influencing the region of interest here.

Gotta hold the camera REAL steady to get those pics. Tripod is best.

Oh, and enable location display in your profile, please. Click "User CP," above, to do it. The forum likes to know where members are.... :)

[sfellini]

I'm not seein' good response in yours above 10 kHz. You're down 10 and 15 dB at the top.

Try that. If you have control of both mid and high with the L-Pads, but can't get response similar to mine, I'd be suspecting the HF drivers need to go in for test and service. Both of them.

With HF at the max level, looks like it's still 5 dB down (this is the
LEFT driver at 1m). Who do I send the drivers to? JBL Pro's site
has the info I need, I guess?

Thanks for your help Zilch. Guess it'll be a few weeks (months? :()
before I'll be posting about these again. Sigh.

Steve (in DC!).

[Zilch]

JBL Pro website will lead you to your local authorized service center. They can take care of these for you in a couple of days probably.

Take them both in for evaluaton.

Tell us what they are once you remove them....

[sfellini]

Tell us what they are once you remove them....

I've removed the six screws which attach the 2344A to the baffle.
It looks pretty well stuck to a gasket. How does one "gracefully"
remove the horn? Crowbar?

[Zilch]

A putty knife, painter's "all in one" tool, spatula, etc., gently break it free.

There's a gasket behind there that's holdin' on. You want it to come off intact, if possible....

[Lancer]

I've removed the six screws which attach the 2344A to the baffle.
It looks pretty well stuck to a gasket. How does one "gracefully"
remove the horn? Crowbar?Dynamite... T.N.T...
Plug the ports and crank AC/DC real loud... the 2235H should pop the horn/cd right outta there...

[GordonW]

Dynamite... T.N.T...
Plug the ports and crank AC/DC real loud... the 2235H should pop the horn/cd right outta there...

:rotfl: "For Those About to Rock" should do it.

Not really "recommended procedure", alas... :D

Regards,
Gordon.

[Guido]

Not really "recommended procedure", alas... :D

But BIG fun :bouncy:

[Lancer]

Oops! My fault! I goofed and posted this in the Technical Help forum! :o: :uhmmmm:

[sfellini]

Tell us what they are once you remove them....

Not having any AC/DC, I resorted to a putty knife. Worked nicely. :)

Steve.

[Zilch]

Those are the correct OEM drivers for late production 4430's, looks like.

We'll soon know if they are O.K. or not. You probably have several factory service centers there.

A pic of the room you're playin' them in would be helpful, too. There's members here whose specialty is room alignment. :)

[sfellini]

We'll soon know if they are O.K. or not. You probably have several factory service centers there.


I just got a call from the local JBL service center and he says the
drivers "sound" ok to him. I asked if that meant they measured
within spec or not and he said he didn't have any special equipment,
and just does a frequency sweep and listens.

I guess this isn't what I expected :blink:. I did want them measured,
right? Isn't that what these guys do? Should I try someone else local,
or ship them off to a service center that someone can recommend?

Steve.

[Regis]

I'd have to say (from a service center perspective, as I used to be a technician) that a JBL Service Center is likely well-equipped to repair your equipment, but as far as actual measurements, that's almost in the realm of the laboratory. I understand that you're trying to set up optimum room eq using the equipment you have on hand, but even Zilch's equipment shows minor differences between the left and right (maybe I'm wrong and feel free to correct me, just playing the 'Devil's Advocate').

Have you tried switching the two as Zilch suggested? Secondly, was there a noticable difference between the two before you got the measuring equipment? If you didn't have the equipment, is the difference noticeable? Zilch did point out L-pads and X-overs as possible culprits and these would have some contribution. I would do an A/B test with both speakers L-pads carefully labeled, to see if there's any differences as well. As far as the passives, try measuring the individual components in the x-overs and see if there's any glaring differences. You'll have to pull the caps to measure capacitance if you can, I don't think you can measure that with the caps in the circuit. Just some thoughts and good luck.

[Zilch]

While I give some credibility to a tech who works with these all of the time saying, "Sounds fine to me," I doubt he could hear a difference in the frequency range of interest here. Ask him to open them up and verify that they have JBL stock diaphragms in them, at least, and that there is nothing apparent amiss in there. For a nominal charge, he might also clean the gaps for you while they're open.

