Announcement

Collapse
No announcement yet.

ASUS P5GC-MX motherboard recap

Collapse
X
 
  • Filter
  • Time
  • Show
Clear All
new posts

    ASUS P5GC-MX motherboard recap

    Here’s another motherboard that needed a full recap: an ASUS P5GC-MX.
    This one was gifted to me some years ago by user Pentium 4, along with a few other goodies. It actually came in working order with no bulging or leaking caps. However, I noted there were United Chemicon KZG caps everywhere on the motherboard. The CPU VRM output (CPU V_core) was the only exception: it had only 2x KZG. The rest was 6x UCC TMV 4V 680 uF caps… which aren’t any good news either.

    So here is what the motherboard looked like with its original caps:


    CPU VRM area up close…
    https://www.badcaps.net/forum/attach...1&d=1611813246

    And the rest of the lower half of the motherboard:
    https://www.badcaps.net/forum/attach...1&d=1611813246

    Also including a back-side picture below, just because I really like the bright yellow PCB on those old ASUS motherboards. (I really wish manufacturers would stop making just black motherboards, like it is nowadays.)
    https://www.badcaps.net/forum/attach...1&d=1611813246

    I didn’t have any immediate plans for this motherboard at the time when I got it, but it was one of the best motherboards I had back then, so I started thinking what kind of system I should build with it. In the meantime, I decided to at least recap the CPU VRM output, since I remembered seeing threads here on badcaps.net that showed UCC TMV caps going bad. Moreover, I was told the motherboard had several 10k’s worth of power-on hours. So that weighted in even further on the decision to do a preemptive CPU VRM recap, rather than wait for caps to start failing on their own.

    Because these Chemicon TMV caps are right along the CPU socket (and in the path of hot air from the CPU heatsink), I thought it would be best to use polymers here. Also, since buck-regulated circuits put more stress on caps, I went after a few of the KZG on the RAM V_DIMM (RAM V_dd) as well.

    And speaking of voltage rails, it wouldn’t make sense to talk about them without showing a cap diagram. So here is that:


    Since no socket 775 CPU runs at over 1.4V on its core, 2.5V polymers on the CPU VRM output (CPU V_core, rail #2) should be fine. For these, I used 6x Sanyo SEPC 2.5V, 820 uF – a slight (but welcome, I’m sure) boost in capacity over the original 680 uF TMV. Meanwhile, the RAM V_dd runs at 1.8V (though some DDR2 RAM could require up to 2.1V), so I figured it’s probably safer to use caps rated for at least 4V there. Unfortunately, I didn’t have any 4V polymers at the time. However, I did have 2x 6.3V, 560 uF Apaq –branded polymers that came from a newer parted ASUS motherboard. So I pulled one 6.3V 820 uF KZG cap from the RAM V_dd rail and then substituted those two Apaq polymers. I then concluded the recap done for the time being, which was this:


    After this I put the motherboard in storage, due to being busy with college (and I still hadn’t made up my mind what components and case to use.) When I pulled it back out of storage again a few months later, I was greeted with this:


    Yup, there they go – KZG sitting in storage and failing. This is not a myth, folks! Unfortunately, I didn’t have enough caps in stock at the time, so I only replaced that single bad KZG with a Rubycon ZLH (6.3V, 820 uF) and then shelved the motherboard once again. Then, once more, I came back to it two months later, pulled it out of storage to count how many caps I need to order for it and….another KZG had failed!


    It was clear now I needed to do a full motherboard recap. But at this point, I already had found another PC for gaming, so I was no longer in a hurry to fix this one. Eventually, however, my parents’ old PC (a Dell Dimension 3000) really needed to be replaced with something newer. So finally I got a bunch of caps and did an almost-complete recap on the P5GC-MX. I only left a few KZG caps on there, simply due to running short on caps again (as I had bought/found other hardware that needed new caps too.) Most of these were in low-stress areas, such as filters for the USB ports and PCI/PCI-E slots, so no big deal if any failed there. After this I built a PC for my parents with this motherboard and an Intel Dual Core E2160 that was also sent to me by Pentium 4. The motherboard worked great without ever skipping a beat for the next few years.

    Then I decided to take a look inside that PC one day somewhat recently, since it had been a few years and I thought it might need a dust cleaning maybe… plus, what a better time to check on the last few remaining 6.3V 820 uF KZG caps left. Sure enough, I found more bad ones! This was the motherboard as of August 2020:


    Basically, 3 more KZG’s had failed of the remaining 9 on the board.
    https://www.badcaps.net/forum/attach...1&d=1611813392
    https://www.badcaps.net/forum/attach...1&d=1611813392
    Two of these were for the USB 5V filtering, and one for the 3.3V rail for the PCI/PCI-E slots. The other 6 looked OK and tested OK.

