JBL crossover issue - polarity? Seeking help. - AVS Forum | Home Theater Discussions And Reviews
Forum Jump: 
 70Likes
Reply
 
Thread Tools
post #1 of 84 Old 10-08-2018, 06:37 PM - Thread Starter
AVS Forum Addicted Member
 
Archaea's Avatar
 
Join Date: Apr 2004
Location: Kansas City, MO
Posts: 11,046
Mentioned: 604 Post(s)
Tagged: 2 Thread(s)
Quoted: 3772 Post(s)
Liked: 3556
JBL crossover issue - polarity? Seeking help.

I bought 13 JBL CBT 70j-1 used at auction.



In way of vetting them, I hooked them all to a tripod and used a mic stand to measure all 13 speakers.

4 of the speakers have a smooth response when measured with omnimic. 9 do not. There is an identical sizable dip at 2Khz on the 9.

Here's a chart of what the speakers frequency response look. See how most have the dip at 2Khz?



Disregard the one very random FR plot in the chart. One of the 13 speakers came with the crossover board melted, so I swapped that board out with the seller, and the replacement crossover board had the same dip at 2Khz as 8 of the others. I just haven't gone back and re-measured all 13 again.

This issue at the crossover point, though I'm not sure how/why? In fairness, my Denon X7200WA doesn't tell me there is a polarity issue when I run Audyssey with these speakers and in the past - when I’ve accidentally wired a speaker out of phase the Denon AVR has told me I have a speaker out of phase. The speakers don't sound bad, and I'm not sure I'd even know there was a problem if I hadn't measured them.

Fortunately, the speakers cleanly come apart into 3 pieces.
  1. Tweeter Array,
  2. Woofer Array,
  3. Cab/Crossover.

I have narrowed the difference down to the cab/crossover I think.

Testing procedure:

I swapped the tweeter array between a good speaker and a bad speaker.

The problem stayed with the cabinet/crossover.

Then,

I swapped the woofer array between the good speaker and the bad speaker.

The problem still stayed with the cabinet/crossover.

When I look at the cabinet/crossover I don't see anything different between the two crossover boards. Furthermore the wiring harnesses all only fit one way, and the speaker wire in the cabinets is cut to exact length with big and small spade terminals which would prohibit anyone from attaching with the polarity backwards - so it doesn't seem possible the wiring is wrong???

Anyone have any pointers? What's my next step in troubleshooting?

Pics follow - more available at request:

Good and bad side by side:



Top part of cabinet/ showing music/speech switch and dispersion switch. All switches in identical positions. This is just to show wiring looks identical



Close up of crossover board connections. See they only attach one way.



Picture of woofer array:



Picture showing the tweeter array only attaches one way. (Disregard the white wire it’s an RGB 5050 LED light I’ve attached to the speaker to mimic AMC Primes red LED look. It doesn’t interfere with anything)




More pics:

Dip FR






Flat FR






Dip FR on Left.
Flat FR on Right


Archaea's 9.8.4 Home Theater Room
(13) JBL CBT 70j-1 | Denon x7200wa | Sherbourn PA 7-350 amplifier | (8) Ultimax 18" sealed subwoofers | (4) iNuke DSP 6000 amplifiers | (4) MB Quart 12" subwoofers mounted direct mounted to Berkline theater chairs BOSS style | Epson 5040UB Projector | Jamestown 144" acoustic transparent 2.35:1 screen w/ Seymour XD fabric

Last edited by Archaea; 10-09-2018 at 05:19 AM.
Archaea is offline  
Sponsored Links
Advertisement
 
post #2 of 84 Old 10-08-2018, 07:52 PM
Senior Member
 
Join Date: Aug 2014
Posts: 449
Mentioned: 6 Post(s)
Tagged: 0 Thread(s)
Quoted: 255 Post(s)
Liked: 238
Nice score!

Man, do you think you could've found a more proprietary, complicated, or unconventional passive crossover to troubleshoot? Sure would be nice if somebody could post a schematic.

WRT phasing of the highs to lows, it looks like the handoff happens at 800Hz, so it's probably not simple polarity.

