Did I actually fix it?

[SluggerB]

I have an ancient AMD 770 motherboard, a BIOSTAR TForce TA770A2. Worked fine enough

I had 8gb of RAM, and wanted to upgrade to 16gb. I bought some RAM off ebay, I will spare you the gory details, but long and short of it, no configuration of RAM would actually boot into Windows. Sometimes I would get a BIOS post, and Windows would crash, sometimes I would just get a black screen. So I gave up and stuck my 8gb back.

A couple days later, I am using the PC, and it just shuts off in the middle of being used. Like it lost power. I reboot, use the PC for a minute, happens again. Reboot again, 30 seconds in the PC freezes. Reboot again, as long as I do nothing the PC runs, but I start using it, and in a minute or so, bloop, power off. I swapped HDD's, same thing, so I know it's not an HDD issue

So I strip the whole thing down,. and want do I see on the motherboard? A bulging capacitor next to my RAM slots. It's the only thing that looks out of the ordinary, so I fire up the soldering iron, fight with the thing for damn near an hour getting it out, replace it (it was a 10v 1000mF capacitor, I only have a 16v 1000mF capacitor on hand, but the internet tells me it should work)

Solder the new cap in, reassemble the PC, and it powered up. I ran the Windows memory test and it completes and no errors. I'm typing from it now and it's already tun longer and one more than what would have caused a shutdown earlier.....

Did I actually just fix it?

[petehall347]

Re: Did I actually fix it?

most important is was it broke before you started ?

[ChaosLegionnaire]

Re: Did I actually fix it?

install back the 16gb ram and see if its okay. it could be the 16gb of ram was flaky because the cap was on its last legs.

[SluggerB]

Re: Did I actually fix it?

install back the 16gb ram and see if its okay. it could be the 16gb of ram was flaky because the cap was on its last legs.

I'm not risking it, the PC stayed on all last night, and I am using it now and it has stayed on, so it seems like I did actually fix the issue. If the dodgy RAM caused the cap to pop, I don't want to risk damaging something else on the board. It was 16 gigs for $25 so it's not super high quality stuff.

It also could be the semi frequent power blips we have here, enough to shut the computer off, so I bought a UPS to eliminate that.

[RJARRRPCGP]

Re: Did I actually fix it?

A shutdown like that can mean poor contact with the CPU heatsink and the anti-core-overheat was triggered.

[SluggerB]

Re: Did I actually fix it?

A shutdown like that can mean poor contact with the CPU heatsink and the anti-core-overheat was triggered.

I had thought of that since I had that exact issue in the past. I checked my temps in the BIOS and my temps were ~20C-23C, WAY lower than what this chip shuts down at (i think it's like 70C or something)

My completely uneducated guess is that the voltages on the cheap RAM were wrong and that caused the cap to go? It is a very very old board (11 years) so all the components are up in age.

I'm still just in awe that I was actually able to fix it

[diif]

Re: Did I actually fix it?

No, it's age that caused the cap to fail not the memory.

[momaka]

Re: Did I actually fix it?

Did I actually just fix it?

No you didn't! (And sorry to rain on your parade like that. )

That motherboard is filled with 6.3V 1000 uF OST RLP capacitors (as if it is some kind of a tradition with Biostar motherboards), and they are ALL likely in various stages of failure at this point.... not so much due to age, but because OST RLP series are CRAP. (In fact, OST in general is crap... save for RLS series, maybe, sometimes, with a pinch of luck. )

How I know?

I buy broken motherboards for fun on eBay when I find one for cheap. I have several Biostar motherboards that I bought exactly because they were listed as not working, and they all came with... guess what? - Blown OST RLP caps. Some had just a few of these blown and the rest looking fine, while others had just about every cap bad. The worst part is that even the ones that didn't appear bloated showed signs of pending failure (typically over 20% higher capacitance than specced.)