    This time around, though, I actually had all the caps needed to complete the recap. So I pulled the rest of the 6.3V 820 uF KZG out and replaced them with more Rubycon ZLH of same voltage and capacity.


    The only United Chemicon KZG that remain now on this motherboard are the 4x 16V 1000 uF units on the CPU VRM high-side. From what I have gathered over the years, 16V 1000 uF KZG don’t seem to be a common failure (I actually don’t recall seeing any.) Moreover, my parents have, like many other folks, started to move more towards using tablets. So this PC would rarely see any use anymore. Thus, as an experiment, I decided to leave the 16V 1000 uF KZG in there. I pulled one a few years back and it tested perfectly in spec, so I think it should be OK. The whole case has good cooling and those caps don’t run hot, so I do expect them to last. As another experiment, I even put some OST RLP for the filters on the USB ports in place of the 6.3V 820 uF KZG, since low-ESR caps are not really needed there. We will see how long those last, but I think they will be fine too. The PC barely gets any use now – mostly when someone needs to print or scan something… or when those silly tablets can’t open a page properly. As such, it sits turned OFF most of the time (but plugged in.)

    For those who might be curious (*ahem* Mr. Per Hansson *ahem* ), I got the following date codes for 6.3V 820 uF KZG’s from this motherboard:
    6(k) DI <-- 4x total; 2x read good, 2x bulged and leaking
    6(k) DK <-- 3x total; all bulged and leaking
    6(k) DG <-- 2x total; one good and one bulged / leaking
    6(k) DH <-- 1x total; bulged and leaking
    6(k) Nc <-- 1x total; still looks OK
    6(k) Na <-- 3x total; all measure good

    And the two 6.3V 1500 uF KZG:
    7(k) 1H <-- 2x total; OK at time of removing, but went bad a few years after in storage on a temporary recap project (XFX GeForce 6800 XT video card.)

    There was also a lone 16V 470 uF KZG for the 12V rail going to the PCI-E slot:
    6(k) DL <-- measured good

    The 4V 680b uF TMV did not have any date codes stamped on the side. However they do have a “71 L3” stamped on top, so maybe that’s their date code?

    And unfortunately, I can’t read the date codes off of the remaining 4x 16V, 1000 uF KZG, due to them still being installed in the motherboard. But if they do fail, I will try to remember to add them here. Given the other KZG caps above, though, I imagine these have a similar date code – i.e. made around 2006 or 2007.

    Anyways, now with all the 6.3V 820 uF gone, I don’t think I’ll be seeing any bad caps on this motherboard anytime soon. But we will see, of course… as always.
    Attached Files
    Last edited by momaka; 01-28-2021, 12:05 AM.

    #2
    Re: ASUS P5GC-MX motherboard recap

    Nice writeup, I guess KZG may actually fare worst being powered off.
    Just like you I have a PC motherboard in storage and they (KZG) get worse with time.
    Of the industrial systems I service it seems that the ones that have not leaked yet are not going to...
    "The one who says it cannot be done should never interrupt the one who is doing it."

    Comment


      #3
      Re: ASUS P5GC-MX motherboard recap

      Just realized something: that board I mentioned, the caps that have failed are momaka's favorite KZG 1500uF 16v
      So maybe you will have to replace yours too eventually
      Attached Files
      "The one who says it cannot be done should never interrupt the one who is doing it."

      Comment


        #4
        Re: ASUS P5GC-MX motherboard recap

        odd those bloated when its an athlon 64 system u have there... msi k8n neo4, isnt it? athlon 64 systems dont really run hot. at least not at the level of the presshots... how's the psu and the ripple on the 12v output during load? dont tell me it failed in storage de-energized?

        Comment


          #5
          Re: ASUS P5GC-MX motherboard recap

          Originally posted by Per Hansson View Post
          Just realized something: that board I mentioned, the caps that have failed are momaka's favorite KZG 1500uF 16v
          So maybe you will have to replace yours too eventually
          I suspect 2004 KZGs and 2005 KZGs are the worst! And those seem to have the same bulging issue as the 3300uFs on my 2004 Asus A7N8X-X, IIRC.