My first thought was the Music/Speech voicing, but that's centered at 5k, not 2k. Still, might be worthwhile to do some measurements on both good and bad units with the switches on and off.

Besides the obvious electrolytic caps as wear items, I'm leery of the SonicGuard light bulbs. Big transients can change their resistance. JBL wires them in parallel with a resistance, so even if the bulbs open circuit some sound will still come through. But depending where they are in the XO circuit, that might change response shape as well as level? Hopefully measuring across the bulbs, cold and in-circuit, would reveal differences between good and bad ones.

Those "sandbox" power resistors can also go bad, and warrant measuring.

And of course solder joints at both the bulbs and power resistors, as they thermal cycle. Get out the magnifying glass, or just see if resoldering a bad XO turns it into a good one.

Here's some spitballing: extra/wrong hardware in the "bad" cabinets near the XO locations, altering the inductor values? Testing with the XO's connected remotely would clear that up, make some extender wires with a male flag on one end and a female on the other.
LTD02 and Archaea like this.
fill35U is offline  
post #3 of 84 Old 10-08-2018, 08:07 PM
AVS Forum Addicted Member
 
notnyt's Avatar
 
Join Date: Dec 2008
Location: Long Island, NY
Posts: 10,199
Mentioned: 316 Post(s)
Tagged: 0 Thread(s)
Quoted: 3732 Post(s)
Liked: 3708
notnyt is offline  
Sponsored Links
Advertisement
 
post #4 of 84 Old 10-08-2018, 08:12 PM
AVS Forum Special Member
 
Samps's Avatar
 
Join Date: Feb 2014
Location: Pittsburgh, PA
Posts: 2,733
Mentioned: 51 Post(s)
Tagged: 0 Thread(s)
Quoted: 1224 Post(s)
Liked: 901
When you change the speech switch does it effect the dip area? If so, I would suspect the switch. Can you swap a good switch with a bad switch?
Samps is offline  
post #5 of 84 Old 10-08-2018, 09:37 PM
AVS Forum Special Member
 
Join Date: Jun 2005
Posts: 4,892
Mentioned: 14 Post(s)
Tagged: 1 Thread(s)
Quoted: 1193 Post(s)
Liked: 663
Quote:
Originally Posted by notnyt View Post
check the bulbs


This. If one was melted then all were likely run hard. Swap bulbs to confirm, unless they’re soldered in.

--
"In many cases there aren’t two sides unless one side is 'reality' and the other is 'nonsense.'" - Phil Plait
Serious Audio Blog 
DS-21 is offline  
post #6 of 84 Old 10-09-2018, 05:29 AM - Thread Starter
AVS Forum Addicted Member
 
Archaea's Avatar
 
Join Date: Apr 2004
Location: Kansas City, MO
Posts: 11,046
Mentioned: 604 Post(s)
Tagged: 2 Thread(s)
Quoted: 3772 Post(s)
Liked: 3556
JBL crossover issue - polarity? Seeking help.

Quote:
Originally Posted by DS-21 View Post
This. If one was melted then all were likely run hard. Swap bulbs to confirm, unless they’re soldered in.
and @notnyt , and @fill35U , thanks for a practical next step in looking closer into the bulbs.

I’ll measure and compare their impedance. Where do I buy replacement bulbs? Or can I just use automotive bulbs? I websearched sonic guard bulbs and got very minimal information and no purchase options, unless I’m using the wrong search term.

——

The speaker that originally measured and sounded way off from the others — I originally suspected most of the tweeters were blown. However, when I opened it up I noticed it had caught fire and fried. The tweeters seemed fine —- but the crossover board wasn’t! When I replaced the crossover board with one provided from the seller the speaker measured in line with the other eight speakers with a dip.







Even melted the port a wee bit!!! It must have been an actual fire on the crossover board.



It unravels another initial mystery too — why one of the grills inside cloth was discolored grey. The smoke must have poured out the ports into the back side of the grill. I originally assumed it was dust???



The back of the crossover boards have no component on them so no sense in taking them out.
Here’s a picture of the back of the fried crossover board that I replaced.


Archaea's 9.8.4 Home Theater Room
(13) JBL CBT 70j-1 | Denon x7200wa | Sherbourn PA 7-350 amplifier | (8) Ultimax 18" sealed subwoofers | (4) iNuke DSP 6000 amplifiers | (4) MB Quart 12" subwoofers mounted direct mounted to Berkline theater chairs BOSS style | Epson 5040UB Projector | Jamestown 144" acoustic transparent 2.35:1 screen w/ Seymour XD fabric

Last edited by Archaea; 10-09-2018 at 05:39 AM.
Archaea is offline  
post #7 of 84 Old 10-09-2018, 06:20 AM - Thread Starter
AVS Forum Addicted Member
 
Archaea's Avatar
 
Join Date: Apr 2004
Location: Kansas City, MO
Posts: 11,046
Mentioned: 604 Post(s)
Tagged: 2 Thread(s)
Quoted: 3772 Post(s)
Liked: 3556
JBL crossover issue - polarity? Seeking help.

0.9 resistance on both good and bad bulbs.

Did I do that right?

Bad




—-

Good









@fill35U , to measure the sandbox resistors I have to de solder them right?

Pic taken from blown crossover board for reference.

Archaea's 9.8.4 Home Theater Room
(13) JBL CBT 70j-1 | Denon x7200wa | Sherbourn PA 7-350 amplifier | (8) Ultimax 18" sealed subwoofers | (4) iNuke DSP 6000 amplifiers | (4) MB Quart 12" subwoofers mounted direct mounted to Berkline theater chairs BOSS style | Epson 5040UB Projector | Jamestown 144" acoustic transparent 2.35:1 screen w/ Seymour XD fabric

Last edited by Archaea; 10-09-2018 at 06:28 AM.
Archaea is offline  
post #8 of 84 Old 10-09-2018, 07:03 AM
Senior Member
 
Join Date: Aug 2014
Posts: 449
Mentioned: 6 Post(s)
Tagged: 0 Thread(s)
Quoted: 255 Post(s)
Liked: 238
Quote:
Originally Posted by Archaea View Post
0.9 resistance on both good and bad bulbs.

Did I do that right?
Looks like the bulbs themselves are OK. But try measuring at their solder joints on the back of the board, or even better scrape off a little of the green coating to get down to the bare copper of the trace next to those solder joints, and measure there. That way you'll also be checking the solder joints to the bulbs.

The "sandbox" resistors are the big white boxy ones marked "20W8(Ohm)J". You should be able to measure them in-circuit. Check their solder joints, too.

Unfortunately, those bulbs come in different ratings. If it were me, I'd buy a selection, make a simple test circuit with a battery and variable resistor (or use a variable power supply), and just use whichever bulb matched the brightness of one of the originals. There's also the possibility that the two bulbs on the board are different types by design.

BTW, fantastic job at supplying so many pics! It'll take a while to trace out a useful schematic to track down some suspects for the dip. This design is pretty much worst case scenario for troubleshooting.
fill35U is offline  
post #9 of 84 Old 10-09-2018, 07:08 AM
Senior Member
 
Join Date: Aug 2014
Posts: 449
Mentioned: 6 Post(s)
Tagged: 0 Thread(s)
Quoted: 255 Post(s)
Liked: 238
If the board marking is correct, these are probably your bulbs:

https://www.simplyspeakers.com/jbl-s...-bulb-sk3.html
Archaea likes this.

Sources/processing: stack of stuff that if it isn't vintage now, it will be soon!
Amps: stacks and stacks of old iron
Main speakers: big DIYSG
Surrounds: Bose graveyard
Subs: a bunch
fill35U is offline  
post #10 of 84 Old 10-09-2018, 08:16 AM
AVS Forum Special Member
 
jcmccorm's Avatar
 
Join Date: Mar 2001
Location: Madison, AL, USA
Posts: 4,492
Mentioned: 19 Post(s)
Tagged: 0 Thread(s)
Quoted: 597 Post(s)
Liked: 444
I guess it's time to swap more stuff out. Can you swap the tweeters with a flat vs dip board?

You haven't tried swapping a crossover board yet right? (probably the easiest next step).
jcmccorm is offline  
post #11 of 84 Old 10-09-2018, 09:21 AM
AVS Forum Special Member
 
Samps's Avatar
 
Join Date: Feb 2014
Location: Pittsburgh, PA
Posts: 2,733
Mentioned: 51 Post(s)
Tagged: 0 Thread(s)
Quoted: 1224 Post(s)
Liked: 901
The voicing switch may adjust the 2k range. So if those switches are bad or calibrated differently that may be the problem. Should be easy to just wire a bad speaker to a good set of switches to test.
Samps is offline  
post #12 of 84 Old 10-09-2018, 09:54 AM
Senior Member
 
Join Date: Aug 2014
Posts: 449
Mentioned: 6 Post(s)
Tagged: 0 Thread(s)
Quoted: 255 Post(s)
Liked: 238
Quote:
Originally Posted by Samps View Post
The voicing switch may adjust the 2k range. So if those switches are bad or calibrated differently that may be the problem. Should be easy to just wire a bad speaker to a good set of switches to test.
Like I said, I suspected that switch at first. But according to all the JBL literature, the voicing introduces a broad *peak* at *5k*. It shouldn't cause a narrow dip at 2k. Still, easy enough to check.

And when checking, first give it about a dozen vigorous twists off and on to help wipe off any dirt or oxidation. Hit it with some Deoxit.
fill35U is offline  
post #13 of 84 Old 10-09-2018, 12:59 PM
AVS Forum Addicted Member
 
notnyt's Avatar
 
Join Date: Dec 2008
Location: Long Island, NY
Posts: 10,199
Mentioned: 316 Post(s)
Tagged: 0 Thread(s)
Quoted: 3732 Post(s)
Liked: 3708
These bulbs look easy enough to eyeball, however testing components on a board without knowledge of the schematic may not yield accurate results. There may be other paths you're measuring through.
notnyt is offline  
post #14 of 84 Old 10-09-2018, 01:02 PM
AVS Forum Addicted Member
 
notnyt's Avatar
 
Join Date: Dec 2008
Location: Long Island, NY
Posts: 10,199
Mentioned: 316 Post(s)
Tagged: 0 Thread(s)
Quoted: 3732 Post(s)
Liked: 3708
notnyt is offline  
post #15 of 84 Old 10-09-2018, 01:11 PM
eXa
Senior Member
 
eXa's Avatar
 
Join Date: Jun 2006
Location: Norway
Posts: 243
Mentioned: 1 Post(s)
Tagged: 0 Thread(s)
Quoted: 30 Post(s)
Liked: 36
Nothing to add. Just wanted to commend you for properly supplying information and pics so its less of a guesswork trying to help you. Awesome!
eXa is offline  
post #16 of 84 Old 10-09-2018, 01:44 PM
AVS Forum Special Member
 
mtg90's Avatar
 
Join Date: Aug 2011
Location: Millbrook, IL
Posts: 1,948
Mentioned: 584 Post(s)
Tagged: 0 Thread(s)
Quoted: 770 Post(s)
Liked: 1588
What's that SW1 switch do when you flip it? The one connected by the black and yellow wires.

The values attached to that RLC filter and bypass switch look light they might affect that range.

Maybe try and unplug the connector and measure to see if the bypass switch is bad?
Archaea likes this.

Last edited by mtg90; 10-09-2018 at 01:54 PM.
mtg90 is offline  
post #17 of 84 Old 10-09-2018, 01:51 PM
Senior Member
 
Join Date: Aug 2014
Posts: 449
Mentioned: 6 Post(s)
Tagged: 0 Thread(s)
Quoted: 255 Post(s)
Liked: 238
@notnyt , thank you so much for the schematic!

You are also correct about one of those resistors being in parallel with other components- R1 is across L7, which has a DCR of 0R15. And that may be the exact area responsible for the dip: the tank of L7 and C8 makes a notch at 2kHz. Bypassed by R1 and switch "SW". So R1, its solder joints, and SW bear some close examination. Apologies to @Samps , as that may indeed be the voicing switch and at 2kHz.

R2 is fine to check in-crcuit, but probably isn't responsible for the dip.

The protection bulbs are in series, bypassed by C10, not a resistor. So checking across C10's leads should be sufficient to check the conditions and connections of both bulbs all simultaneously. The bulbs only protect the tweeters, BTW.
Archaea and notnyt like this.

Sources/processing: stack of stuff that if it isn't vintage now, it will be soon!
Amps: stacks and stacks of old iron
Main speakers: big DIYSG
Surrounds: Bose graveyard
Subs: a bunch
fill35U is offline  
post #18 of 84 Old 10-09-2018, 06:46 PM - Thread Starter
AVS Forum Addicted Member
 
Archaea's Avatar
 
Join Date: Apr 2004
Location: Kansas City, MO
Posts: 11,046
Mentioned: 604 Post(s)
Tagged: 2 Thread(s)
Quoted: 3772 Post(s)
Liked: 3556
JBL crossover issue - polarity? Seeking help.

Quote:
Originally Posted by Samps View Post
When you change the speech switch does it effect the dip area? If so, I would suspect the switch. Can you swap a good switch with a bad switch?


Thanks for the idea - in order to test this I just hooked my multimeter up to the music/speech switch on both cabs after disconnecting the yellow/black two wire lead. In Music mode there is infinite resistance on the switches in both boxes. In Speech mode there is .6 resistance on one box and .5 resistance on the other box (that delta is likely a accuracy error on my cheapo multi-meter)

EDIT - verified - I began to suspect my multimeter was tool old and too cheap to be accurate - and why start a project like this with all this great help and be giving uncertain measurements. I ran to Wal-Mart and picked up a auto range multimeter for $21 - (I know - big spender) - any rate - when I retook these measurements I still got infinite resistance on music mode, but now 0.0 on the speech mode.

Both switches measure the same (.5 in speech mode on old centech multimiter and 0.0 on new HyperTech multimeter)

(1 (infinite) in music mode) with both meters.



All my original FR measurements occurred with the switch in music mode which is infinite resistance. (checked and double checked at the time). All switches have a nice satisfying click when they pop into place. These are not infinite pots - but rather position pots. (if that's the right description?)



From the JBL manual - the music/speech switch should so the following: (looks like more like 500hz to 6000hz boosting)




Archaea's 9.8.4 Home Theater Room
(13) JBL CBT 70j-1 | Denon x7200wa | Sherbourn PA 7-350 amplifier | (8) Ultimax 18" sealed subwoofers | (4) iNuke DSP 6000 amplifiers | (4) MB Quart 12" subwoofers mounted direct mounted to Berkline theater chairs BOSS style | Epson 5040UB Projector | Jamestown 144" acoustic transparent 2.35:1 screen w/ Seymour XD fabric

Last edited by Archaea; 10-09-2018 at 08:58 PM.
Archaea is offline  
post #19 of 84 Old 10-09-2018, 07:32 PM
AVS Forum Special Member
 
mtg90's Avatar
 
Join Date: Aug 2011
Location: Millbrook, IL
Posts: 1,948
Mentioned: 584 Post(s)
Tagged: 0 Thread(s)
Quoted: 770 Post(s)
Liked: 1588
Did you check the 8 Ohm R1 resistor like fill35U said? If that resistor or its solder joints are bad then the inductor L7 and capacitor C8 will create a much sharper/deeper notch then that circuit is supposed to.
mtg90 is offline  
post #20 of 84 Old 10-09-2018, 09:15 PM - Thread Starter
AVS Forum Addicted Member
 
Archaea's Avatar
 
Join Date: Apr 2004
Location: Kansas City, MO
Posts: 11,046
Mentioned: 604 Post(s)
Tagged: 2 Thread(s)
Quoted: 3772 Post(s)
Liked: 3556
JBL crossover issue - polarity? Seeking help.

Quote:
Originally Posted by mtg90 View Post
Did you check the 8 Ohm R1 resistor like fill35U said? If that resistor or its solder joints are bad then the inductor L7 and capacitor C8 will create a much sharper/deeper notch then that circuit is supposed to.


If only I understood which resistor that was?

Suspected bad board pulled. Pics follows

Wiring










Sent from my iPhone using Tapatalk

Archaea's 9.8.4 Home Theater Room
(13) JBL CBT 70j-1 | Denon x7200wa | Sherbourn PA 7-350 amplifier | (8) Ultimax 18" sealed subwoofers | (4) iNuke DSP 6000 amplifiers | (4) MB Quart 12" subwoofers mounted direct mounted to Berkline theater chairs BOSS style | Epson 5040UB Projector | Jamestown 144" acoustic transparent 2.35:1 screen w/ Seymour XD fabric
Archaea is offline  
post #21 of 84 Old 10-09-2018, 09:28 PM - Thread Starter
AVS Forum Addicted Member
 
Archaea's Avatar
 
Join Date: Apr 2004
Location: Kansas City, MO
Posts: 11,046
Mentioned: 604 Post(s)
Tagged: 2 Thread(s)
Quoted: 3772 Post(s)
Liked: 3556
JBL crossover issue - polarity? Seeking help.

Derp...answer was staring me right in the face on the board's silk screen.



R1 SQZ 20W8RJ is the label. It's the grey sandbox resistor at the top of the board (opposite the 8 pin tweeter connection). You can also see the label for the R2 resistor in the above pic.



Measuring now!

Multimeter shows R2 as 7.9 ohm on each board, but R1 as 0.0 ohm?












Both measuring 0 ohm on resistor 1 both known good and suspect bad board doesn’t make sense?

Archaea's 9.8.4 Home Theater Room
(13) JBL CBT 70j-1 | Denon x7200wa | Sherbourn PA 7-350 amplifier | (8) Ultimax 18" sealed subwoofers | (4) iNuke DSP 6000 amplifiers | (4) MB Quart 12" subwoofers mounted direct mounted to Berkline theater chairs BOSS style | Epson 5040UB Projector | Jamestown 144" acoustic transparent 2.35:1 screen w/ Seymour XD fabric

Last edited by Archaea; 10-09-2018 at 10:05 PM.
Archaea is offline  
post #22 of 84 Old 10-09-2018, 09:54 PM - Thread Starter
AVS Forum Addicted Member
 
Archaea's Avatar
 
Join Date: Apr 2004
Location: Kansas City, MO
Posts: 11,046
Mentioned: 604 Post(s)
Tagged: 2 Thread(s)
Quoted: 3772 Post(s)
Liked: 3556
JBL crossover issue - polarity? Seeking help.

Suspected bad crossover measuring both grey sandbox resistors. R1 and R2. R1 reads 0.0, R2 reads 7.7-7.9.



Same for known good crossover board!!!??? I don’t know anything, but that seems odd!

Was hoping that was it - seemed a fix I could easily handle.











I appreciate all of your help!!!


What's next to try?

Archaea's 9.8.4 Home Theater Room
(13) JBL CBT 70j-1 | Denon x7200wa | Sherbourn PA 7-350 amplifier | (8) Ultimax 18" sealed subwoofers | (4) iNuke DSP 6000 amplifiers | (4) MB Quart 12" subwoofers mounted direct mounted to Berkline theater chairs BOSS style | Epson 5040UB Projector | Jamestown 144" acoustic transparent 2.35:1 screen w/ Seymour XD fabric

Last edited by Archaea; 10-10-2018 at 05:19 AM.
Archaea is offline  
post #23 of 84 Old 10-10-2018, 05:15 AM
Senior Member
 
Join Date: Aug 2014
Posts: 449
Mentioned: 6 Post(s)
Tagged: 0 Thread(s)
Quoted: 255 Post(s)
Liked: 238
R2 is good.

To measure R1 in-circuit, you would need to unsolder one leg of L7 (and make sure the voicing switch is open).
But for the trouble of messing with L7, you might as well just completely unsolder R1 instead and measure across it. If it's good, put it back in and take another sound measurement. By resoldering it, you'll have taken care of any bad solder joints.

Sources/processing: stack of stuff that if it isn't vintage now, it will be soon!
Amps: stacks and stacks of old iron
Main speakers: big DIYSG
Surrounds: Bose graveyard
Subs: a bunch
fill35U is offline  
post #24 of 84 Old 10-10-2018, 07:18 AM
AVS Forum Special Member
 
Samps's Avatar
 
Join Date: Feb 2014
Location: Pittsburgh, PA
Posts: 2,733
Mentioned: 51 Post(s)
Tagged: 0 Thread(s)
Quoted: 1224 Post(s)
Liked: 901
Is it possible that any of those solder joints that are very near to the adjacent circuit are bridging the gap?

Some of these look very close or even spilled over onto the neighboring network. Or that one with four dots almost looks like they are touching.

Attached Thumbnails
Click image for larger version

Name:	Screen Shot 2018-10-10 at 10.16.41 AM.png
Views:	24
Size:	21.2 KB
ID:	2466602  
Attached Images
   
LTD02 likes this.
Samps is offline  
post #25 of 84 Old 10-10-2018, 08:08 AM
Senior Member
 
Join Date: Aug 2014
Posts: 449
Mentioned: 6 Post(s)
Tagged: 0 Thread(s)
Quoted: 255 Post(s)
Liked: 238
Good eye, Samps!

If the 4-pin connector was shorted, the entire woofer section would be shorted to ground, but from the sound measurements all are working.

Likewise, if the area in your bottom right pic was shorted, none of the tweeters would work.

But if the area in your bottom right pic was shorted, this would short across L2. I don't think that would cause a 2k dip, but since this is a high-tech phased array, maybe??

The other thing is, it would have to be bad in the same way in the same place on all the bad units, which is unlikely. OTOH, I've worked on gear where the same hand soldered joint in the same location was unquestionably bad in the same way over several months of factory production, so anything's possible...

Sources/processing: stack of stuff that if it isn't vintage now, it will be soon!
Amps: stacks and stacks of old iron
Main speakers: big DIYSG
Surrounds: Bose graveyard
Subs: a bunch
fill35U is offline  
post #26 of 84 Old 10-10-2018, 10:05 AM
AVS Forum Special Member
 
mtg90's Avatar
 
Join Date: Aug 2011
Location: Millbrook, IL
Posts: 1,948
Mentioned: 584 Post(s)
Tagged: 0 Thread(s)
Quoted: 770 Post(s)
Liked: 1588
Have you measured a bad speaker with the speech mode switch enabled? If it is a bad R1 resistor the switch would bypass that network and remove the dip. If it's not related to that network then the dip should still be present in the speech mode.
Archaea likes this.
mtg90 is offline  
post #27 of 84 Old 10-10-2018, 10:06 AM
AVS Forum Special Member
 
Samps's Avatar
 
Join Date: Feb 2014
Location: Pittsburgh, PA
Posts: 2,733
Mentioned: 51 Post(s)
Tagged: 0 Thread(s)
Quoted: 1224 Post(s)
Liked: 901
Response to post 25:

Yeah, I thought it was unlikely too. But if the design of the board has a difficult to solder section, I could see it failing repeatedly.

Brings up a good point though. What component would be something that could be consistently bad on this many units?
Archaea likes this.
Samps is offline  
post #28 of 84 Old 10-10-2018, 12:24 PM
Senior Member
 
Join Date: Aug 2014
Posts: 449
Mentioned: 6 Post(s)
Tagged: 0 Thread(s)
Quoted: 255 Post(s)
Liked: 238
More food for thought:

On the burned board, C12-2 is the cap that erupted. I've seen those types of metallized film caps fail that way before. I don't know the root cause, but when they fail they must have had a good amount of energy through them to at least burst them open.

Why would C12-2 be passing a lot of current? It's a 2uF cap paralleled with a 100uF one(C12-1). If the caps were ideal, not much should be going through C12-2. Perhaps C12-2 was added as a tweak? But the tolerance on C12-1 is almost certainly +/-10% or greater, swamping the value of C12-2.

C12-1 is an aluminum non-polarized electrolytic, the schematic designates a Dissipation Factor of <4%. C12-1 still probably has a higher ESR (couple orders of magnitude) than C12-2, but again most of the current should be going through C12-1. C12-1 is rated at probably... 0.25 amps? So much less should be going through C12-2, and through a much lower ESR! At least at audio frequencies, and we're already downstream of iron-core L9, and in series with L10.

Besides, they're in a notch filter with an effective 10R series resistance. Even less opportunity to fail.

Howabout this scenario: loose woofers connection overvolts C12-2 by inductive kick, C12-1 sucks it up by dielectric reforming, but C12-2 blows a pinhole which leads to internal arcing. Arc is a low resistance path at all frequencies, and the internal fire is fed...

Curious how with C12-2 completely destroyed, there was no difference in the measured response near the 111Hz area of the notch filter it was part of!

I vote for mailing the bad XO to mtg90 for forensic analysis.

Sources/processing: stack of stuff that if it isn't vintage now, it will be soon!
Amps: stacks and stacks of old iron
Main speakers: big DIYSG
Surrounds: Bose graveyard
Subs: a bunch

Last edited by fill35U; 10-10-2018 at 12:28 PM.
fill35U is offline  
post #29 of 84 Old 10-10-2018, 03:11 PM
AVS Forum Special Member
 
Samps's Avatar
 
Join Date: Feb 2014
Location: Pittsburgh, PA
Posts: 2,733
Mentioned: 51 Post(s)
Tagged: 0 Thread(s)
Quoted: 1224 Post(s)
Liked: 901
Here's a wild idea that may be complete BS, but what if you made two jumper wires with alligator clips on each end. Then connect those across each component on the good board and the bad board, then measure response. Eventually you should get to the component that is bad.
Samps is offline  
post #30 of 84 Old 10-10-2018, 08:06 PM
AVS Forum Special Member
 
mtg90's Avatar
 
Join Date: Aug 2011
Location: Millbrook, IL
Posts: 1,948
Mentioned: 584 Post(s)
Tagged: 0 Thread(s)
Quoted: 770 Post(s)
Liked: 1588
Quote:
Originally Posted by fill35U View Post
More food for thought:

On the burned board, C12-2 is the cap that erupted. I've seen those types of metallized film caps fail that way before. I don't know the root cause, but when they fail they must have had a good amount of energy through them to at least burst them open.

Why would C12-2 be passing a lot of current? It's a 2uF cap paralleled with a 100uF one(C12-1). If the caps were ideal, not much should be going through C12-2. Perhaps C12-2 was added as a tweak? But the tolerance on C12-1 is almost certainly +/-10% or greater, swamping the value of C12-2.

C12-1 is an aluminum non-polarized electrolytic, the schematic designates a Dissipation Factor of <4%. C12-1 still probably has a higher ESR (couple orders of magnitude) than C12-2, but again most of the current should be going through C12-1. C12-1 is rated at probably... 0.25 amps? So much less should be going through C12-2, and through a much lower ESR! At least at audio frequencies, and we're already downstream of iron-core L9, and in series with L10.

Besides, they're in a notch filter with an effective 10R series resistance. Even less opportunity to fail.

Howabout this scenario: loose woofers connection overvolts C12-2 by inductive kick, C12-1 sucks it up by dielectric reforming, but C12-2 blows a pinhole which leads to internal arcing. Arc is a low resistance path at all frequencies, and the internal fire is fed...

Curious how with C12-2 completely destroyed, there was no difference in the measured response near the 111Hz area of the notch filter it was part of!

I vote for mailing the bad XO to mtg90 for forensic analysis.
I don't know, I feel like most the time those electrolytic capacitors blow out one of the ends. I also think it looks more like the 30uF C11 that ruptured.

I propose this alternative hypothesis. Woofer connector somehow comes loose and technician notices that this speaker's level is low so he cranks it up. Without the load of the woofers on the 2nd order filter C11 acts as a short at the resonant frequency of the now LC circuit and soaks up tons of current from the cranked up amplifier. L7 and L9 get real toasty from all the current being passed and cause the heat shrink to melt/burn, C11 eventually gets hot enough to explode/catch fire. Also since the amp was cranked up so far up both bulbs on the HF side blow out as well.
mtg90 is offline  
Sponsored Links
Advertisement
 
Reply DIY Speakers and Subs

Thread Tools
Show Printable Version Show Printable Version
Email this Page Email this Page


Forum Jump: 

Posting Rules  
You may not post new threads
You may not post replies
You may not post attachments
You may not edit your posts

BB code is On
Smilies are On
[IMG] code is On
HTML code is Off
Trackbacks are Off
Pingbacks are Off
Refbacks are Off