2245 questions - buzz

[valbird]

I recently picked up a couple of JBL 4645 boxes with 2245 drivers in them. I'm trying to check things out, and here's what I've noticed:

-surround replacement done less neat than I like
-one box buzzes when the cone is tapped
-almost all the mount screws have their phillips heads cammed out

So I decided to pull the driver on the box that buzzed when tapped. Had to drill out 4 mounting screws. Noticed that the front gasket (thick rubber) was not aligned with screws, and was of two different thicknesses.

For the drill out, I kept a vacuum near the drill bit to capture the shrapnel. It took a while because I didn't want to chance breaking the bit or slipping.

Expected to find a screw or something on the back side, making the buzzing. I did not. I did find a piece of particle board.

Out of the box, the driver made the normal damped thump when tapped. Putting some bass into it, I found that it was distorting significantly with a manually induced excursion of about 3mm. Seems like too little, but I haven't checked Xmax for this driver yet, etc. Since everyone is home, I have stopped my play for the day.

I need to get some more mounting screws, as all but one I would not put back in, and 4 have been destroyed by drilling. As a result, this box will be open for a while.

Does anyone have some suggestions on how to test this, and determine if I have a problem?

As near as I can tell, the less than ideal bad foam job is not skewing the cone off center, but I could be wrong.

I did run this driver through the crossover (DCX2496) and using a cutoff of 100 Hz there was audible buzzing. However when I gave it broadband music, I could not readily detect any buzzing. For now, I will assume that the piece of wood (15mm x 3mm x2mm) may have been captive somewhere making that noise.

Thanks.

[pos]

Putting some bass into it, I found that it was distorting significantly with a manually induced excursion of about 3mm. Seems like too little, but I haven't checked Xmax for this driver yet, etc.

Hi

Xmax is quite higher than that, and I found this driver to be particularly silent during large excursion.
Did you allow the air to escape the rear vent of the magnet during your test?

In my tests, even with lots of room around the rear vent, I did hear the noise from the air in the vent before any other distortion from the moving assembly itself. My drivers had the square foil cal on top of the round vent, which does not help with air noise...

[valbird]

Venting was not an issue. I guess my concern is that there may be VC problem or an alignment problem. My recall is that Xmax is well over .25 inch, but I have not looked that up yet.

I have not compared it to the other driver, but since that one is in the enclosure, it would be harder to do.

[maxwedge]

xmax is 25mm but but checking that with your hands is not really an acceptable method of testing. You really need to apply a sine wave signal to it either from your pc, a cd or a signal generator. Just hook it up sitting on a table sitting and listen for voice coil rub, which is what it sounds like to me. These aren't long excursion drivers even though 25mm seems like a lot but remember that that is the max as in failure point! I have a pair but don't use them as subs (bottom in 4 way) and when I crank them they really don't move a whole lot, although a lot of air is pushing out the ports.:)

Edit: I meant xmech.:)

[valbird]

I agree. My 40 yo Wavetek generator has taken a run somewhere. I need to find generator / sw to use, and about .5 Hz sine with 500 Hz or 1kHz imposed on top of it will help me see what is going on.

[pos]

the calculated xmax is around 7mm one way.
the xmech is much higher than that, but remember this driver is quite fragile.

[valbird]

the calculated xmax is around 7mm one way.
the xmech is much higher than that, but remember this driver is quite fragile.

I'm seeing about half that.

I have another concern, and that is while the re-surround was not the neat job I do, or am used to seeing, the dust cap appears to be original. This has me wondering if the cap was removed and the VC shimmed when re-surrounding.

The previous owner did say that he re-surrounded the speakers with JBL OEM parts (he does business with a large JBL distributor), but that he never tested them afterwards.

Somehow I have the feeling that I might end up re-surrounding this driver...but I'd like to verify what is happening, and if I can why, which is why I am calling on the collective wisdom here. I've re-surrounded probably a hundred or more drivers, never once had a problem, and I have no experience with the 2245 driver.

pos, why do you say the driver is fragile? More so than other drivers?

[subwoof]

the internal air filter foam is made of the same crap as the original surround - it too must be replaced. if it's the original cap, that is a sure sign that the refoam was a kitchen table project. If you take the rear foilcal off ( if it has one ) you can see it.

My refomed speakers have the dust covers removed, new filters and the gap is shimmed while the glue sets. and at this time a good tech can slide shims around the former to see if there is any deformation and can clean out debris.

and BTW - jbl has never sourced the foam to dealers, repair centers, etc.

[valbird]

Subwoof, there is no foilcal on the 2245H. There is a screen.

I cannot determine for certain that the dustcaps came off during the refoam. However, I can say that the dust caps on these are slightly larger than the ones on a known 2245 from 1984.

I can determine that with the existing surround, there is less than a 0.5mm variability in the end of the cone to the basket, which is consistent with what I measure on the other 2245H. Similarly, the cone is symmetrically placed within the basket.

I put a couple of watts of pinknoise, rolling off at 100 Hz, and found that coincident with random higher excursions of the cone, there was raspiness, like when there is a fried voice coil. It is not readily audible at lower levels.

