Defect capacitor on JBL 4344 ??

[mech986]

One question please: Can "Mylar (polyester) film caps with small polypropylene bypass" capacitors produce this liquid when they are dead, like chimical capacitors ?

Or, if it possible, I will find other crossover filters with "up to date" components...
What do you think ?

Thanks for the kind words.

In my experience (IME), typical film/film-foil/metallized film capacitors will not leak or drip like this, assuming they have the typical end cap seals of epoxy. Much older JBL film caps were encased in a cardoard tube and wax impregnated. High heat from lots of signal current could theoretically heat the cap hot so the wax version may show signs of melting but I've never seen that.

The typical polyester epoxy closed caps are usually rated at 50 deg. C., 85 deg. C., and up to 125 deg. C operating temperature so I doubt the epoxy end caps or even the cap itself is the source of the mysterious drip material, at least for modern caps.

Regards,

Bart

upon yet another look at your original picture, it does appear that the large film cap and possibly the small 3.0uf film cap in question have an endcap material, and the endcap material on the 3.0uf unit looks damaged. Why is the question - excessive current? Heat from a soldering job? Get us a pic of the other end of the 3.0 uf cap and we can then compare.

If you have any question at all, then desolder the cap, and measure on a cap meter. But at that point, like above, I'd recommend replacement to be sure and safe.

[Madien3388]

Thank you very much Bart. Indeed, only this endcap material seems to be damaged, and I agree with you, "high" temperature or irregular glue may have done this. The other endcap is OK as seen in this picture.
Having such endcap could have changed the capacitor value?
If this shunt capacitor was dead, what should I notice in the sound ?

I'm going to buy the capacitor, just in case first (I only did basic soldering when I was at school, I have no significant experience and no material). Which ones to choose : Kimber or Auricap ?

Thanks again, it is great to have JBL fans all over the world

Damien

[Ian Mackenzie]

Both brands are ver similar and I suggested them as they are similar to a paper and oil (without the coloration and do not exhibit the brightness of the trendy polyropolyne metalised capacitors) and would blend well with the mylar that are used elsewhere in the crossover.

You need some tools.

A sharp short knife to gently cut and lift the hot melt glue.

You will find the glue will lift quite easily but take care.

A solding iron 50 watt, solder wick or solder sucker (mechanical) to remove the solder on the joint (so not heat and force)

A hot melt glue gun. Cut the leads to length and strip the insulation and ensure the parts fit the pcb holes before applying hot melt glue.

Apply a modest amount of glue and insert the capacitor. The glue will cool in a few minutes. The solder leads to the pcb foot prints with resin core solder.

The board should clip apart and the terminals should allow you to remove the to pcb to work on it.

Use an strong light and take your time.

Ian

[Madien3388]

Thank you very much Ian. Now I have to be sure to translate correctly your pieces of advice, it is very technical for me

Here is my conclusion following all your pieces of advice :

The glue under all the 4 capacitors on this filter is the same (same colour, same aspect). All capacitors seem all original JBL ones (Mexico). So I think this capacitor was damaged when the filter was originaly built. The endcap was damaged by soldering, glue ...
As all speakers (internaly and cosmetics) are absolutely in MINT conditions, previous owner was as at least as crazy and maniac as me and took a lot of care of them. So, if the capacitor was really damaged, he should have noticed it in the past after so many years, shouldn't he ?

As it is always annoying to have little default, I'm going to buy spare capacitors and wait until I have done deeper listening tests before replacing it. It seems reasonable.

Is it possible to test the capacitor without unsoldering it from the filter ?

Kimber and Auricap are very similar but not equivalent: will the speakers sound differently with them ?

Many thanks again for your help.
Damien

[Ian Mackenzie]

Are you asking for advice?

Plese read my previous post

The read the Sonic Craft web pages

[grumpy]

Is it possible to test the capacitor without unsoldering it from the filter ?

It is -possible- to test the entire crossover... the question might be, is it possible for -you-?
It's not trivial, and requires a both a measure of instrumentation that is not
particularly common and a fair bit of experience.

If they were mine, I probably wouldn't do anything. Have you asked Chas about his
stock spare parts? It seems that the priority is having them -look- 100%... at least
that is the impression that I am getting.

If you are planning on mucking with the caps, I'd suggest you replace them in pairs
(one on each L/R) crossover in the same location/value, so you don't have a
different itching concern to keep you awake at night.

Best regards,

-grumpy

[Madien3388]

Thank you for your help !

Ian, I'm going to "learn" what you have told me in your previous message, so that I will be able to intervine on the filter (I think it is not so complicated to solder the capacitor, but indeed it has to be done carefully).

During this "learning" phase, I'm going to buy the spare capacitors and I will do many listening tests with friends of fine to find out if the speaker's sound differs from the other speaker or not.

As said previously, there might me nothing on this capacitor as all capacitors of this filter seem to have the same "age". I think, is is an assumption, that the capacitor was llittle damaged when it was originally mounted in the filter 25 years ago. I do not think high current could damage only one endcap of only one capacitor. Entire capacitor would have been damaged and probably other components would have been damageed too.