In fact, I was actually looking at this exact same motherboard model - TForce TA770A2 - listed on eBay for parts about a month ago, because it had blown caps. I was going to buy it, but eventually someone beat me to it while I was making up my mind if I should get that motherboard or wait a few more days and get another item I was looking at.

In any case, you will want to get rid of ALL of those OST RLP caps if you want that motherboard to work reliably in the long term.

For those OST RLP caps, you can use either Rubycon ZLH 6.3V 820 uF (in stock from Digikey, Mouser, and badcaps.net too!) or Rubycon ZLQ 6.3V 1200 uF... or mix and match between these (that's what I do, usually.) United Chemicon KY 6.3V 1000 uF will work too, as would Nichicon HE with those specs. Those OST RLP c[r]aps don't have really good specs anyways, so even quality Japanese GP caps will work in a pinch - at least for a temporary recap anyways.

[SluggerB]

Re: Did I actually fix it?

No you didn't! (And sorry to rain on your parade like that. )

I'm not going to spend the time to recap the entire board, if it starts to flake again I'll replace them as needed. So far since replacing this one it's been rock solid so I am happy. And I would like to build a new system anyway. The one that went bad was a 10v 1000mF, not 6,3v, those still seem to be chugging along.

I'm still uber proud of myself for fixing it, this was the first thing of my own that I actually care about/use all the time that I managed to fix with my nacent electronics knowledge. Don't burst my bubble!

[ribcage]

Re: Did I actually fix it?

11 years is not old at all. I am pretty sure your 16GB ram would work now that you replaced that cap, I had a very similar issue. But practice soldering before soldering on mobos, one hour of heat can easily render the mobo unusable, with proper iron, heat and technique, you should get the cap out in 5 seconds.

[momaka]

Re: Did I actually fix it?

Don't burst my bubble!

Yeah, you're right.
Sorry if I indeed sounded like an ass.

It's just that sometimes, a few of us long-time badcappers can get a little "triggered" when we see this 'replaced just the bad caps and it works fine now' scenario. And it's not really because we are trying to make you do things a certain way (because "our way" is "better", or whatever lol), but rather because I think we've all been through this path that you went through at some point... and it's always eventually turned out that more bad caps would pop along if you left them the first time.

Now, you may get lucky, and the board may just stop POSTing or it may just crash randomly. But sometimes, bad caps can make other components fail, and that makes some repairs very not worthwhile... and one reason why this thread exists: This is Why You NEVER Ignore Bad Capacitors!

So, on that note, I have to concur with the advice that ribcage suggested about practicing your soldering a bit more and then tackling this board again.

The one that went bad was a 10v 1000mF, not 6,3v, those still seem to be chugging along.

It's all the same. OST RLP tend to have the same failure mechanism. If you have a cap meter, you'd probably find that actually may or may not be true for all of them.

My observation (and I'm sure of a few other members here too who have been replacing bad caps for a long time) is that when one cap from a certain brand fails on a piece of equipment, the rest from that brand and series may (and likely will) follow suit shortly, if not already (but just aren't showing it yet.) This is especially applicable to known problematic cap brands / series.

Here is one ASUS P5GC-MX I fully recapped recently:
https://www.badcaps.net/forum/showthread.php?t=92451
It started off with just one cap going bad a way while back... then another... then another... and so on. Luckily, I got the majority of the important caps recapped before putting it in full-time service. So when more of the last few KZG caps failed, it didn't cause any problems, because I already knew what those caps did and that they couldn't affect the stability of the board. But I replaced them, nevertheless, when I pulled out the system again for a general inspection and cleaning. 30 minutes later, and no more KZGs.

Now do things as I say and not as I do!
If you look at the pictures on that ASUS P5GC-MX I posted, you may spot I used a few... OST RLP caps, of all things, for the filter on the USB ports. These aren't original to the board. I used those in my recap to replace the blown KZGs (so again, do as I say and not as I do ). Of course, it should be noted that in my case, these OSTs came from a motherboard that had cap problems from a different brand and series, and none of the OST were failed or from high-stress areas (all read good on my meter too.) I have them placed in a low-stress non-critical areas on the P5GC-MX (i.e. the USB port filters), so they should be OK there and cannot cause any stability issues or make other component fail. So while not a good idea, you can sometimes get away with "tricks" like that.