Furthermore, when pressure is placed on the back of the cone at 1 and 7 o'clock, there is immediate raspiness, and at other symmetric positions there is no such raspiness.

I'm thinking that there is some voice coil damage, and that I'll be ordering a cone. Normally JBL dealers won't sell you a cone alone, but I have done probably 100 other drivers, so I'll probably just get a cone kit. Before I do, I would like to prove out the cone. One way that occurs to me is to put a scope on it, and see what happens when things break up (raspiness). If the voltage spikes then, it could be a voice coil, or a connection to it opening up.

I'm still open to any suggestions or ideas. Thanks.

[pos]

pos, why do you say the driver is fragile? More so than other drivers?

Modern drivers have an xmech much higher than their xmax, and various protection system (differential driver for modern JBL, double spiders with variable compliance for precision devices, etc...). The 2245H has a very low Fs and basically respond to any signal. As the excursion does x4 when you divide the frequency by two this can be quite dangerous. When used in a BR box this driver really needs a HP filter under its tuning frequency.


Furthermore, when pressure is placed on the back of the cone at 1 and 7 o'clock, there is immediate raspiness, and at other symmetric positions there is no such raspiness.
That looks like a typical case of badly centered refoam job. Try replacing the surrounds, it might fix your problem (if the VC is still ok)

[valbird]

Modern drivers have an xmech much higher than their xmax, and various protection system (differential driver for modern JBL, double spiders with variable compliance for precision devices, etc...). The 2245H has a very low Fs and basically respond to any signal. As the excursion does x4 when you divide the frequency by two this can be quite dangerous. When used in a BR box this driver really needs a HP filter under its tuning frequency.


That looks like a typical case of badly centered refoam job. Try replacing the surrounds, it might fix your problem (if the VC is still ok)

Thanks for your clarification, and I agree that the compliance is very high, and that damaging excursions are possible and likely. When doing reinforcement work we typically ran 48 db/oct HP filtering above the tuning point of the boxes so that we didn't over drive (exceed a given excursion distance).

As for the concept of a badly centered refoam job, I remain unconvinced that the refoam is indeed the problem. The cone in the basket, although that is not a true measure, seems centered, and the dust covers are larger than factory, which suggests that the refoamer may have centered the VC when refoaming.

Even the shimming of the VC is debatable for refoaming. I've probably done a 100 or so refoam jobs, and I've always used shims, but there is a school of thought which relies on the spider and a LF signal to perform the centering. Again, while I haven't used that technique, it has merit, as in the end the spider ends up doing much of the support, with the foam providing a lesser alignment contribution.

So what's wrong with this driver? I don't know. It could even be something like the VC attachment to leads or some other VC problem. For all I know they VC may have been damaged by over excursion at some point, prior to the refoam.

What I do know is that there was interference from a piece of particle board with the cone, partially caused by failure of the particle board at the T-nut. The refoamer did allow in an email that after refoaming, and checking the driver out, the drivers were reinstalled into the box, and were not connected to a signal after that. I found that interesting, because I always have reinstalled drivers, and then tested in the box with a signal generator and also with music.

Anyone have any pointers to VC failure modality and non-disassembly evaluation?

[pos]

You have clearly more experience and knowledge than I do on refoam jobs (I only did a handful, and always with the easy 30Hz auto centering technique).

Have you checked that the foam pad that covers the inside hole of the magnet vent is still in place? These things do rot and tend to scatter in the gap when they do... (better replace them with something else)

[grumpy]

It would seem that if you're unable to find an electrical problem
(I'd suggest a tone generator rather than pink noise, often easier to
hone in on an issue), and a dust-cap-ectomy doesn't produce clues,
there are few options beyond pulling the cone kit.

Signs of shifted magnet? Completely intact tinsel leads (not touching cone
during larger excursions)... that you can excite or prevent the issue by
limiting cone movement manually in certain areas does seem to indicate
an interference issue (bubbled up varnish on the coil?, slightly distorted spider,
bottomed out/mushroomed coil former)...

Good luck.

[richluvsound]

Sorry to barge in ..... can I put a 2245 cone in a 2241 basket ?

Rich

[grumpy]

2240 basket?, yes. Someone else will have to answer your question more thoroughly
re gap, VC size/depth/offset, mag field, frame/cone dimensional compatibility, etc...

[richluvsound]

Cheers Grumpy . There was a list about what fits what ,but I can't find it ...

[valbird]

Got some more time with the driver tonight, and reinstalled it in the box. After installing in the box, what I found was that slight pressure at the 12 o'oclock caused the buzzing again. Changing the fold in the foam, I could alter substantially the buzz pattern, so my conclusion is that the VC in not properly centered. I'll order surround and a dust cap, and redo it. If that doesn't work, then I'll put a new VC in it, assuming that there is nothing wrong with the magnet or basket.

The good news is that I have to drive this thing pretty high until it is perceptible, and even higher until it is measurable. So I can use it for softer listening, or I can just bridge the <100Hz together and run it in to the apparently good sub, until I have the chance to fix it.

Thanks for all the ideas.

After I get this fixed, I have to measure the box performance and decide what kind of EQ I might wish to implement That will be several months away, because there are several intervening projects.

Thanks again, everyone.