          The KZGs failed when not powered up, IIRC! IIRC, when sitting one day in 2015, saw them bulging and leaking!
          Last edited by RJARRRPCGP; 02-03-2021, 12:32 AM.
          ASRock B550 PG Velocita

          Ryzen 9 "Vermeer" 5900X

          16 GB AData XPG Spectrix D41

          Sapphire Nitro+ Radeon RX 6750 XT

          eVGA Supernova G3 750W

          Western Digital Black SN850 1TB NVMe SSD

          Alienware AW3423DWF OLED




          "¡Me encanta "Me Encanta o Enlistarlo con Hilary Farr!" -Mí mismo

          "There's nothing more unattractive than a chick smoking a cigarette" -Topcat

          "Today's lesson in pissivity comes in the form of a ziplock baggie full of GPU extension brackets & hardware that for the last ~3 years have been on my bench, always in my way, getting moved around constantly....and yesterday I found myself in need of them....and the bastards are now nowhere to be found! Motherfracker!!" -Topcat

          "did I see a chair fly? I think I did! Time for popcorn!" -ratdude747

          Comment


            #6
            Re: ASUS P5GC-MX motherboard recap

            Originally posted by Per Hansson View Post
            Just realized something: that board I mentioned, the caps that have failed are momaka's favorite KZG 1500uF 16v
            So maybe you will have to replace yours too eventually
            I've seen 1500 uF 16V KZG fail before too on various motherboards I've looked at on eBay. But it's a rare occurrence. Nevertheless, I do replace them when I have a chance, as I have lots of Nichicon HN (H05 and later) along with MCZ (T05 and later.)
            Mine above are 1000 uF (@ 16V.) Not that this means they can't go bad... it's just that I haven't seen a whole lot of these fail. Will certainly keep an eye out on them from time to time.
            I also already replaced the 1500 uF 6.3V ones too... and they bulged in storage (with mild use as test caps) a few years after.

            Originally posted by ChaosLegionnaire View Post
            odd those bloated when its an athlon 64 system u have there... msi k8n neo4, isnt it? athlon 64 systems dont really run hot. at least not at the level of the presshots...
            True.

            But how hot the caps run can also depends on who made the motherboard.
            Generally speaking, MSI VRMs tend to be pretty well-designed and run cool - at least on most of their motherboards I've seen. ASUS - mostly good, but occasionally not so good too. Gigabyte was good with the old boards, but AM2 and newer have their VRMs run silly-hot sometimes.
            And Biostar and ECS seem to run hot no matter the CPU, if the various ones I have in my collection are to be used as an example.
            Last edited by momaka; 02-03-2021, 02:29 AM.

            Comment


              #7
              Re: ASUS P5GC-MX motherboard recap

              Momaka, thanks for this. I learned a lot from your post. How are you able to determine what the purpose of all the caps is and what rail they are located on? I have a P5GV-MX board and it worked great but when I took it out of storage, it did not power up at all. There are only 3 KZG caps all 16V 1000uf 8x25mm and in the VRM arrea. No bulging, I see a code on the back 59 over C/3). Maybe indicates September 2005? What should I use to replace those KZGs?

              Comment


                #8
                Re: ASUS P5GC-MX motherboard recap

                Originally posted by ChaosLegionnaire View Post
                odd those bloated when its an athlon 64 system u have there... msi k8n neo4, isnt it? athlon 64 systems dont really run hot. at least not at the level of the presshots... how's the psu and the ripple on the 12v output during load? dont tell me it failed in storage de-energized?
                Yup, you identified it correctly.
                Originally if memory serves only one cap was failed.
                The others have failed in storage, well, the first one could have failed de-energized too in hindsight, I don't remember how long it sat here
                It's a leftover from some PC someone never picked up from here...
                Last edited by Per Hansson; 02-03-2021, 01:02 PM.
                "The one who says it cannot be done should never interrupt the one who is doing it."

                Comment


                  #9
                  Re: ASUS P5GC-MX motherboard recap

                  Originally posted by bigbeark View Post
                  Momaka, thanks for this.
                  You're welcome!

                  Originally posted by bigbeark View Post
                  How are you able to determine what the purpose of all the caps is and what rail they are located on?
                  Usually just a multimeter (and a bit of intuition, as desktop motherboards tend to have their voltage rails in more or less "standard" places - i.e. CPU VRM and caps by the CPU, RAM VRM and caps by the RAM slots, and etc.)