At risk of sending you on a wild goose chase, I'd then call the other nearby JBL service centers and inquire if any of them have means to compare the performance of your drivers to known good ones above 10 kHz.

Another option would be to tell this tech to go ahead and put new diaphragms in them. That'll be about $300 for the pair, probably, and may not solve the problem. At least you'll know the drivers are to spec then, though. I've seen other members here recommend this as standard practice with any acquisition of used compression drivers.

The one case in which I did that here, I concluded there was probably nothing wrong with the old diaphragms in the first place. I did have success with returning a different pair to the JBL factory for evaluation and rehab, but that was after I had obtained definitive performance measurements by Mr. Widget indicating they were not up to spec.

I'm convinced there's a problem somewhere there. Once you get them back, move that "Int/Ext" switch on the crossovers back and forth a couple of times. It may be that the contacts of that switch are dirty, is all....

[edgewound]

A real easy first test to find out if they're JBL diaphragms is to measure the DC resistance of the voice coil with a ohmmeter. D8R2425 should measure 3.2 ohms dcr +/- 8%. This is what's in a 2426H driver. Maybe it's a JBL diaphragm but possibly 16 ohms? which would measure 6.2 ohms dcr +/- 8%. If it's an aftermarket 8 ohm it'll likely measure close to 8 ohms dcr. Lotsa people all over the place are peddling aftermarket diaphragms and claiming they're identical to the JBL...well....THEY'RE NOT!!!! They sound different....darker..usually because thery'r pulling less current due to the higher dcr.....whew!!


I know...those 4430's are big and heavy.....but it's best to have them tested as a system...visually and sonically...by someone willing to open up and take a peek to check the innards. I'd venture to say that most Service Centers don't have sophisticated measurement facilities...me included, but I do have an RTA and test mic that works ok.

Hope this info helps:)

[Don Mascali]

May I ask Where you sent them?

:D I am near you and inquiring minds want to know.

[sfellini]

A real easy first test to find out if they're JBL diaphragms is to measure the DC resistance of the voice coil with a ohmmeter. D8R2425 should measure 3.2 ohms dcr +/- 8%. This is what's in a 2426H driver.

Thanks for the suggestion. I picked up the drivers yesterday and
was able to measure them. One is 3.5 ohms and the other is 3.6.
That's more than 8%, but just barely.

I also mounted the drivers on their horns and sent pink noise
through them and measured (R & L respectively shown below).
This will take the crosover out of the picture.

There is a notch at 4k in the R driver which corresponds to the
dip in the near field response that I first posted. So I think something really is off with this driver :(.

Is this likely to be fixed with a diaphram replacement?

Also for your amusement, I'm including a pic of the test rig here
at Fellini Audio Labs.:p

Thanks for all the help I've been getting...
Steve.

[sfellini]

May I ask Where you sent them?


Hi Don,

Joe Rodgers PAS
12401 Twinbrook Parkway
Rockville 20852
(301) 984-3822
Transducers, Electronics, EON, LSR & 6208, JRX, MPro &
SoundFactor

I would not reccomend these guys based on the condition
of my drivers when I got them back: each had a chip or
two taken out of the magnet :banghead:.

Steve.

[Zilch]

Pull the mic up outta the horn throat a foot or so. Repeat test.

http://audioheritage.org/vbulletin/showthread.php?t=6735&page=1&p=64864

I see your protection capacitor there. Good job! :thmbsup:

[johnaec]

Pull the mic up outta the horn throat a foot or so. Repeat test.I agree! 'Way too many anomolies when jammed in the throat. I'd suggest 1-3 feet, though the farther away the more likely getting some room interference.

John

[sfellini]

Pull the mic up outta the horn throat a foot or so. Repeat test.


and


I agree! 'Way too many anomolies when jammed in the throat. I'd suggest 1-3 feet,

Ok, at 12 inches the notch is gone and they look pretty close (R, L).:applaud:

Thanks!

Next they go back into the cabs.

[johnaec]

Wow - that's about as perfectly a matched pair as I've ever seen. It also looks like you've got a real smooth standard 6db/octave bi-radial rolloff, (5db graph divisions?). If the crossovers are in spec, I'd think they should sound pretty good!

John

[Zilch]

Next they go back into the cabs.Nope. Now run a wire to them from one of your crossovers. Substitute an 8-Ohm resistor for the woofer in that cabinet. Test both drivers on that one crossover. Then repeat with the other crossover. Small wilre nuts will facilitate this process....

