Assistance with identifying replacement caps on ASUS P5VDZ-MX

[crosswinds]

Hello,

So, pulled a motherboard out from storage, was working when put away, wont post when I took it out of storage today.
Since it was working, and it looks to be essentially time that killed it I figured its a good idea to replace the caps before digging further.
Looks to be only good quality caps on this board, so might not be the issue. But it is closing in on 20 years now so, wont hurt i guess

Its a ASUS P5VDZ-MX board.

I would really appreciate some assistance on what the best replacement caps would be for these:

Panasonic(M) FL 1000uF 16v
Panasonic(M) FJS 820uF 6.3v
Rubycon MBZ 470uF 16v

And the most difficult of them all. It looks to be polymer caps from what I can gather. Through-hole with markings:
691
561
04v
(I attach a image of these)

[andrewsawesomr]

For the Rubycons: Panasonic EEUFR1E471B

[stj]

i would just go full polymer

[crosswinds]

Thanks for your input!
The most difficult ones to identify values for are the ones with red markings on. I do suspect those might be polymer, but not sure. Not really finding any sources online for these markings.
Any clues on those?

[stj]

match the markings
http://capacitor.web.fc2.com/solidcapacitor.html

[crosswinds]

ALright so with that it looks like it is a Fujitsu cap, Series 691, 560uF capacitance and 4V. Does that sound reasonable?
I do not thing i can get hold of this exact series or manufacturer. But I can get hold of Panasonic caps for example.

Can all the other ones be changed to polymer caps? Any pitfalls here in term of ESR and so?

[crosswinds]

Hello! So I pulled the capacitors out of circuit.

Measured ESR and the capacitance on all of them. On the ESR meter, all of them but 3 measured 00 on the meter, the other three was 0.01-0.02. This is with the "EVB" ESR meter bought a couple of years back. Not sure if its valid for these types of caps though.
However, according to those readings it does look like all of them had good ESR? Thats a bit unfortunate since I was hoping the issue was caps.

Though, capacitance measurements was a little bit more "all over the place" with:


Not sure if these measurements could indicate some form of issue? Anyway, since they are now out, I need to order anyway

Any input is appreciated!

[stj]

firstly, polymers other than ancient ones in blue or purple sleeves from oscon dont really fail,
the ones in the pic are 560uf - they are numbered with the last digit being the number of zero's
you can replace the others with polymer if you can match the diameter and you dont want the best.
the best polymers will be far too low esr compared to the original caps.
thats why there are missing parts between those polymers - the board was designed to be stuffed with electrolytics

try looking at Kemet polymers - they are very cheap.
or get surplus caps from aliexpress.

[crosswinds]

Yeap, the polymers most likely are good then. Probably the others also. I jumped to quick on the caps in this case
However, it was the only thing i could think of that would die on the "shelf"

I picked panasonic FR for the electroytics on the list, i guess that would be fine?

And the replacements for the polymers i got to this https://www.mouser.se/ProductDetail/667-4SEPC560MX

[stj]

FR is fine for anything other than the cpu - which is why it had polymers

[crosswinds]

Great!
Thanks for your help with this.

What do you think about the polymers i linked? Would those be valid for the CPU?

[momaka]

I highly doubt the issue is with the caps here. All of the electrolytic and polymer caps are good brands, so they are extremely unlikely to fail... and your ESR readings confirm this. So personally I wouldn't waste any time with a recap here.
FWIW, I have audio gear with 30-40 year old Japanese electrolytic caps, most of which still measuring in spec after all this time.
Electrolytic caps don't (easily) go bad sitting on the shelf. It takes many (10+) years of no use and a very sensitive circuit for any issues to arise... at least for the good Japanese brands.

How was the motherboard stored? In an anti-static bag? Was is also handled properly if/when it was taken out of the anti-static bag?
Elevated humidity / air moisture can also tarnish contacts overt time. So it may be worthwhile to re-seat the CPU in the LGA socket. Also, if BIOS is in a socketed holder, try re-seating that too. And same goes for RAM - do a few insertions and removals of RAM sticks in the slots to clear any possible oxidation.

Oh, and I'm assuming you checked and tested the board with a good CMOS coin battery, just to rule that out as an issue.

Last but not least, do you know what kind of cooler was last used on the CPU? And how about now?
Although not super common, but I've seen the BGA fail on CPU sockets due to stock Intel pushpin -type coolers... and I always recommend to avoid these whenever possible. Many of them warp/flex the board too much, which can cause damage over time.

[crosswinds]

Hello,

Thanks for your input!

I agree that its a long shot with the caps.
However, it was first resort as I know it worked before being stored, im the original owner of the board, and it has always been well taken care of. So i figured this will be one of the things that actually time takes a toll on the shelf.
This combined with that i have another asus board, same socket, around same years,however with more hours on, that has alot of bulging caps.

So for your questions:
It was stored in a static bag on a shelf in the office. Not like a damp humid place.
CMOS coincell new
I have reseated CPU, tried another cpu.
Reseated bios eeprom.
tried different rams.

I tried it with a postcard, and it was giving me flaky results. Stopping pretty early, but with different results some times, so not always the same.

The thing with the cooler is interesting. It has always had the original intel cooler on it. I remember back in the day I never liked the force that you had to put on it. I read now that this might been a issue.
I can see that there is a slight warp around the cpu area. So maybe this very well can be the issue. Which would be very unfortunate.

However, since I removed the caps to really be able to check them, I did order new for it, as i do not want to give up without some fight at least. Was already going to order for the other board anyway.

[ChaosLegionnaire]

try clearing cmos as well using the jumper if u havent. i noticed some boards, if left in storage for a decade or more and the cmos battery is left to die gradually that way, the cmos settings somehow become corrupt and the board just refuses to post unless u replace both the cmos battery AND clear the cmos then u power it on.

another method u can try is momaka's hammerfist blow method. he uses it on nvidia shitset motherboards with bad solder bumps in the chip. inspired by him, i have also used it on a motherboard i bought that unfortunately turned out faulty. when i shoved and pushed my hand really hard on the mobo nb heatsink, the board successfully posts and boots the os to desktop successfully but as soon as i take my hand off, the board insta freezes, hangs and stops working, so its useless unfortunately. i'm not a muscleman with 20 inch biceps and triceps so i cant use the board with one hand permanently shoved on the board while using the other hand on the keyboard and mouse. its terribly inefficient too esp. if one hand gets tired, its not easy to change to the other hand without freezing up the system and losing everything i did so far.

so u can try shoving and pushing your hand really hard on the mobo nb heatsink or the cpu heatsink to see if that changes the post codes u get but unfortunately, its also a via chipset board. via chipsets arent too reliable but they're not bad. just decent, average or okay. i have a couple of faulty via boards from the athlon 64 era. so if that changes the post codes or the board suddenly starts working with one hand shoved on it, u do indeed have a board with bga problems from bad flexion of the pcb. so yes, divert your attention to pre-emptively recapping the other working board with imminently failing caps.

[crosswinds]

Hello!

Thanks all for your tips!

So, I received caps to both boards today.

Soldered all in today on this one and as suspected, Nothing changed.
However, I did also get a new post-card and it is consistently stopping at C1 with the DRAM LED on the postcard lit. So not sure if the other postcards was flaky or something changed with new caps. The latter is doubtful though.

C1 is memory from what i can gather. And i tried 3 different sticks, always the same code.

Postcodes are as follow after powerup: 00 -> FF ->CF ->C0 -> C1

I did not try to bend the board, to be honest. The warping is for sure not bad at all, so i really want that to be the last resort.
Any further hints on what to check? I do have access to oscilloscope, multimeter & all in it for learning.
Did also clear cmos with the jumper.

If its recommended to make a new thread for the board, please let me know, as this just started as a identify cap thing, but now it turned in to a quest

[crosswinds]

Alright. I had a small development with this.

Armed with very little knowledge about motherboard component level troubleshooting, some time and a multimeter. I started to randomly check the transistors on the board.

since we have c1 memory post code, I started around the RAM slots. I found one that is shorted between Gate - Source, which im pretty sure is not supposed to be?
Will need to pull it to really confirm. But i dont think something else in the circuit might cause this gate-source short symptom? Am i barking up the wrong tree?

Please see the attached image, it is the one closest to the screw hole(if i did not confuse them now writing, but one of these three). The others does not show sign of this.

[stj]

ceramic capacitors can go short circuit a lot - see if the transistor has a parallel ceramic cap

[crosswinds]

ceramic capacitors can go short circuit a lot - see if the transistor has a parallel ceramic cap

Great tip,Thanks for it!

I did a quick check for any ceramics in the near viscinity of the transistor. But nothing there.
Will need to trace it out thoroughly to be sure though.

[momaka]

Hello!
Soldered all in today on this one and as suspected, Nothing changed.
However, I did also get a new post-card and it is consistently stopping at C1 with the DRAM LED on the postcard lit. So not sure if the other postcards was flaky or something changed with new caps. The latter is doubtful though.

Some POST cards are indeed flaky... well, more like, it depends on which PCI slot they are inserted. The one I have is like this - on some motherboards, I don't always get POST codes, depending on which PCI slot I used. Also, it doesn't appear to use gold contacts but rather tin contacts on its edge connector, so sometimes it takes a bit of "wiggling around" in the PCI slot for it to read codes.

If its recommended to make a new thread for the board, please let me know, as this just started as a identify cap thing, but now it turned in to a quest

I think it's OK.
You could ask the mods to move it to the Motherboard repair section if you like.

Armed with very little knowledge about motherboard component level troubleshooting, some time and a multimeter. I started to randomly check the transistors on the board.

since we have c1 memory post code, I started around the RAM slots. I found one that is shorted between Gate - Source, which im pretty sure is not supposed to be?
Will need to pull it to really confirm. But i dont think something else in the circuit might cause this gate-source short symptom? Am i barking up the wrong tree?

Please see the attached image, it is the one closest to the screw hole(if i did not confuse them now writing, but one of these three). The others does not show sign of this.

Generally, a short between Gate and Source (or Gate to anything) on a MOSFET is never a good sign.

However, in this case, you can see on that bottom left MOSFET you pointed to (close to the screw hole) that both the Source and the Gate are connected to the same light(er) yellow trace... which means they are electrically connected together. Now, if you're wondering why, sometimes manufacturers do this to turn MOSFETs into power diodes. With Gate tied to Source, only thing left on the MOSFET is the Source-Drain "body" diode. It's usually capable of handling a few Amps of current... though it's not a very efficient method, since MOSFET body diodes tend to have large voltage drop (so they heat up a fair bit.) Of course, this could also be done purposefully too, to drop voltage down... sort of like a linear regulator.

I agree that its a long shot with the caps.
However, it was first resort as I know it worked before being stored, im the original owner of the board, and it has always been well taken care of. So i figured this will be one of the things that actually time takes a toll on the shelf.
This combined with that i have another asus board, same socket, around same years,however with more hours on, that has alot of bulging caps.

Let me guess.. that other board has brown UCC "KZG" caps? - If so, those are known to be problematic and nothing to do with the board sitting on the shelf or having many hours of use.

However, since I removed the caps to really be able to check them, I did order new for it, as i do not want to give up without some fight at least.

Make sense, I suppose - once the old caps are removed, you could get new ones.

That said, save the old caps from this board, as they should all still be perfectly usable. If you dabble with retro hardware, you'll find then useful sooner rather than later.

try clearing cmos as well using the jumper if u havent. i noticed some boards, if left in storage for a decade or more and the cmos battery is left to die gradually that way, the cmos settings somehow become corrupt and the board just refuses to post unless u replace both the cmos battery AND clear the cmos then u power it on.

+1

I can't remember now on what board I ran into this issue. But yes, I always check manual on which way to clear CMOS. On many boards, removing the CMOS battery (and disconnecting the PSU) is enough. But I do recall running into one board that only partially cleared the CMOS with the battery removal.

another method u can try is momaka's hammerfist blow method. he uses it on nvidia shitset motherboards with bad solder bumps in the chip.

Ah, yes, that be the ECS mobo on my Gateway GT5656 PC. It has a GeForce 6150 chipset that's going on the way out. Haven't reflowed it yet, because touching (putting a bit of pressure) on the chipset heatsink usually gets it going and it stays "fixed" for a while. When it gets worse, I'll probably pull it and give it a re-heat. Only reason I don't want to do that is because I would also have to remove some of the nearby caps, and it's just a chore doing that.

inspired by him, i have also used it on a motherboard i bought that unfortunately turned out faulty. when i shoved and pushed my hand really hard on the mobo nb heatsink, the board successfully posts and boots the os to desktop successfully but as soon as i take my hand off, the board insta freezes, hangs and stops working, so its useless unfortunately. i'm not a muscleman with 20 inch biceps and triceps so i cant use the board with one hand permanently shoved on the board while using the other hand on the keyboard and mouse. its terribly inefficient too esp. if one hand gets tired, its not easy to change to the other hand without freezing up the system and losing everything i did so far.

You know there's a more "elegant" method to doing this.
https://www.badcaps.net/forum/troubl...28#post1489828

via chipsets arent too reliable but they're not bad. just decent, average or okay. i have a couple of faulty via boards from the athlon 64 era.

Interesting you say this, as it seems to confirm my suspicion that the Via NB in my ASUS K8V-SE Delux is also starting to die. The last couple of times I powered it on, I got artifacts on boot. I surely thought it must have been my XFX GeForce 6800 XT AGP card that's going bad. Swapped it for a GeForce 7600 GS AGP that I knew worked fine and... same thing with the artifacts. So I checked both cards on a different motherboard, and they are both fine. Put them back in th K8V-SE Delux, and back to artifacts, at least on a cold boot.

So I suppose certain Via chipsets could also be problematic?
If so, that would be very disappointing. That means from all of the retro hardware that's out there, only Intel chipsets can be trusted in the long run. (And maybe a few AMD ones??)

[crosswinds]

Hey,

Thanks for all the info here!

Even though it killed my hopes for the mosfet to be the culprit. I was really hoping for it. But I see now in the picture what you mean.
Did not even think this was a thing to do it as you said, but it sure do look like it.

Back to the drawing board with this one. Eventhough I dont really know what the next step should be