Could be either the power supply section and/or audio power amplifier section. Hard to say until examined closely the internals...

Can ignore the leads that are connected to the bottom side, just repair the leads that are connected to the top side. Check carefully the location to place the "L" wire on the top as to prevent unwanted connections.

If use that method, then pull them by tilting the capacitor side to side with one lead/leg at a time. The problem is that there is solder between the leads/legs of the capacitor and the tube surface inside that still remained. Usually can be pulled out easily but if the leads/legs are tight inside the hole then the soldering iron had to stay a little longer (to completely melt the solder inside the hole), and that usually can result in the tubes and solder pads coming off due to prolonged heating (as the soldering iron melts the glue than holds them in place). And yeah, desoldering double sided board can sometimes be very difficult as well, especially when there is solder still inside the through hole. The most recommended tool for this type of job is: HAKKO | Desoldering / Rework | HAKKO 808 which is rather pricey but gets the job done. Example of real word usage...

This post has been edited by lex: Feb 16 2014, 01:44 PM

For those capacitors after the voltage regulators, any low impedance capacitors will do. The main purpose is to filter out remaining ripples from the voltage regulator and rectifier stage (which could have slipped the first hurdle). Those Nippon Chemi-con GXE series have higher operating temperatures (up to 125C) targeted for harsh and very hot environments such as inside car engines (and should be also suitable for this speaker system). Generally prefer capacitors that can handle high ripple currents and also low impedance as well.

Is that R30 resistor have a value of 100 Ohms? What was the wattage of the resistor and zener diodes? Do note that both resistors and zener diodes have wattage ratings. If too small wattage then it may blow (due to overheating and conducting electrical current past specifications). Check the pins at the transformer for cold solder joints. There may be a weak MOSFET in the other side/half of the circuit (which could be damaged in the first failure). Thus replace those with new ones as well. Also check that fusible link (0 Ohm resistor). Anyway, did you remove all the degraded glue and replace all the capacitors as mentioned here?

This post has been edited by lex: Feb 22 2014, 07:03 PM

Try using higher wattage parts, plus check D16, R39, C30 and C58 (replace all of them if necessary). Also on the other half, replace Q1. As for the R31 (fusible link or zero Ohm resistor), make sure it has no resistance at all (virtually zero Ohms). Do inspect carefully for possible cold/cracked solder joints on the transformer pins. Re-check the L links on the through hole repairs (make sure the top side has sufficient solder and contact with it)...

Make sure all degraded glue has been cleanly removed especially the feedback areas, and places there there are exposed traces and solder pads. Clean with isopropyl alcohol also, as sometimes the chemicals from the degraded glue breaking down may remain on the surface (will look like stains, and is usually conductive and/or corrosive!).

This post has been edited by lex: Feb 24 2014, 01:20 PM

For most of these switching type power supply circuits, especially close to any choke or transformer (such as those used as part of snubber network), the voltage rating of the capacitor should be around at least 1kV (usually to account for any forward and/or back EMF which has usually higher voltage than the supply rail). For that C58, just try using one 3.3nF capacitor with 1kVDC rating (although it looks to be main part of an oscillator startup/feedback loop)...

Hardly matters here anyway as even the audio signal is very low (typically under 2VRMS) and well below 50V peak level (of course not surpassing the voltage of the power supply). The bias voltage held at these capacitors is about half the voltage of the supply rails which is pretty low (again, much lower than 50V)...

First, check what was the voltage output at the main power section (for the power amplifier section). Did you change all the capacitors as shown here? Also check for faulty components, especially that fusible link (zero Ohm resistor) at the main power section...

Its 22 Ohms. That resistor is part of a RC snubber network around the recitfiers. I would recommend Google "resistor color code" to find references to resistor values, as well as online resistor color code calculators.

Well, a fan mod would do better to cool down the components though would be more work than just attaching an extra heatsink. I've had another one repaired not long ago. Not as pretty as the first few shown here (due to very stubborn degraded glue)..

Have not yet encountered such problematic control pods, though the usual problems I've encountered is the glitchy standby button (which causes quick on-then-off sometimes) due to worn microswitch inside. The other one, was the stuck volume button caused by misaligned plastic contact internally (can be fixed easily). As for the problem described above, thanks, will look into it when/if I've encountered one of these...

Unfortunately I'm not living in the U.K. You may try finding someone with electronics repair experience (ie. good with soldering jobs) to help you from your own country at Badcaps...

This post has been edited by lex: Jun 20 2014, 08:12 PM

As mentioned in this thread, first steps would be to replace those capacitors and clean up all the degraded glue (including those on the other boards besides the power supply one). Sometimes bad capacitors do not necessarily bulge/bloat. They can be also be leaking from the bottom, and/or simply became "dried" without any apparent signs. As for other failed and/or faulty components, then you have to look/observe carefully and may require some tools (such as a multimeter, plus some knowledge on diagnosing/checking for failed/faulty components). For some components, you may need to pull it out to check (ie. cannot be checked in circuit itself)..

That hot air treatment is not recommended at all, and will eventually kill the subwoofer (which is what happened to the original person who posted that "quick fix", his subwoofer finally went kaput). That only works if there are dying/failing capacitors on the power supply. For your information, when capacitors start failing their capacitance is becoming reduced. Heating up electrolytic capacitors will increase its capacitance, thus "restores" some capacitance to allow the power circuits to continue to function. However in the long run, heating up those capacitors will cause the electrolytes inside to boil and degrade, thus finally goes bad and the subwoofer stops working altogether. Also heating up the boards inside the subwoofer dries up and evaporates the conductive chemicals from the degraded glue. This restores normal circuit functions. However as the degraded glue continue to decay and breakdown, more conductive chemicals are produced and causes circuit confusion (due to creation of "phantom circuits"). Using that hot air heating it up only temporarily restores some circuit functions (until the next round). Please note that direct heating can also damage (and degrade) other components as well...

Some photos may help. What is your level of electronics knowledge? In particular, checking for failed/faulty components?

This post has been edited by lex: Jun 23 2014, 02:10 AM

Both D6 and D7 are DIACS. The part number is "BR100/03" and you should be able to find it on most online stores. D22, D23, D24 and D25 are actually a 15V zener diode. Can just choose any available 15V zener diode (and preferably with higher wattage), for example "1N4744A"...

Resistors should not be 40% off (that's too much). The most I would tolerate would be 10% off, and the most recommended tolerance is 5% or lower. Thus do check those resistors again. Anyway some of those resistors mentioned are simply 0R (zero Ohm) links such as R31 and R10 thus your measurement could be off due to the error offset on your meter. Thus do check the error offset of your meter, as it could have caused incorrect readings. As for properly measuring the resistors, you have to remove them off the board first otherwise surrounding connected components (including other resistors) could affect the measurement...

They (Creative Labs) no longer make beasts like these anymore. This GigaWorks S750 subwoofer is capable of reproducing frequencies below 35Hz or below 40Hz (the lowest frequency which most run-of-the-mill subwoofers are capable of), as mentioned in its specifications a very low 25Hz...

This post has been edited by lex: Jun 25 2014, 02:50 PM

The run-of-the-mill 1N4004 is just a normal/standard diode thus not suitable. That D12 is right after the transformer and this is a switching type power supply, thus you have to use either a schottky diode or a fast recovery diode. In other words, not the same...

This post has been edited by lex: Jun 27 2014, 08:44 PM

Did you managed to fully revive the unit? Does it still suffer from very low volume after repairs? Check those 0R resistors situated right behind those big four 470uF 200VDC capacitors, in case the low volume problem still remains...

Looks like quite an amount of oxidation there (by the chemicals from degraded glue) as the copper layer looks dark and/or dull, instead of being shiny. After removing all those chunks degraded glue, try cleaning those exposed spots with IPA (isopropyl alcohol) to remove any remnants of chemicals from the decaying glue...

This post has been edited by lex: Jul 22 2014, 07:28 PM

C40? Is that a ceramic disc capacitor? If it is then usually that is 10nF (or 10,000pF) and the recommended voltage rating is usually 1kV since its part of the snubber network around the rectifiers...

This is one of the power amplifier boards, and there are two of them. In fact both of them are identical. You can check the value of the capacitor on the other power amplifier board, in the same location. If that capacitor is in parallel with that big (1000uF 100VDC) electrolytic capacitor next to it, then very likely its a 100nF capacitor with voltage rating of at least 100V (following the big capacitor, can use capacitors with higher voltage rating also). Anyway, that capacitor may not be the only failed component on that board. Possibly that big (1000uF 100VDC) electrolytic capacitor next to it has some problems as well...

This one is a 4.7uF 100V ceramic capacitor. It should follow the same voltage as that big capacitor next to it...

The one to get rid off are those that have turned yellowish to brownish color. The white colored glue is still fine, as its a different type of glue than the ones used on the rest of the boards...

