PDA

View Full Version : problem recone old driver with new diaphram



B&KMan
04-29-2005, 11:49 PM
Hello funtastic members,

It's me again with my weird english and a weird problem...

I Have a 4343 and change last week my caps for HF (2420) and UHF (2405) driver for bigger and more decent caps.
I change in same time my UHF diaphram and soldering with silver soldering...

Before jobs my position Lpas is close to off -4 -5 dB...

After the job... put LPads at max +3 +4 dB and response is just under limit to correspond flat response with the rest of system..

I' choking !!!, I verified circuit resistance, response driver, phase and impulse and not problem appear...

i doubth the flexibilty of the diaphram is realy more flexible after 30 years, What time burning a new diaphram ???

If I compare, the 2420 is appear in same level before and after...

the response curve appear excellent in direct and realy near field (3 inch of tweeter)

anybody have more tecnical info of burning or rodage diaphram effet time

thanks for all...

jean.

:)

johnaec
04-30-2005, 08:17 AM
So you did the crossover change and diaphragm change at the same time? I'd recommend changing only one thing at a time so if the sound changes you know the single cause.

Can you re-install the original UHF diaphragm? If sound is still changed it must be something in the crossover changes.

John

Robh3606
04-30-2005, 08:28 AM
Sure your solder joints are OK?? When you measure the DCR?? How do the 2 diaphrams compare are they the same??


Rob:)

Earl K
04-30-2005, 09:09 AM
Hi Jean

- Here are a few reasons for obtaining lower SPL figures in your UHF gear.
- Review to see if one or more match your particular experience .

(a) The replacement diaphragm is very easily "twisted" in the gap when the central phase plug is removed/reinstalled . This happens if the phase plug is tightened ( or loosened) without being firmly held as the central bolt is turned.

(b) The usage of aftermarket diaphragms. ie non - JBL diaphragms .

(c) The installation of the heavier 075 (2402 ) diaphragms instead of the 077 ( 2405 ) diaphragm . With a high crossover point this combo would measure less loud .

(d) The replacement diaphragm is "binding ( or touching )" in the gap resulting in a loss of efficiency . ( Proper setup of a diaphragm requires a tone generator.)

:cheers:

Robh3606
04-30-2005, 09:34 AM
The 2405 has pins to align the key the diaphram. Is there still enough slop with the pins where it would actually rub?? They look pretty tight. I have never changed one and always thought the pins would at least keep you safe from any gross misalignment. The 2404 however doesn't have pins??? Wonder why??

Rob:)

Earl K
04-30-2005, 10:17 AM
Yeh Rob

- Maybe my point "d" is bogus / hard to know without a 077 open in front of me .

:cheers:

B&KMan
04-30-2005, 12:05 PM
Hi Jean

- Here are a few reasons for obtaining lower SPL figures in your UHF gear.
- Review to see if one or more match your particular experience .

(a) The replacement diaphragm is very easily "twisted" in the gap when the central phase plug is removed/reinstalled . This happens if the phase plug is tightened ( or loosened) without being firmly held as the central bolt is turned.

YES THIS IS THE REASON I CHANGE MY OLD BECAUSE I TWISTED FOIL IN SEARCH THE %$?&? HIDEN SCREW..
BUT I'M SHURE I NOT TWIST CAP.

(b) The usage of aftermarket diaphragms. ie non - JBL diaphragms .
THIS IS A BRAND NAME FACTORY OLD IS BLUE AND NEW IS YELLOW
D16r2405 2 x 150$$ cnd each


(c) The installation of the heavier 075 (2402 ) diaphragms instead of the 077 ( 2405 ) diaphragm . With a high crossover point this combo would measure less loud .
(I HAVE A LITTLE EXPERTISE IN THIS AND I CLEAN GAP , VERIFIE, ALIGNEMENT, AND NO PROBLEM IN THERE... THE MOTOR IS 2405 16 OHMS AND THIS IS A ORIGINAL DIAPHRAM...

(d) The replacement diaphragm is "binding ( or touching )" in the gap resulting in a loss of efficiency . ( Proper setup of a diaphragm requires a tone generator.)

WHAT IS METHOD FOR THIS TEST PLEASE ???

(I RUN WHITE NOISE AND MIC IN NEAR FIELD : RESULT RELATIVELY flat response just around 20K is really higher (6 dB))

I GO TO OPPORTUNITY TO MY SPEAKER IS OPEN FOR CLEAN-UP THE GAP OF THE 2420 AND CLEAN CONTACT OF INSIDE SCREW POST. WOW BIGGER DIFFERENCE IN RESULT... BUT NO LOOSE SPL...

:cheers:

p.s. I put keyCaps for clarity,


thanks for other replys too, I tried to synthetise all questions...


Well, in other response question : YEs I check DCR but diaphram is install so is 6.3 and 6.4 ...

I check polarity too the driver have a positive born at black and I performed, after build, a impulse test response fo check phase... all is good...


----- in other way: the hole pin for centered diaprham is really tight , absolutely non loose, maybe little stress in one side ...

I remark the quality is not perfect : the flat cable for connection post is past on back on the coil and create a little wave in coil this is no good but tw diaprham is same problem...

I check carefully the crossover before and after, a have a kodak pict before and after and nothing wrong connection appear... in other side same affect in two corossover...


Let me know all theorical test procedure ???


MMMMMM I feeling I unbuild the speaker this week-end..... dammed :banghead:

Jean.

Earl K
04-30-2005, 12:47 PM
Hi Jean

Okay - Glad to see you are using real JBL parts / etc. / I had to ask .


WHAT IS METHOD FOR THIS TEST PLEASE ???