I really want to learn how to replace some components on the filter, as I will probably need to do it in the future (the filter components won't last forever, will they ?...).

My questions are more : is it really recommended to replace components ? does it change the sound significantly ? Does it improve the speakers ?
(I have seen the Clark experience on 4345 speakers and as said in the post, the modifications were made for Clark project only and may not be appropriate for other persons).

Grumpy, I have been told that it is possible to test the capacitor only. Indeed, it will be too difficult for me to test the entire crossover... Do you know about testing only one capacitor (without dismounting it)

Thanks a lot!
Damien

[grumpy]

I have been told that it is possible to test the capacitor only. Indeed, it will be too difficult for me to test the entire crossover... Do you know about testing only one capacitor (without dismounting it)

One can test for value and leakage. To do a simple/adequate test with
hobbyist level equipment (DVM with capacitor reading) would require lifting
at least one leg from the circuit board.

In-circuit testing would require a signal generator, oscilloscope, and a few
years at university. You could check that it wasn't shorted out (assuming there
are no inductors or low value resistors in parallel), and that it passes
signal (not open to AC waveforms).

Leaving the boards alone (if differences between speakers is not heard) is
a good path. Replacing worrisome elements, in pairs, on both speakers, is
also a good path. Do keep in mind that a soldering iron can easily ruin
the circuit board, damage parts, damage you... and that removing parts
is much more difficult than installing them (adding to Ian's brief instructions
and admonitions).

Hope all goes well and you get much satisfaction and enjoyment from your
speakers!

-grumpy

[Madien3388]

Many thanks Grumpy. Indeed, it is a little risky ...

A friend of mine told me that this "shunt capacitor" could be linked to the crossover of the horn. In the worse case, if its value has changed, it could have modified the crossover which could result in damaging the horn.
So I want to be sure that it is OK.

I have been adviced to call a shop near my home where people have good experience in this kind of operation. I will remove the board and ask them to measure the value of the capacitor.

Damien

[Madien3388]

A friend of mine has just told me that he has a LCR meter. I am not an expert at all in this, but according to what everyone told me, even with this device I still need to remove one leg from the capacitor, don't I ?
Thank you very much
Damien

[Mr. Widget]

...even with this device I still need to remove one leg from the capacitor, don't I ?

Yep.

From looking at your pics I'd assume the damage was done during manufacturing... had anyone noticed earlier it could have been replaced under warranty. It may not actually affect performance though.


Widget

[Ian Mackenzie]

I would suggest your conclusions are assumptions.

A lot of people would never remove the woofer least know a strange looking capacitor and even do nothing about it

The worst case senario is the capacitor value may have drifted and most likely reduced or open circuit.

The effect would be the horn filter would be non optimum but it would not damage the compression driver.

One simple test is to remove the capacitor and test its value with a suitable meter. Its is measures within 5% I would say its good. Compare to the other one.

Decide as you wish.

But I can tell you significant improvements can be made by bypassing the biamp switch and using better capacitors. The simpler way to do this is have someone assemble a charge - coupled newtork and remove the stock network and put it in a safe place.

[Madien3388]

You're right Ian.
I'm going to find someone to help me to measure the value of this capacitor. So I will know if it is good or not.

Regarding the improvment of the filter by bypassing the bi-amp switch, I would be ready to try this but only if I can keep the orignal crossover filter. I'm not sure I have understood well, but the charge - coupled newtork you speak about would replace the entire original network ?
That means that I have to find replacement networks so that I can tweak them and keep the original networks in a safe place (in case of resell).
That is correct ? Does it need do remove the little plate (where 4344 is written) in the front of the speakers to remove the crossover filters ?

And, where to find such networks and do you know someone trustable to do that ?

And again many thanks for your help, I really appreciate it
Damien

[Ian Mackenzie]

You're right Ian.
I'm going to find someone to help me to measure the value of this capacitor. So I will know if it is good or not.

Regarding the improvment of the filter by bypassing the bi-amp switch, I would be ready to try this but only if I can keep the orignal crossover filter. I'm not sure I have understood well, but the charge - coupled newtork you speak about would replace the entire original network ?

Yes
That means that I have to find replacement networks so that I can tweak them and keep the original networks in a safe place (in case of resell).

That is correct ? Yes Does it need do remove the little plate (where 4344 is written) in the front of the speakers to remove the crossover filters ?

No

And, where to find such networks and do you know someone trustable to do that ?

It depends on how patient you are (its takes a while to organise) .Member 4313B has been kindly assisting some members in his own time. I have made the odd network from time to time.

And again many thanks for your help, I really appreciate it
Damien

See above

[Madien3388]

See above

Thanks Ian.

So I will test the capacitor this week.

I'm not in a hurry to upgrade the crossover filters of my 4344. But I think it could be great to take advantage of these amazing speakers with the benefit of up to date new and modern filters.
If you have information on this in the next days or months, of course I will be very interested.

I will let you all know when I will have tested the capacitor. I hope everything will be fine.
Damien