OCZ-520 Looks Good, but Won't Work

[Jonny Accelerant]

I suspect the cap(s) immediately next to the large bundle of cables coming into the PSU. I'm thinking mechanical action/stress/movement has "done something" to the caps, as one of them in particular is REALLY wobbly. It won't rotate, so it feels like both leads are still intact, but you can't really see anything so IDK. Wonder if caps can go bad due to movement stress where the lead stays unbroken, but the cap is broken anyways.

Also, there is another thread (which I've read) on badcaps here. Mine is a 520 Watt and that one is a 420 and the components look different although the overall design looks the same. Cap manufacturers are different.

Did a several thorough visual inspections top, side & bottom, multiple times and I've gone through 3 different cameras trying to get the best possible resolution, which I'll call "adequate".

I can find no evidence of burning, heat, leaking, bulging, nor any other defect that might indicate a problem, but the PSU won't work. It killed one motherboard and I got stupid and put it in another and luckily the PSU died before it killed another motherboard. From what I've read, when caps go bad they allow "ripple" which I assume is variation in voltage and that is really bad for motherboard caps, and that is (one way at least) how bad PSU's kill motherboards.

So caps are on the list of suspects. I have a few ideas on how to troubleshoot this, but would rather get the 1st step, basic advice first before going off on some wild tangent. This particular repair is stemming from a post I made several days ago asking if, in general, repairing PSU's for money is worth doing, and the consensus was "not really", but for the sake of personal development and a little side money. This is the best of the 20 or so PSU's I have laying around, so I thought I'd invest my time & energy into the one that would give the most bang for the buck, compared to the numerous POS PSU's that, even when repaired are still POS's.

As usual, I appreciate all help, even misguided attempts at help. This will be the 4th piece of electronics I've successfully repaired (if successful) based on knowledge and help I've received on this forum, and I remain appropriately appreciative of that. As an aside, I spent almost a full year in the USCM ground radio repair school learning to troubleshoot and repair military radios to the component level and once graduated, spent the remainder of my 4 year contract and never touched a soldering iron.

They used to tell us it cost in the neighborhood of $100,000 to train us, and how appreciative we were supposed to be, and then we never used a single dime's worth of that training. Now, 25 years later I've put that background to use for the 1st time ever, and have repaired more equipment in the last 2 months than I did in 3 years in the USMC.

To give you-all a sense of perspective about "military training" vs. online educational opportunities. And also where you tax payer dollars go. They could probably send those jarheads here, and save the taxpayers millions, is my point.

[kaboom]

Re: OCZ-520 Looks Good, but Won't Work

Check 5vsb. The aux supplies have to work in order for the main supply to come up.

I bet it's high or missing.

Purple to black/case.

If 5vsb ran away, it'll burn/destroy the chipset, which has standby power any time the power supply has line voltage.


I would investigate that loose lead/cap before applying power- no need to damage anything else.

[Jonny Accelerant]

Re: OCZ-520 Looks Good, but Won't Work

Check 5vsb. The aux supplies have to work in order for the main supply to come up.

I bet it's high or missing.

Purple to black/case.

If 5vsb ran away, it'll burn/destroy the chipset, which has standby power any time the power supply has line voltage.


I would investigate that loose lead/cap before applying power- no need to damage anything else.

I'm posting the 20/24 pin pinout for future reference.

As I interpret your post, you are saying that the 5 (vsb? Did "volts" get upgraded?) is always on? Even when the PSU is off? I'll check it tomorrow, it's getting late and plus I can't remember where I left my multimeter. Thanks for the help. I'm thinking, no matter what the specific problem is, I should recap the whole thing as a matter of GP (General Purpose). The other thread (link above) someone said something to the effect that when one goes, they all go and made nasty sounding noises about the quality of the caps in the unit.

When I bought it, it was supposed to be a high-end unit. It's heavy as hell and note the super-heavy layer of solder on the runs. Never seen that before, ever. I need to find the badcaps.net list of "known shitty capacitor manufacturers" (obviously titled more professionally) and see if any of these are on it. Most of the caps in the photos have the manufacturer's name faced away, so it's going to be difficult to identify them. Another argument in favor of breaking out my nifty new Desoldering Tool ($15.00 from Radio Shack) and dropping all the caps off and then identifying them. Maybe testing them? Some of the old vets here seem to be willing to test and if they test good put them back on the board, whereas I'm more inclined to throw them all away and buy new, since (in my calculator) the time & energy spent in disassembly and reassembly is worth WAY more than the caps, new or used, and that having to do it twice is stupid, whereas throwing away old but good caps and replacing them with new and good caps seems dramatically less wasteful, considering the worst-case scenario. I pay more in shipping than I do in cost of caps anyways, so I don't see the point in removing and then resoldering the same old caps back on. Maybe these are crusty, wizened old misers counting their pennies on an old wooden table by candlelight each night? Caps on average cost 50 cents each, and the average repair is 10 or so caps, so to save a fraction of $2.50 you resolder the old components back on the PC board?