With that said, you don't really need to do a full-blown recap on your board. But if you do get a chance in the future (i.e. time and interest), you should replace at least all of the "important" caps on your board - that is, those that filter power to the RAM, chipsets, and CPU. The ones by the PCI/PCI-E slots and USB ports don't usually matter too much and can be left (that is, if you wanted to save on time and use less caps.)

I'm still uber proud of myself for fixing it, this was the first thing of my own that I actually care about/use all the time that I managed to fix with my nacent electronics knowledge.

Oh I remember that feeling too when I did my first fix - it was good!
On that note, I do congratulate you on yours as well.

[SluggerB]

Re: Did I actually fix it?

Yeah, you're right.
Sorry if I indeed sounded like an ass.

It's just that sometimes, a few of us long-time badcappers can get a little "triggered" when we see this 'replaced just the bad caps and it works fine now' scenario.

I wouldn't be surprised if the board is on it's way out and needs full recap, but I'm just interested in getting it into a stable limp along mode. If it starts crashing again I don't mind replacing the caps 1 by 1


But you may be more prescient than you think, I woke up this morning to a windows messaging saying it recovered from a BSOD and checked the up time and it did spontaneously reboot around 2AM this morning . It may be been a fluke, if not I'll be back at it again

[SluggerB]

Re: Did I actually fix it?

Yeah, you're right.
Sorry if I indeed sounded like an ass.

Ddue no worries the help I have gotten on this forum has been tremendous and I am forever grateful to have found such a great resource for learning about this stuff


I took a look at the error message and it mentioned it was a blue screen error. When I had the bad cap, I never got that blue screen error, the PC would just shut down like it lost power. So they may be entirely unrelated. Either way I just snagged a new motherboard for a new PC build since it showed up back in stock (well, new to ME, I'll be moving from ancient hardware to old hardware lol), so in a month or so I should have a new system and this PC will be mostly cannibalized/retired.

[momaka]

Re: Did I actually fix it?

I wouldn't be surprised if the board is on it's way out and needs full recap, but I'm just interested in getting it into a stable limp along mode.

I know what you mean: just keep it working unit a replacement arrives shortly - I've been guilty of doing that myself ... though with nothing I use on a daily basis.

If it starts crashing again I don't mind replacing the caps 1 by 1

But see, that's the thing with bad caps - it may not start crashing and instead could just instantly burn out a regulator or MOSFET somewhere the next time a cap fails. It all depends which cap fails and how well-designed the motherboard's power circuits will be to handle that failure. So you could well wake up to a room full of magic smoke one morning. I don't mean to scare you or anything like that, of course, as the chances of that happening are relatively low. But I have seen it and had to repair plenty of hardware that has gone through that (and made the repair that much more complicated.)

That said, and with a bit of luck, your motherboard will probably be able to make it until your new one arrives shortly. But IMHO, it's a decent AM2+/AM3 motherboard and you should still recap it if you'd like to have a spare (or as a hobby / soldering practice??)

But you may be more prescient than you think, I woke up this morning to a windows messaging saying it recovered from a BSOD and checked the up time and it did spontaneously reboot around 2AM this morning . It may be been a fluke, if not I'll be back at it again

I mean, it could be the mobo, or it could be something else. But in my experience with these OST capacitors, once a few start failing, the others will soon follow. Could be days, could be weeks, or it could be months, and the failure may progress slowly or suddenly.

Ddue no worries the help I have gotten on this forum has been tremendous and I am forever grateful to have found such a great resource for learning about this stuff

Yeah, I feel exactly the same about this place... which is why I've hung along for so long.

When I had the bad cap, I never got that blue screen error, the PC would just shut down like it lost power. So they may be entirely unrelated.