Actually you can use just any ceramic capacitor of the same capacitance (4.7uF) and voltage rating (100VDC). Typically I would choose NP0/C0G, U2J, X7R and X8R type ceramic capacitors (more stable across entire range of temperatures) rather than the cheaper Z5U and Y5V ceramic type capacitors (capacitance and impedance varies a lot with temperature). That big capacitor beside it is a general purpose capacitor. Usually if you find a ceramic capacitor in parallel with it then its one way which some manufacturers cut cost. Instead of using more expensive specialised low ESR/impedance capacitors, they would use cheaper general purpose capacitor with a cheap ceramic capacitor. The strength of ceramic capacitor is that it has very low ESR/impedance at high frequencies (typically in the switching power supply range). Thus that capacitor serves as a low ESR/impedance side to the bigger electrolytic capacitor (functions to bypass/filter and sink high frequency ripples off the rails that the bigger general purpose electrolytic capacitor unable to handle). And if the power rails are too noisy (often due to power supply failing) then that ceramic capacitor will be taking all the "heavy load" (trying to remove the ripples)...

This post has been edited by lex: Aug 29 2014, 01:06 AM

Glad it worked. Should give those speakers a new lease of life...

Have you cleaned the connectors on the power amplifier boards? Also you may have to clean the audio controller board and the I/O board (where the audio, control pod/dock and speaker connectors are located). Especially the I/O board, which is right under the audio controller board, has lots of degraded glue especially at the connector solder pins. That one is kinda hard to clean and require some proper tools...

Yups, those are the boards I've mentioned. Particularly that I/O board is just full of glue. I would recommend that you scrape it off, as there could be degraded glue hidden under it (and causing problems with some of the audio inputs there). If you do not want to desolder the board off (which is a lot of hard and difficult work, particularly since this is double sided PCB) then try usuing a thin long flat head screwdriver plus very fine but strong tweezers to remove those glue. Also remove any glue on the audio controller board especially those that cover the I.C pins. Have to be delicate as these are (delicate) SMD type components thus proceed with care and caution...

First off, that 470uF 200VDC capacitor is of wrong dimensions. Its too tall at 50mm. Plus the originals are "snap-in" types. Thus please follow of the capacitors I've selected, which should be this one: Mouser > Passive Components > Capacitors > Aluminum Electrolytic Capacitors > Aluminum Electrolytic Capacitors - Snap In > Panasonic EET-ED2D471BA. And looks like that one is going to be EOL'ed (thus grab it while still available, because that one has one of the best specifications compared to the rest of the field)...

Those are not really resistors. Sometimes they are called 0 Ohm resistors (single black band means "0"), but they are actually fusible links (that looks like resistors)...

That glue to help hold the components (especially big ones) in place for easier soldering (in the factory). Other than that it has no use at all. Thus just remove those glue. And do not use anymore glue. The prolonged heat inside the subwoofer can degrade (cause chemical breakdown) those glue. Do clean with IPA (isopropyl alcohol) the areas where the glue was removed...

Seems Panasonic decided to quit producing snap-ins. That Panasonic TS-ED series (actual datasheet here) has the best specifications among all, with ripple current handling of 1850mA@120Hz while the rest of the field have around 1400mA to 1600mA ripple current handling only (at 120Hz). Alternatives would be Nichicon GU series (which has lower ripple current handling)...

If they are not broken (should read as shorted on the multimeter) then no point replacing them. Yes, those are the "0 Ohm" resistors (actually fusible links)...

That standby power section may be able to automatically adjust to new higher voltage (since its a standard flyback design), but the main power section may not be able to use a higher voltage (does not seem to have any feedback section).

Have you tried removing the degraded glue from the audio controller and I/O boards, as well as the connectors on the cable that links the audio controller board to the power amplifier boards (as shown in earlier posts)?

Not sure what's the actual difference between both 110V and 230V versions (since I have not worked on a 110V version), but you can try out and see if swapping/changing those connections will work...

This post has been edited by lex: Sep 26 2014, 12:20 AM

Just a 60W soldering iron, ordinary de-soldering pump (solder sucker) and solder flux. Make sure you have a soldering iron of sufficient wattage (minimum of 40W, recommended at least 60W or higher) because this is a double sided PCB. Just becareful of those smaller solder through holes with thin surrounding pads. Make sure you do not damage them with a much hotter soldering iron (avoid heating it up too long).