Seems nuts to me. OCD past the point of eccentricity, and well-into Loonyville, USA. Unless there's a reasoned explanation?

[Jonny Accelerant]

Found It!

After literally the 4th time giving it a close visual inspection, with an eye to removing the heat sinks, I found one leaking capacitor (see attached).

So now the question. Can I do this without removing the heatsink(s)? I pulled the screws out from underneath, but the heat sinks are fastened horizontally to some components on the pc board, and how do I get those screws off?

Also, I'm thinking the best way to do this is to desolder each cap one at a time and note what it is and what it's named on the pc board, before trying to identify and order the replacements. Two other threads on similar PSU's say to replace all the caps, and I assume they mean the electrolytic and not the ceramic disks and others.

[kaboom]

Re: OCZ-520 Looks Good, but Won't Work

I'm posting the 20/24 pin pinout for future reference.

It's the purple wire on the 20/24 pin plug.

As I interpret your post, you are saying that the 5 (vsb? Did "volts" get upgraded?) is always on? Even when the PSU is off? I'll check it tomorrow, it's getting late and plus I can't remember where I left my multimeter. Thanks for the help. I'm thinking, no matter what the specific problem is, I should recap the whole thing as a matter of GP (General Purpose). The other thread (link above) someone said something to the effect that when one goes, they all go and made nasty sounding noises about the quality of the caps in the unit.

5VSB- 5V standby

Most ATX power supplies are actually several. There's the main supply for +/-12, +/-5, and +3.3. -5 isn't used much anymore and most newish units omit this output. This main supply can take several forms. Older ones were a half bridge with both a power transformer and a smaller driver transformer. Usually had a TL494/KA7500 and a LM339. Deers (oh dear) combined both devices into a single "chip of the year." Switching devices were, from the small side, a pair of 2SC4242s, to larger MJE13009s. 5/12, both polarities, went thru a common output inductor, usually a type 26 toroid, sometimes 52, while +3.3 had its own as it was regulated seperately. This was a so-called mag-amp regulator, which got its power from the same secondary windings as the 5V outputs, not 5Vsb, which I will get to.

These things are pulse width regulated, and the transformer ratio allows for margin at high load/low line. What this means is that the actual voltage peaks before the output inductor may approach 20V for the 12V output. You'll remember that current through an inductor cannot instantly change. It follows then, that when the 20V pulse comes from the rectifier, the DC output, after the inductor, also does not instantly go up to 20V.

The inductor and output capacitors integrate, and the DC output voltage becomes a function of duty cycle, high-side DC voltage, transformer ratio, and various losses, like transformer leakage inductance, rectifier junction drop, etc.

Now that 5vsb. This supply is on anytime there's line voltage. Usually, there are two aux supplies; the 5vsb, and another one for the actual aux supply, which feeds the PS-ON circuitry and PWM of the main supply.

There are two main types. A one-transistor self-oscillating converter made of discrete components, among which is a shunt resistor that develops a voltage proportional to current drawn by the switch transistor. The top of this resistor ultimately feeds the base of a second transistor. This arrangement limits output in the event 5vsb is shorted by something. But it does not lock out; should something short 5vsb to common, current will still flow- magnitude of which is determined by that shunt resistor and other resistors that determine the ultimate value. The second transistor also removes drive from the first one in order to regulate the output voltage. It does so in an interesting way.

This is the so-called two-transistor circuit. .

In order to explain a most unpleasant failure of such a supply, we must first fully understand it. There are three windings on that 5vsb transformer- primary, secondary, and tertiary or feedback. The actual circuit is just a blocking oscillator with primary and feedback phased accordingly. But the feedback winding does double duty. The AC component is coupled to the transistor to make oscillation sustain, while a DC component is rectified, dropped w/ a resistor, and used to charge a cap- this is the "critical cap." The voltage on this cap is compared to that of a Zener- when it exceeds that by a given amount, that little transistor (2ac945) comes on and removes drive from the first transistor, regulating the output.

"But how does keeping the voltage on that cap nearly constant keep 5vsb constant?" Good question! And the basis for regulation! The voltage on that cap and Zener is turns-ratio-derived to the secondary's voltage. An output inductor isn't needed, since the action of that little transformer (flyback converter) takes its place. Even so, similar PWM action described above happens.

So if 5vsb load increases, both 5vsb voltage and feedback DC voltage decrease, so the duty cycle increases, restoring voltage on the critical cap. If load drops, both output and critical cap voltage go up, that '945 comes on more and reduces the duty cycle of the main switch.

Here's the kicker, literally. Suppose that critical cap goes leaky, to the point of not charging with each cycle. A reference is never developed! That '945 never comes on to limit the duty cycle. The supply runs full blast- hey, gotta try to get the voltage on that cap up! But with a bad cap, that's not happening. Now the supply is running full blast and 5vsb is now 15vsb! The motherboard is long since history... Happened to lots of emachines back in the P3 days.

By contrast, the other standby supply uses something like a TNY-xxx, a TinySwitch. A little more refined and much less likely to run away, generally these have an optoisolator with a dedicated feedback pin.

Finally, there can even be a third supply powering the whole mess. A boost-converter active PFC! Takes line voltage AC and boosts it to 360-420VDC. Both the standby and main supplies are fed from this in such units, though usually the booster is inactive when the main supply is off.

There are several types of main supply, but these things are usually half bridges, or single- or double-forward converters.

From there, how the secondaries are derived can also vary. Most are group regulated on 5/12, with that coupled output inductor; 3.3V having its own. Others, usually much larger, higher wattage ones, just have a single 12V output. Here, there will be buck modules to generate the other voltages, another case of power supply-within-power supply. The +12V output will be divided, with shunts and current monitored, so if current on a given set of output leads gets too high, the supply shuts down.

When I bought it, it was supposed to be a high-end unit. It's heavy as hell and note the super-heavy layer of solder on the runs. Never seen that before, ever. I need to find the badcaps.net list of "known shitty capacitor manufacturers" (obviously titled more professionally) and see if any of these are on it. Most of the caps in the photos have the manufacturer's name faced away, so it's going to be difficult to identify them.

I've been there. There's too much stuff lately that only exists to get you to spend money, rather than simply be what it ought to.

The green ones are Teapo SC. The brown "teapos" somehow resemble Hermeis to me. The stuff has become rather reliable otherwise, barring stuck fan and overheating, or lightning. How else do you "guarantee" repeat business than by deliberately using junk caps?

We're not in the '40s - '60s anymore when electrolytic caps were rather cranky. No. Never have we been so spoiled with cap choices yet today's stuff has junk caps....

Another argument in favor of breaking out my nifty new Desoldering Tool ($15.00 from Radio Shack) and dropping all the caps off and then identifying them. Maybe testing them? Some of the old vets here seem to be willing to test and if they test good put them back on the board, whereas I'm more inclined to throw them all away and buy new, since (in my calculator) the time & energy spent in disassembly and reassembly is worth WAY more than the caps, new or used, and that having to do it twice is stupid, whereas throwing away old but good caps and replacing them with new and good caps seems dramatically less wasteful, considering the worst-case scenario. I pay more in shipping than I do in cost of caps anyways, so I don't see the point in removing and then resoldering the same old caps back on. Maybe these are crusty, wizened old misers counting their pennies on an old wooden table by candlelight each night? Caps on average cost 50 cents each, and the average repair is 10 or so caps, so to save a fraction of $2.50 you resolder the old components back on the PC board?

I've been saying this for years. With what it takes to get them apart, cleaned up and ready to work on, they could've had good caps from day one. 20 year old computer monitors, with good Nichicon caps- all I had to do with them was blow the dust out, clean them up and maybe tweak focus and G2. Cap replacement? No way- they'll outlast the CRT.


If going to that much trouble, I'd put new ones in. You're not that far from Mouser if you want to order from there.

[kaboom]

Re: Found It!

After literally the 4th time giving it a close visual inspection, with an eye to removing the heat sinks, I found one leaking capacitor (see attached).

So now the question. Can I do this without removing the heatsink(s)? I pulled the screws out from underneath, but the heat sinks are fastened horizontally to some components on the pc board, and how do I get those screws off?

Those heatsinks also have tabs soldered to the board. You might be able to get it if you remove the ones in front.

Also, I'm thinking the best way to do this is to desolder each cap one at a time and note what it is and what it's named on the pc board, before trying to identify and order the replacements. Two other threads on similar PSU's say to replace all the caps, and I assume they mean the electrolytic and not the ceramic disks and others.

I always do one at a time, and usually in a certain order when dealing with your under-the-heatsink problem.

Yes, just all electros, especially the small ones!

[Per Hansson]

Re: OCZ-520 Looks Good, but Won't Work

Yea, the PSU has Teapo and what looks like Asiacon.
The Teapos are known to fail without bulging...
So if you want to repair this thing best to replace all Teapo & Asiacon caps...

[kc8adu]

Re: OCZ-520 Looks Good, but Won't Work

looks like a cwt build.these have a cap which goes high esr and makes the +5stby overshoot.
its near the standby transformer.
its an ok unit if you replace all caps with quality ones.

[pdavid]

Re: OCZ-520 Looks Good, but Won't Work

This is a Topower platform from 2004. I have one maxed out with APFC. I remember mine had all caps except primaries blown. Even some of the small ones were dried or blown. On the fan controler too...

[Wester547]

Re: OCZ-520 Looks Good, but Won't Work

Yea, the PSU has Teapo and what looks like Asiacon.
The Teapos are known to fail without bulging...
So if you want to repair this thing best to replace all Teapo & Asiacon caps

Those aren't actually Asiacon, they look much more like JunFu and JunFu = Asia-X = Fuhjyyu (which is worse than Asiacon IMO, as Asiacon = Hermei). So yes they should all be replaced.

By contrast, the other standby supply uses something like a TNY-xxx, a TinySwitch. A little more refined and much less likely to run away, generally these have an optoisolator with a dedicated feedback pin.

Sort of off-topic but everell recently performed a DM311 mod on an FSP300-60ATV and it did not prevent +5VSB from going overvoltage (it went up to 17V). The culprit was a bad optocoupler on +5VSB which a bad flyback transformer caused, also taking out the proprietary FSP supervisory chip on the secondary side.

Of course. the unit failed in that manner before he modded it, but I think it suggests that just having a PWM-generated +5VSB circuit on the primary is not enough for overvoltage protection. I believe this is also somewhat confirmed by the fact that Delta's GPS-450AA and GPS-350BB PSUs both use TNY276 to generate +5VSB but if you look at their actual PDF documentations, no mention of overvoltage for +5VSB is made. Look up DPS-350QB's documentation (which I have not seen the internals of) and there is actual mention of +5VSB overvoltage protection. But like you said, you need something to lock out or "latch". And, as you said, the PWM-generated circuit still provides better regulation than the two-transistor circuit, is probably easier to troubleshoot, and so forth, though I have still noticed the presence of what appears to be a critical "startup" capacitor in these circuits (but not always). And although I'm not really that knowledgeable on this topic, I believe there are a few ways to achieve proper protection (though I may be wrong on this so don't take this with anything more than a grain of salt):

1) Though it would be rather inefficient to do so, use a linear-regulated +5VSB output, which is another variation of the one transistor circuit on the primary of +5VSB (as evidenced by the lack of an optocoupler for +5VSB and a linear regulator on the secondary). This may not be able to allow it to "lock out" but I believe it would be far more effectual regulation and protection for +5VSB than using the regular approach.
2) Having a properly configured zener on the output of +5VSB that would blow should the voltage go higher than 5.5V-6V.
3) Having a crowbar protection circuit (SCR) for +5VSB on the secondary.
4) Having a +5VSB sense wire that would tell the unit to shut down should a certain amount of current be drawn and a certain voltage be reached.

Unfortunately, it does sound like Johnny's unit went overvoltage on +5VSB and killed that one motherboard. So I also strongly admonish that it's recapped before used again in another machine.

[Pentium4]

Re: OCZ-520 Looks Good, but Won't Work

Definitely a Topower. Awesome PSU, that uses crap caps and sometimes weak 12V rectifiers.

Oh boy....Teapo SEK everywhere? Any one of those could be bad, especially the small ones. Check the ones on the 5VSB circuit first. Those purple caps are Aisa-X.