Click that crossover "Int/ext" switch back and forth a couple of times to be sure dirty contacts there aren't a problem. Maybe test first, then click to see if there's a difference. Set controls at "0".

This is SCIENCE now at Fellini Labs.... :p

[duaneage]

I am betting on cold solder joint somewhere or even the L-pads. Try both drivers on each crossover. If both drivers have the same notch on one xover eureka!


Vibration inside the cabinet (there is plenty of that!) may have shaken a solder joint loose. A resistor could be damaged, check the resistors on the board.

[Zilch]

Notch is not the problem. That was an artifact of having the mic too close to the throat.

The problem is (was) lack of HF extension....

[sfellini]

...run a wire to them from one of your crossovers.

Quick question: do I get to the xover by removing the 2235 or from
the back?

[Zilch]

For what you need to do, just remove the woofer.

I am sorry the idiots chipped your magnets. Of all people, they should know how to handle these.... :(

[sfellini]

Now run a wire to them from one of your crossovers. Substitute an 8-Ohm resistor for the woofer in that cabinet. Test both drivers on that one crossover. Then repeat with the other crossover.

Click that crossover "Int/ext" switch back and forth a couple of times to be sure dirty contacts there aren't a problem.

I've done the above tests. First of all, identical results were obtained
before and after clicking the int/ext switch for each crossover (not shown).

Here are the results for LL0, LR0, RL0 and RR0, where the first letter
is the Left or Right crossover, the second letter is the Right or Left driver,
and the digit is the HF setting on the xover.

Note that for a given driver, there is more output above 8k in the Right
crossover. HOWEVER, with a +1 dB boost in the HF control, they look
pretty similar (see the following post).

Also, for a given crossover, the Right driver seems flatter in the 3-6k
region (I didn't try compensating for this with the MF control). I should
try this because previous results show the drivers are identical without
a crossover.

Together this means the LL combination should give the worst response
and the RR the best, and this is in fact the case (no wonder the two speakers measured different).

[sfellini]

HOWEVER, with a +1 dB boost in the HF control, they look
pretty similar.

Here are some more FRs, showing the effects of boosting the HF control.
One complication: on the Right crossover, the +1 dB stop is as far as the
control will go, so it's probably really +2 db?

LL1, LR1, RL1, RR1:

[Zilch]

Well, you're lookin' quite good, actually. Compare to reference:

http://audioheritage.org/vbulletin/showthread.php?t=6735&page=1&p=64993.

If you pull the mic back to 1M, I think you'll find the response to be smoother.

So, your right crossover has been boosting the HF about 1 dB, because the control is not aligned with the front panel markings.

I just checked, and the control board is mounted from the front panel with screws under the foilcal. You might be able to reach in from behind and rotate the HF control that little bit if it's slightly loose.

The knob itself is 3/8" deep, and can be pulled off from the front with an IC extraction tool which will hook under the rear edge of the knob. That'll let you put it back on in the correct position, if that's important to you. If you damage the knob, though, replacements do not seem to be available from JBL parts.

[sfellini]

Well, you're lookin' quite good, actually. Compare to reference:

http://audioheritage.org/vbulletin/showthread.php?t=6735&page=1&p=64993.


Yeah, it was pretty ugly. I think much of it was due to incorrect
measurement methods. Having the drivers on the floor and
having the mic hang over them is way more precise than trying
to point the mic at the speaker on a makeshift boom.

Zilch, you've been a great help, and it was a good learning
experience. Way more to go... :blink:.

Thanks again,
Steve.

[Zilch]

Stand and mini-boom I use was like $25 at Guitar Center. Nicely adjustable for height and angle.

It looks like you've zeroed in on part of the problem, at least, and perhaps corrected some of the rest of it in the process.

Remount the drivers in the cabinets and do a final test. :D

[sfellini]

Remount the drivers in the cabinets and do a final test. :D

Got my tripod & boom.

So to conclude this thread :p, here are the final results
(L & R at 24"). And to answer the question asked in the
thread's title: "no" :D.

(Now, how does one accurately measure bass response
with the RTA? :blink:).

Steve.

[Titanium Dome]

This has been helpful to me as well.

Thanks to all participants for a constructive thread. :thmbsup: