EEVblog 1672 - JBL REPAIR: You'll never guess the fault

[EEVblog]

Repair of a JBL Partybox 310 speaker. Follow along as Dave goes down the repair rabbit hole. Try and guess the exact fault before Dave does, it's a doozy!

00:00 - JBL Partybox 310 Speaker fault symptoms and initial failure speculation
02:57 - Teardown
06:23 - Is it a firmware lockup issue?
07:17 - Checking the power switch
09:13 - Main PCB's
10:52 - Hello IT, have you tried turning it off and on again
11:46 - PCB microscope inspection
13:39 - Bench testing and, thou shalt test voltages
16:31 - Looking at the keypad PCB again
17:41 - Replicate the fault using an external PSU
22:55 - Schematic, we have Schematic!
24:14 - The power flow block diagram
28:44 - The penny drops
29:45 - Design TIP
32:04 - Without the schematic I'd be chasing a red herring down a rabbit hole
33:10 - Manual Bypass time to get it working!
34:50 - I called it at the start, but assumed incorrectly
36:33 - OK, let's trace WHERE this fault actually is
37:41 - No, it's can't be, surely not...
39:52 - But WHAT caused it?
42:05 - The repair
44:20 - Assembly back together and testing



Part 2:

[MLXXXp]

@EEVblog, I hope for thoroughness you replaced the ECHO - button that had the bad "clicker", while you had access to it. Yes, I know it probably still worked fine but, at least for me, my OCD would have forced me to change it.

[EEVblog]

@EEVblog, I hope for thoroughness you replaced the ECHO - button that had the bad "clicker", while you had access to it. Yes, I know it probably still worked fine but, at least for me, my OCD would have forced me to change it.

In all the excitement, I forgot!

[Kleinstein]

The design of the 2 PCBs for the buttons is really odd: not just the rediculous number of zero ohm links, but also having 2 PCBs side by side with the extra cable.

[kleiner Rainer]

My opinion: the power button is the only button that has a voltage across it all the time - even when the speaker is off, so any corrosive liquid has all the time it needs to corrode the copper trace. The other traces nearby are not corroded because there is no voltage on them when off.
Even rain water is conductive enough to destroy circuit boards given enough time - I found that out when a phantom power unit for a Sennheiser artificial head microphone got rain soaked - a few hours with 9V dissolved a trace on the pcb!

Greetings,

Rainer DG1SMD

[EEVblog]

The design of the 2 PCBs for the buttons is really odd: not just the rediculous number of zero ohm links, but also having 2 PCBs side by side with the extra cable.

Yes, odd, unless theree another model without half the controls. But even then you just don't populate them.

[EEVblog]

My opinion: the power button is the only button that has a voltage across it all the time - even when the speaker is off, so any corrosive liquid has all the time it needs to corrode the copper trace. The other traces nearby are not corroded because there is no voltage on them when off.
Even rain water is conductive enough to destroy circuit boards given enough time - I found that out when a phantom power unit for a Sennheiser artificial head microphone got rain soaked - a few hours with 9V dissolved a trace on the pcb!

Yes, that's what my Part 2 video is about.

[Langdon]

My opinion: the power button is the only button that has a voltage across it all the time - even when the speaker is off, so any corrosive liquid has all the time it needs to corrode the copper trace. The other traces nearby are not corroded because there is no voltage on them when off.
Even rain water is conductive enough to destroy circuit boards given enough time - I found that out when a phantom power unit for a Sennheiser artificial head microphone got rain soaked - a few hours with 9V dissolved a trace on the pcb!

Greetings,

Rainer DG1SMD

Acid rain?

Damn sulfur dioxide!

[golden_labels]

Great! 45 minutes of a thrilling detective story.

We also have a separate thread about the 0 Ω links (from EEVblog2).

[Simmed]

JBL procurement manager : guys we have a surplus of over 1million 0R smd resistors, let me know some ideas what to do with it ok ?   

[tszaboo]

can someone post a link to that service manual please?

[temperance]

The resistor in top right corner solder joints looks like it wasn't heated properly during production. The resistor termination looks untouched to me.

For the "etched" away trace. I've seen PCB's with acid trapped underneath the solder mask causing all sorts of problems from shorts between traces to open traces.

Edit: if those things are build down to cost. Then why are there ferrite beads on the lines for the push button lines? I mean there are (not on this page of the schematic) resistors going to whichever supply voltage powering the micro controller and the voltage divider outputs are going to an ADC. The capacitors such as C1853 shouldn't be there at all. You only need an RC low pass towards the ADC. 1 resistor and 1 capacitor instead of all the ferrite beads and all the capacitors.

An other problem with this construction is the very low current trough the switches.

Edit 2: the service manual after googling twenty seconds:
https://adjuntos.yoreparo.com/default/0006/72/b7b4200db612927fec6dc038ae1fd9922dfbbc47.pdf

Beware, it seems to be drawn in Orcad 

[TheSteve]

Great vid Dave! See, we need more repair videos.  And yes it takes a lot of broken stuff to get good vids, that is the challenge.

[richnormand]

Re: broken trace.

You mentioned a possible liquid spillage.  Yet nothing was found around the keys and other mating pieces or on the membrane itself.
Similarly you could see the trace all around when you talked about the seal around the membrane.

Looks like the break on each side of the trace is pretty sharp. Have you looked at the membrane to see if the missing part of the trace might be embedded in a glue that was all around the edge?
These units suffer lots of low frequency vibrations and this could be metal fatigue with the trace glued at that spot vibrating all the time, or someone dismantled the unit before and ripped it out?

Interesting detective work.
You may have saved it from an early trip to the junk heap.
Cheers.
 

[JoeyG]

Liquid Damage to an electronics product  should not be a warranty repair as the liquid can seep under IC's  and other places and Electrolysis  take place. Often even a full clean of the PCB to remove moisture/rust and Electrolysis  can not be guaranteed  and the effort to remove all ICs as not financially viable

[Bettingbennie]

Hello, does anyone know what the problem might be?

The speaker sometimes won't turn on. The battery has 8 volts, and after unplugging and reconnecting the battery, I can turn the speaker on. All the buttons work, except for the Bluetooth one. Pressing it briefly or holding it down doesn't work. When I press the Bluetooth button, the battery lights do come on briefly. There's also no pairing sound en light when pressing the Bluetooth button. It's also not in demo mode. I've also tried several complete resets, but without success.

[EEVblog]

The speaker sometimes won't turn on. The battery has 8 volts, and after unplugging and reconnecting the battery, I can turn the speaker on. All the buttons work, except for the Bluetooth one. Pressing it briefly or holding it down doesn't work. When I press the Bluetooth button, the battery lights do come on briefly. There's also no pairing sound en light when pressing the Bluetooth button. It's also not in demo mode. I've also tried several complete resets, but without success.

Definitely not a keypad PCB fault then if the button is actually detected.
Service manual is here:
https://www.eevblog.com/files/JBL-Partybox-310-Service-Manual-v1-4.pdf

[Bettingbennie]

Thanks, I've replaced both pushbutton circuit boards and tested a new battery, but unfortunately without success.

Could it be the motherboard? If so, can I simply replace it, or does it need to be updated afterwards?

According to JBL/Harman customer service, it's not possible to manually update the JBL 310 via USB. Only certified installers can do this. Is it worth trying all of this?

[EEVblog]

Could it be the motherboard? If so, can I simply replace it, or does it need to be updated afterwards?

I can't think of any reason why you couldn't just replace the motherboard.