I just repaired another one of these a few weeks back (Branded Athena Power), and every single cap except the high voltage caps tested bad. That's right, 100% of them, even the ones on the daughter-board. Only 50% of them with vents actually bulged. Brands consisting of Fuhjyyu, Teapo, Yihcon, CS, and GoldLink. Did any of the fans seize?

[keeney123]

Re: OCZ-520 Looks Good, but Won't Work

I suspect the cap(s) immediately next to the large bundle of cables coming into the PSU. I'm thinking mechanical action/stress/movement has "done something" to the caps, as one of them in particular is REALLY wobbly. It won't rotate, so it feels like both leads are still intact, but you can't really see anything so IDK. Wonder if caps can go bad due to movement stress where the lead stays unbroken, but the cap is broken anyways.

Also, there is another thread (which I've read) on badcaps here. Mine is a 520 Watt and that one is a 420 and the components look different although the overall design looks the same. Cap manufacturers are different.

Did a several thorough visual inspections top, side & bottom, multiple times and I've gone through 3 different cameras trying to get the best possible resolution, which I'll call "adequate".

I can find no evidence of burning, heat, leaking, bulging, nor any other defect that might indicate a problem, but the PSU won't work. It killed one motherboard and I got stupid and put it in another and luckily the PSU died before it killed another motherboard. From what I've read, when caps go bad they allow "ripple" which I assume is variation in voltage and that is really bad for motherboard caps, and that is (one way at least) how bad PSU's kill motherboards.

So caps are on the list of suspects. I have a few ideas on how to troubleshoot this, but would rather get the 1st step, basic advice first before going off on some wild tangent. This particular repair is stemming from a post I made several days ago asking if, in general, repairing PSU's for money is worth doing, and the consensus was "not really", but for the sake of personal development and a little side money. This is the best of the 20 or so PSU's I have laying around, so I thought I'd invest my time & energy into the one that would give the most bang for the buck, compared to the numerous POS PSU's that, even when repaired are still POS's.

As usual, I appreciate all help, even misguided attempts at help. This will be the 4th piece of electronics I've successfully repaired (if successful) based on knowledge and help I've received on this forum, and I remain appropriately appreciative of that. As an aside, I spent almost a full year in the USCM ground radio repair school learning to troubleshoot and repair military radios to the component level and once graduated, spent the remainder of my 4 year contract and never touched a soldering iron.

They used to tell us it cost in the neighborhood of $100,000 to train us, and how appreciative we were supposed to be, and then we never used a single dime's worth of that training. Now, 25 years later I've put that background to use for the 1st time ever, and have repaired more equipment in the last 2 months than I did in 3 years in the USMC.

To give you-all a sense of perspective about "military training" vs. online educational opportunities. And also where you tax payer dollars go. They could probably send those jarheads here, and save the taxpayers millions, is my point.

I would use some clean air to clean the dust especially about those thru-hole ICs. It may do nothing, but they can fail from conduction thru the dust. Also, those electrolytic caps can fail and still show no outward signs of failure.
The military must of messed up in the latter years in training because I worked with four military train techs in the early 80's and they were very proficient in soldering and troubleshooting.

[Jonny Accelerant]

Progress Report

The primary problem with my OCZ-520 12U P022-F ver:2.2 is the leaking and bulging capacitor C24, hidden under the larger of the two heat sinks, which is why I missed it the 1st 3 times I did a visual inspection.

Which is not to say that this is the ONLY problem. I posted the above for search engine purposes on the chance that someday, someone else will be looking for that exact text.

Last night and part of this morning I removed all 22 capacitors from the P022-F ver:2.2, except for the remaining 3 small capacitors on the daughter board, as I could not see a way to remove the daughter board from the main board without desoldering each and every pin, so I'm going to roll the dice and hope the daughter board is good, and will remain so indefinitely. I'd be interested in learning what the purpose of the daughter board is.

Total time in desoldering was about 5 hours. I worked slowly and methodically, and documented each capacitor on a .txt document as it was removed, then went back afterwards and measured each of their physical dimensions.

This was by far the most difficult and complex repair I've attempted. The super-heavy solder underneath the board made my small desoldering iron have to work very hard. A few capacitors required a very long time to heat the solder before it would liquify, and one in particular required the assistance of my soldering iron to provide additional heat in order to get the solder to melt.

Another unexpected difficulty was in learning how to navigate the underside of the PC board without any labels or markings showing which components were which, nor even what general "class" they were. Diodes, transistors, capacitors, resistors, IC chips, copper wires, heat sinks, inductors and who knows what all else need to be clearly identified from beneath the board so that you do not accidentally desolder something that you might not be able to get back in. I have a loose resistor now because I was not as disciplined as I am right now.

Some hints I learned was that diodes have their leads bent inward towards each other, capacitors and resistors usually have their leads come straight out, wires have a distinctive look, as do IC chips. The super heavy solder runs made things a real confusing mess, however, once you got a few components out, you could use the empty holes in the board as reference points. Flipping the board over & back again, over & over again, was a real, ROYAL pain in the ass as the heavy cables didn't want to "flip" and sometimes I'd have to take a quick look, get a mental image, try to flip the board, the cables would get in the way, I'd have to fuck with them, move them around, no not that way dammit, how about this way, shit this stupid cable won't bend that way, don't let it flop into the soldering iron, there, that's good, now what did that board look like again?

Back to square one. I had to suppress the urge to take a large hammer to the goddam thing at least 3 times, and quit early last night (midnight) because the rage was getting increasingly difficult to control and I really like this PSU. I've destroyed things in the past, and most of the time I feel quite good about it, although sometimes beating the person that designed it in the head with a hammer would feel even better, I decided that it could wait, and it did and this morning I was "fresh" and got the job done with a minimum risk of "Component Repair Rage", which in case you do not know is a Federally recognized disability and if you can find a Doctor willing to give you the diagnosis, you can spend the rest of your life sitting on the couch watching Jerry Springer and wondering (and then finding out) who the Daddy is.

So now I have 22 capacitors that I need to find replacements for. Planning on using Mouser as they are close and I like their search function better. Prefer Panasonic caps, don't care about spending a couple of extra dollars.

Took a couple of pictures, showing C24 in all it's leaking and bulging glory, and also did a "group-shot" showing the entire class together, in case anyone is interested in what they all look like. Teapo and Asia'x, as many have feared, soon to be replaced with genuine american-quality capacitors that just happen to be manufactured in Japan.

EDIT***
Can't upload these two photos. Wonder if there is a limit for uploaded files?

[Behemot]

Re: OCZ-520 Looks Good, but Won't Work

I can supply you with the caps, got Chemi-Cons KYA 3300 uF/6,3 V D10x25 mm, these are hard to get, and Samxons RS 3300/16 D10x30 mm (these are impossible to get anywhere but from me). That is just equal to replace combination of 2200- and 4700uF caps. Got the rest as well but that you can get pretty much everywhere…

[CapSwapper]

Re: OCZ-520 Looks Good, but Won't Work

Teapos

[RJARRRPCGP]

Re: OCZ-520 Looks Good, but Won't Work

Teapos

Not scared of Teapos unlike "KM"s. A lot of "KM"s are real bad without bulging or leaking.

[CapSwapper]

Re: OCZ-520 Looks Good, but Won't Work

I pulled so many bad Teapos I could fill a landfill

[goontron]

Re: OCZ-520 Looks Good, but Won't Work

I pulled so many bad Teapos I could fill a landfill

throw them in a fire and watch the show

[c_hegge]

Re: OCZ-520 Looks Good, but Won't Work

throw them in a fire and watch the show

+1. This.

[Jonny Accelerant]

Uploading Photos

I tried to upload some "progress report" photos the other day and the forum was giving me the finger. Today I could upload photos on a new thread, so maybe I can upload these, which may help someone else and trying to figure out which leads to desolder was very difficult.

"underneath.jpg" shows the underside of the main pc board, with red circles marking those leads that are buried underneath a heavy layer of solder, and the blue circles mark the lead pairs for each capacitor. This was not double-checked, so it may have some errors, but most of these are correct, at least. Also I made some mistakes so if the desoldered area is not marked by a circle, it is probably a mistake I made that will have to be corrected when soldering in the replacement capacitors. One resistor was accidentally desoldered, as well as at least one heavy metal post for a large inductor, I think. This was a sloppy, imprecise job and that is why I am posting this image, so as to help prevent it from happening to someone else.

"daughterboard.jpg" shows the daughterboard (whose purpose is still unknown to me) with the original capacitors that I am NOT going to replace.

"leaking.jpg" shows the leaking capacitor C24 after desoldering and "Leakycap.jpg" shows the C24 before it was removed.

The rest are general images showing the board from various angles after the capacitors have all been removed.