Logitech Speakerset Repairing

[danger89]

Hi,

My Logitech Speaker Set (just 3 years old) let me down. 
So, I though let's open this thing and find out what's going wrong inside!

After some search I noticed some bright sparks (almost firework) coming out of the bottom of the PCB ahhaa. So I knew enough. I removed the PSU PCB plate.
Tada!

Conclusion: a capacitor was broken and cracked open. Some glass liquid is dropped out (no idea what this is? But don't eat it I guess). I removed the broken cap and cleaned it. As you can see here:


Now, the hard part: it's SMD component. Meaning there is no indication what value is has? Or other specs (250V? ESP?), you just ** don't know....

I hope, now you come into the picture. Can you give me some advise or something to help me further? I already bought a LCR Meter (currently shipping). I didn't had one yet.

To be complete: it's a Logitech Speaker System Z906:


Thanks in advance! 

Kind regards,
Melroy van den Berg

Full-HD pictures:

[HighInBC]

I have encountered burnt SMD capacitors before. I have never heard a good solution to figuring out what they used to be. If anyone knows this wisdom it would be of great general use.

[Rasz]

1 check those transistors
2 power it up with no cap
3 caps below transistors appear to be the same, probably ~1uF

[crispy_tofu]

I have encountered burnt SMD capacitors before. I have never heard a good solution to figuring out what they used to be. If anyone knows this wisdom it would be of great general use.

Same, that would be helpful.  Following!

[alanb]

Have you checked that there isn't an identical part for the other Chanel. If there is you could remove it and measure the value.

[danger89]

1 check those transistors
2 power it up with no cap
3 caps below transistors appear to be the same, probably ~1uF

1. What do I need to check? The value? If they still work? If so, how? I did already a datasheet 'typical (reference) design' look-up without luck.
2. And then? Maybe it will cause more damage? So I need to fix that too...
3. True, they seems the same. Probably ceramic low ESR caps. I bought a LCR meter (see below), hopefully I can figure out the value asap.



Thx!

[danger89]

Have you checked that there isn't an identical part for the other Chanel. If there is you could remove it and measure the value.

On the other channel I only see a diode, no capacitor. As you can see on the picture. I will try to measure one of the caps below the transistors very soon I hope  .

[PA0PBZ]

On the other channel I only see a diode, no capacitor. As you can see on the picture. I will try to measure one of the caps below the transistors very soon I hope  .

Other channel? These are clearly diodes (D20 and D2?) and not transistors, so it could as well be part of the power supply.
What is the exact problem, it does not power up anymore?
Is there a connector on the other side of the PCB on the row of pins? Where is it going to?

[danger89]

Other channel? These are clearly diodes (D20 and D2?) and not transistors, so it could as well be part of the power supply.
What is the exact problem, it does not power up anymore?
Is there a connector on the other side of the PCB on the row of pins? Where is it going to?

yes you are right. D20 and D2 are diodes, sorry I was getting confused.
The problem is indeed it doesn't powerup anymore, and I heard some sparkling. When I opened the case and connected the speaker again: I saw flames .. it was almost firework! WAUW. But atleast I knew where I needed to look at, the bottom side of the power-circuit PCB board. Then I saw this capacitor was clearly broken.
Yes there is a connector for the input voltage (230V) and there are 2 other connectors going to an other separate board. Uhm.. I will take more pictures today, is that OK? I think that makes is more clear. This second board is probably the amplifier board itself, and not the power-circuit board.

Kind regards,
Melroy van den Berg

[PA0PBZ]

Hi Melroy,

Yes, a few more pictures would help. I don't think the failure is just the capacitor, it would normally not completely cut the power.
Can you measure the 2 diodes to see if they are still acting as a diode? If they seem weird disconnect them from the pcb and measure again.

[danger89]

Hi Melroy,

Yes, a few more pictures would help. I don't think the failure is just the capacitor, it would normally not completely cut the power.
Can you measure the 2 diodes to see if they are still acting as a diode? If they seem weird disconnect them from the pcb and measure again.

Ok no problem. Well, sometimes there was just a bit of power. Within a second the devices goes off again. You needed to press several times the power button over and over again to turn it on for a second or less. But as you can see the capacitor is really really broken (see pictures). What I mean is that is damage can lead to total power failure.

I can try to measure the diodes in the evening Thanks for the tip.

[Rasz]

hah they are diodes not transistors, didnt payu attention :/
this cap fried because it shorted, running without a cap wont damage anything
check those diodes for shorts (and all the diodes around while you are at it), + power input (where power comes from socket)

[danger89]

Hi Melroy,

Yes, a few more pictures would help.

I uploaded more pictures here for convenience:
http://photo.melroy.org/thumbnails.php?album=1

ps.
'Power Circuit Bottom Side' picture is where the fun is happening.

ps.ps. there is also a Download link if you need the original file for some reason.

[danger89]

check those diodes for shorts (and all the diodes around while you are at it), + power input (where power comes from socket)

I will thanks. I checked around the power input, I didn't saw anything strange. No blow-up caps and no broken fuses what so ever. Bottom left in the main power input:
Top Image

At the bottom of the PCB, the top-left is now the main power input on this image (I flipped the board over):
Bottom Image

[Rasz]

1 did you try turning it on without the cap?
2 those 2 thin traces originating near burn area and going all the way to the other side of the board to terminate at the connector are interesting, this small part appears to be market RT1 = smd NTC Thermistor = they knew this part of pcb would be cooking = W T F?

[danger89]

1 did you try turning it on without the cap?
2 those 2 thin traces originating near burn area and going all the way to the other side of the board to terminate at the connector are interesting, this small part appears to be market RT1 = smd NTC Thermistor = they knew this part of pcb would be cooking = W T F?

1. I will try that soon, after I measured both diodes (D2 & D20) first.
2. Can you circle the place where you find this RT1 component on the picture?

[Rasz]

2. Can you circle the place where you find this RT1 component on the picture?

it is the one marked RT1 on the silkscreen, surely you can see it on actual board if I can see it on your picture
btw clean that burned spot with IPA

[isopropilick]

I will facilitate the work, my little friend your cap is this (obviously i cut 1 trace to made the capacitor measurement and reconnected to make the voltage measurement): 2.2uf at 100v

[danger89]

I will facilitate the work, my little friend your cap is this (obviously i cut 1 trace to made the capacitor measurement and reconnected to make the voltage measurement): 2.2uf at 100v

WAUW!  I can't thank you enough isopropilick! I measured the cap below (which kinda looks the same as the original cap). This gives me the following result: 1.73uF (using a cheap meter). Still this is one of the caps which looked similar at the left bottom in the picture and NOT, I repeat NOT, the original.

See:


Nevertheless, I'm 100% more sure about your measurement, because of the following points:

• You got a FLUKE multimeter
• You are measuring the original / right capacitor, the one that is broken at my side!

Maybe you want to send a nice high detailed photo near the capacitor, so I can see how I need to solder it back again --> I mean over here the top layer PCB also a burned away  .
Thanks.


==========================
Ps.

I also measured the diodes near to the broken cap. Diode voltage: 0.365V (and the other-way-around: 'LO' as expected).  And 4 MOhm resistant at one way and 'LO' to other-way. Can somebody agree with me, that the diode is atleast still OK?


Ps. ps. I'm also interested how you measured the voltage (100V) of the cap? The set-up and multimeter settings.

btw clean that burned spot with IPA

I ordered IPA thx for the hint, still I don't know how easy it is to clean. It's not flux what you are seeing, but burned capacitor liquid.

[Rasz]

its mostly carbon buildup saturating solder mask, will be ok once you clean it with alcohol
I am guessing isopropilick estimated cap voltage based on working voltage he measured
depending on a type caps can have a HUGE tolerance error (Dave talked about it on one of the recent videos), for fun go to mouser/farnell and look at available parameters for 2uF ceramic cap

[danger89]

depending on a type caps can have a HUGE tolerance error (Dave talked about it on one of the recent videos), for fun go to mouser/farnell and look at available parameters for 2uF ceramic cap

You have a vid link?

Yes, I already checked Farnell. I need to measure the case of the capacitor (size):



Sorry it's Dutch 

[isopropilick]

So yo need more specs? i can provide all what you need (tomorrow when arrive at home)... only request 

[danger89]

So yo need more specs? i can provide all what you need (tomorrow when arrive at home)... only request 

All? WAUW!

Well I think only the size (L"xW") of the casing. Maybe tolerance? The most important details so I can order the correct one.
And again, I really would like to know HOW you did the voltage measurement. So I can do the same in the future for other peoples to help them 

EDIT:

Ps. I also read about the microphonic behavior of ceramic (layered) capacitors. I need to watch out which cap I will pick. Due to the fact it's a speaker-set, sound waves can enter the cap causing voltage on the pcb routes. And also thinking about the impedance level. I need to find a cap which works on the right frequency to have the best amount of impedance (I'm not talking about ESR only).

[Rasz]

I meant tolerance ranges

in general you dont need to stress to much about cap values, its all ballpark anyway (unless radio circuits or fast digital transmission lines)

[danger89]

I meant tolerance ranges

in general you dont need to stress to much about cap values, its all ballpark anyway (unless radio circuits or fast digital transmission lines)

Is class 2 ceramic capacitor ok?