Oh no, don't look at it like that. Let me give you a good story here (at least I think it is), if you don't mind.

Some years ago, I used to volunteer for the IT department of a local non-profit. They got these used HP DC5000 SFF desktops donated to them, and I did the dust cleanup and OS setup on these - all HDDs cloned identically using the same exact Windows XP Pro configuration (since we had a legitimate XP VLK key.)

Most of these desktops worked "OK" - or so it appeared on the surface. While indeed most did not crash, I noticed during cleaning that at least 90% of them (about 30-odd machines) came with various caps bulged inside. After a few weeks of use, certain machines started getting "strange" errors: one or two refused to POST anymore or would just BSOD. But there were also a few very "interesting" ones. These would crash only under certain circumstances, like watching a video or loading a few very specific websites in Internet Explorer (yes, I.E. - back then, it still was considered somewhat "OK" to use )

Now you may think there are too many factors involved to pinpoint such an issue so specifically - like what video was played specifically and what else the user was doing when the issue happened. But that's not the case at all! We were able to re-create the issue 100%. Since all of these were the same HP DC5000 machines with same exactly cloned HDD, they were absolutely identical in all aspects. So on the machines that crashed specifically with video and IE, I put their HDDs in other DC5000 that I just recapped for the IT dept. And guess what? - All the issues disappeared. But if I swapped back the HDD in the original machine, the issue re-appeared. To re-assure myself that nothing else was playing here, I even went as far as swapping the CPU and RAM between the PCs. However, each time the issue followed the PC wit the faulty caps.

The more interesting (and perhaps funny) part IMO, is the issue itself: it did not cause the PC to reboot or BSOD (unlike some of the other same machines with bad caps) - IE simply crashed with a generic Windows memory error message or IE would just shut down without warning - as if someone had hit the "X" button. That's it - no hard crash or anything. And this issue was 100% re-creatable.

So in short, bad caps can cause all sorts of random issues: from reboots and BSODs to Windows errors or simply the application shutting down without warning. As I was recapping these machines, we saw more of them come up with this issue, along with others like BSODs and whatnot. Thus, don't expect that bad caps can cause only one specific issue - it can be anything.

In the case of the above motherboards, the bad caps were everywhere on the boards, though different boards would have bad caps appear in different spots.

On that note, I even created a thread here on how to poly-mod these boards:
https://www.badcaps.net/forum/showthread.php?t=23503

[sam_sam_sam]

Re: Did I actually fix it?

Oh no, don't look at it like that. Let me give you a good story here (at least I think it is), if you don't mind.

Some years ago, I used to volunteer for the IT department of a local non-profit. They got these used HP DC5000 SFF desktops donated to them, and I did the dust cleanup and OS setup on these - all HDDs cloned identically using the same exact Windows XP Pro configuration (since we had a legitimate XP VLK key.)

Most of these desktops worked "OK" - or so it appeared on the surface. While indeed most did not crash, I noticed during cleaning that at least 90% of them (about 30-odd machines) came with various caps bulged inside. After a few weeks of use, certain machines started getting "strange" errors: one or two refused to POST anymore or would just BSOD. But there were also a few very "interesting" ones. These would crash only under certain circumstances, like watching a video or loading a few very specific websites in Internet Explorer (yes, I.E. - back then, it still was considered somewhat "OK" to use )

Now you may think there are too many factors involved to pinpoint such an issue so specifically - like what video was played specifically and what else the user was doing when the issue happened. But that's not the case at all! We were able to re-create the issue 100%. Since all of these were the same HP DC5000 machines with same exactly cloned HDD, they were absolutely identical in all aspects. So on the machines that crashed specifically with video and IE, I put their HDDs in other DC5000 that I just recapped for the IT dept. And guess what? - All the issues disappeared. But if I swapped back the HDD in the original machine, the issue re-appeared. To re-assure myself that nothing else was playing here, I even went as far as swapping the CPU and RAM between the PCs. However, each time the issue followed the PC wit the faulty caps.

The more interesting (and perhaps funny) part IMO, is the issue itself: it did not cause the PC to reboot or BSOD (unlike some of the other same machines with bad caps) - IE simply crashed with a generic Windows memory error message or IE would just shut down without warning - as if someone had hit the "X" button. That's it - no hard crash or anything. And this issue was 100% re-creatable.

So in short, bad caps can cause all sorts of random issues: from reboots and BSODs to Windows errors or simply the application shutting down without warning. As I was recapping these machines, we saw more of them come up with this issue, along with others like BSODs and whatnot. Thus, don't expect that bad caps can cause only one specific issue - it can be anything.

In the case of the above motherboards, the bad caps were everywhere on the boards, though different boards would have bad caps appear in different spots.

On that note, I even created a thread here on how to poly-mod these boards:
https://www.badcaps.net/forum/showthread.php?t=23503

Thank you for posting this

I do have a question for you about this

Can you tell if it an issue with the mother board if you do not see any capacitors blowing short of just testing them or is this just the capacitors that you can see the issues with them

[momaka]

Re: Did I actually fix it?

I do have a question for you about this

Can you tell if it an issue with the mother board if you do not see any capacitors blowing short of just testing them or is this just the capacitors that you can see the issues with them

It varies and really depends on what capacitors are on the board and what their function is. Some crap cap brands will not look bulged but can still be bad (either ESR too high or capacitance too high from excess leakage or capacitance low.) And depending on where the bad caps are located, you may or may not have an issue with stability.

In the case of the P4SD above - all had bad caps, but just in different spots. Some had just a few bad Rubycon MCZ around the CPU... but due to them being in parallel with a few polymers, these usually did not cause issues, so long as only 1 or 2 were bad. The RAM V_dd also didn't always cause problems when there were bad caps on it, but more often it did. And the small Teapos scatterred around the rest of the board - these rarely caused problems. But in one instance, the mobo wouldn't detect any IDE or USB drives, and in another it crashed with high HDD activity. So there really wasn't that much of a pattern between the failed caps and the failure modes, aside from that one case with Internet Explorer that happened on a couple of motherboards specifically and which was fixed with the recap of those boards.

So to answer you question shortly - there really isn't a sure way to say if the caps are the issue or not on a board that appears to have no bad caps, but the caps themselves are known problematic brands and series. In such cases, I just remove a few "key" filter caps and replace with good ones if board is unstable. If it continues to be unstable, it may not be the caps and might be something else (nVidia chipsets, for example, are known to have poor reliability.)

[SluggerB]

Re: Did I actually fix it?

But see, that's the thing with bad caps - it may not start crashing and instead could just instantly burn out a regulator or MOSFET somewhere the next time a cap fails. It all depends which cap fails and how well-designed the motherboard's power circuits will be to handle that failure. So you could well wake up to a room full of magic smoke one morning. I don't mean to scare you or anything like that, of course, as the chances of that happening are relatively low. But I have seen it and had to repair plenty of hardware that has gone through that (and made the repair that much more complicated.)

I understand that. All my data is backed up weekly to an external USB hard drive. So if this PC just imploded one day randomly, I could just unplug the drive, pop it in my laptop, and wouldn't skip a beat.

Aside form that one reboot, everything is still solid, so the system seems stable enough to get me by until it is replaced, which was on my radar anyway. This board is maxed out CPU and RAM wise and I want something with more upgrade headroom while it is still available to be purchased

That said, and with a bit of luck, your motherboard will probably be able to make it until your new one arrives shortly. But IMHO, it's a decent AM2+/AM3 motherboard and you should still recap it if you'd like to have a spare (or as a hobby / soldering practice??)

I'll probably have the new system up an running on a week or so as other parts arrive, all I really need is RAM and everything else is coming over, I am keeping the same CPU until chip prices come down. I may hold onto the board for cap replacement practice though.