(I RUN WHITE NOISE AND MIC IN NEAR FIELD : RESULT RELATIVELY flat response just around 20K is really higher (6 dB))

- Hmmm, I've been looking around for a diagram of JBLs' recommended diaphragm setup guide. So far I can't find it online .

- Jean ; PM me your email address & I'll email you 2 jpgs of JBLs' recommended diaphragm replacement instructions .

- I just looked at them / there isn't anything earth shattering in them ;

- Here's a quick overview ;

(a) You should use a sine wave within a defined range when replacing a diaphragm . The voltage and frequency range are specified for each driver - but that info is not included on the info I have . For a 077 I would sweep from around 4500 hz to 10000 hz . I'm not sure how much voltage to use but since I don't like loud sounds I never use very much .

(b) You are listening for buzzes that manifest themselves when the coil is too close to the top-plate walls ( giving a so-called "binding - coil " ) . JBL recommends using a small rubber mallet to "tap" the top-plate / to center the voice-coil in the gap / if one suspects the voice-coil is binding to one-side of the gap .

(c) NOTE : one really can't use PINK or WHITE Noise for diaphragm centering since these buzzes really won't show themselves properly with those type of sources ( since those sources already "sound" like noise ) .





:cheers:

scott fitlin
04-30-2005, 01:23 PM
You used new and bigger, and different brand capacitors! This will change the sound, and in some cases, new caps need break in time to sound their best. Many people say new caps, especially some Blackgate and other esoteric brands actually sound congested, and band limited till they break in.

This could be your problem as well.

B&KMan
04-30-2005, 01:25 PM
(c) NOTE : one really can't use PINK or WHITE Noise for diaphragm centering since these buzzes really won't show themselves properly with those type of sources ( since those sources already "sound" like noise ) .

:cheers:


Thanks very well,

I sended PM

I know the pink nois or white noise is not good for this but I know is iexist procedure for dual channel analyser test white noise and in Ceptrum function response expose the power of harmonic. this method is developping for fixing complex search harmonic problem in rotary machine. probably I run test if I have a couple of info On jbl factory: exemple frequency resonnance type choose by factory for test: phase, rub, loose particle, level of second and 3 harmonic...

I hope you have this precious info for complete bench test...

:cheers:

Jean.

B&KMan
04-30-2005, 01:31 PM
You used new and bigger, and different brand capacitors! This will change the sound, and in some cases, new caps need break in time to sound their best. Many people say new caps, especially some Blackgate and other esoteric brands actually sound congested, and band limited till they break in.

This could be your problem as well.

I recognize tottally : this point is true !!! but why loose 4-5 db power in dB ??

I run with RTA and mic in really near field and the response is ok but more distance the response is drop fast in top of range frequency (20K)...

Strange... Maybe It is a rodage effect is more signifiant in very high frequency effect...
and I mesure the transition effect in ... step by step...

:)

B&KMan
04-30-2005, 04:10 PM
I have relatively technical review for automatic test quality control driver...
10 pages... many surprising info... I hope to send in this thread If you want...


I found a tech method for test personnalised test for driver... or details specs of JBL for this driver...

Anybody have hint or tips ???

thanks

Jean.

B&KMan
05-01-2005, 02:36 PM
... This could be your problem as well.


Well, the old caps is 1.0uF and rated at 200V

The new is same uF value but 630V

the less sensitity is possible because the voltage rated is higher ???


Jean.

Earl K
05-01-2005, 03:48 PM
Well, the old caps is 1.0uF and rated at 200V

The new is same uF value but 630V

the less sensitity is possible because the voltage rated is higher ???

- No Jean, the higher voltage rating will not effect or attenuate the signal going through it . A lower voltage rating might "clip" the waveform if it was run up to the maximum rating of the cap . That would be heard as obvious distortion .

- Unfortunately in a situation like you presently face which has some sort of mysterious paradox in play / I always have to retrace my steps for the modifications that I made / and then start putting things back the way they were / just to find the actual problem . I'm way more of a "lab" oriented guy than a theortical sort .

- In the end / this method of "empirical" study has proven to be the fastest and most enlightening way to many a resolution / along with a great hands'-on education .

- I'd start by replacing the caps that you swapped out in the UHF circuits . If you don't want to do that , then at least measure the old caps capacitance. Its' possible ( though not probable ) that identical pole positions within each uhf hipass were "short-circuited" by an over-voltage situation. Running 2 pole vs 3 pole might account for the earlier loud tweeters. ( granted, this is a long shot )

- Another somewhat disturbing possibility is that the silver solder you used has actually "cooked" the new caps that you put in and altered the values upwards. Smaller uf values would "steal" db sensitivity . ( Again, quite a long shot ).
- Are you heat-sinking the capacitor side of the lead when you solder the new connections ?

<. :cheers:

B&KMan
05-01-2005, 04:12 PM
-
- Are you heat-sinking the capacitor side of the lead when you solder the new connections ?
<. :cheers:

I put a little touch silver soldering in contact of on hot iron for permit better fast transfer. and just is meld I put iron on opposite side of cap for heat contact and put wire silver in inner side ... meld is fast appear and keep 3-5 second for gond soldering enveloppe and for the flux is go out... and out without moving part , no wind, no vibration... the response appear excellent and of course the old caps is maybe toast and partial short...

but I not put a crocodile clip for prevent damage because the lengt is relatively long (I do not cut). Not caps appears trace of overheat but ....


:banghead: :biting: :banghead:


I'm feeling the rodage of new steel suspender+new coil create a resistive motion and less sensitivity around 3 dB is relatively possible...
Of course if you change all the balance is respected. and just little subtil pump-up volume satisfied for same power result...


But your experience is not go in this way, so , I open speaker again this week for complete test, let me know if you have other idea for test without unsoldering or unmounted unnecessay the driver and caps.

Thanks again for time and effort in mystery ...