                  My usual process is to first identify all caps that are in parallel with each other. After this, I try to identify buck-regulated rails first - that is, voltage rails that are generated with a circuit that has MOSFETs coils/inductors/toroids. The CPU is the easiest one to start with, because the MOSFETs, caps, and inductors/toroids for the CPU VRM are almost always right next to the CPU socket. After that, it's the other rails. Then from all of those caps that I identified are in parallel, I try to see which ones connect to voltage rails on the PSU (3.3V, 5V, and 12V) by doing a resistance test between the positive lead on those caps and each pin on the ATX 20/24 pin connector. Once that is done, I look at the other parallel caps that still haven't been associated with a rail and try to see to which MOSFETs/regulator ICs they connect to, again via resistance test. Once all caps are traced to some MOSFET or PSU pin, I then connect the motherboard to a PSU and power it On, followed by measuring the voltages on the MOSFETs/regulators that connect to these caps. And then finally, from the voltages, I can usually tell what the caps do (if I haven't been already able to determine that from their position on the board already.)

                  So all you need is a multimeter, really. Once you do that on a few boards, you'll eventually start to see the "pattern" of how caps are connected and to which rails... and then identifying them becomes pretty easy.

                  On that note, if you wanted to look at something more simple, here is a GPU (graphics card) I did the same thing so I can see what caps I could use where:
                  https://www.badcaps.net/forum/showthread.php?t=73234

                  Originally posted by bigbeark View Post
                  I have a P5GV-MX board and it worked great but when I took it out of storage, it did not power up at all. There are only 3 KZG caps all 16V 1000uf 8x25mm and in the VRM arrea. No bulging, I see a code on the back 59 over C/3). Maybe indicates September 2005? What should I use to replace those KZGs?
                  I doubt those KZG caps are the problem. Even if they were faulty, the board should have powered up, but possibly not POSTed due to too much noise on CPU V_core. Don't bother replacing them until you can get the board to at least power up. On that note, does the board act totally dead - as in, fans don't even try to spin up for a split moment? If so, you may have other issues.

                  Given also your issue with the BFG 100 Watt PSU you posted in another thread, I suggest to check whatever switch you are using at the front panel connector. Perhaps that switch (or the connector contacts for the FP header) is getting worn out and not making proper contact. Use a flat head screw driver when in doubt.

                  Originally posted by Per Hansson View Post
                  It's a leftover from some PC someone never picked up from here...
                  Their loss, IMO.
                  Those old socket 939 boards are pretty cool. Granted this one has PCI-E 16x and not AGP slot... which would be nice for a retro system. But still, it's a good oldschool mobo. Too bad it's a GeForce chipset - not the most reliable things. But at least it's a GeForce 4 and not a 6100/6150 (or collectively, "GF 6th gen"), which have horrible track record. The GF 4 is OK with good cooling... so that small fan on the chipset really is needed. That's the other thing that bugs me about GeForce chipsets - they all really do need active cooling. And the ones that come with small fans are just way too loud to the point of being very annoying.
                  Last edited by momaka; 02-05-2021, 02:10 AM.

                  Comment


                    #10
                    Re: ASUS P5GC-MX motherboard recap

                    umm they're not geforce chipsets. geforce is the video card line for nvidia. nforce is the motherboard chipset line for nvidia and the shitset on that board is the nforce 4 not geforce 4.

                    Comment


                      #11
                      Re: ASUS P5GC-MX motherboard recap

                      Originally posted by ChaosLegionnaire View Post
                      umm they're not geforce chipsets. geforce is the video card line for nvidia. nforce is the motherboard chipset line for nvidia and the shitset on that board is the nforce 4 not geforce 4.
                      Oops... yeah, you're correct about that. Perhaps my mind's logic spell check wasn't working at the time when I posted this.

                      In any case, it's an nVidia chipset... and not reliable. So let's leave it at that.

                      Comment


                        #12
                        Re: ASUS P5GC-MX motherboard recap

                        Just a short update about some of the caps... I'm glad I did the original preemptive recap on the CPU VRM. Indeed the information I've read in other posts about United Chemicon TMV capacitors failing turned out to be true here as well, after many years. In my case, 1 out of the 6 TMV I removed from this board has also finally failed. Post and pictures about it can be found here. I'm not sure about the other 5. They still look OK visually, but may not be. Currently, they are sitting on an EVGA GeForce 7600 GT "toy" video card that doesn't work properly, so I'm not going to bother to remove them (yet.) But when they fail, I will update this thread. The one failed TMV here happened back in September of last year (see picture date, which is correct.) So very likely the cap failed even a little bit before then. And I imagine the other TMV caps are probably on their way out too.

                        Thus, if you have one of these P5GC-MX motherboards... or really any motherboard with these United Chemicon TMV caps, consider replacing them - especially if the board has sat on the shelf for a long time. Seems that when in use, they might have a chance of lasting longer. But with infrequent use, they tend to fail. Granted I did the preemptive recap back in 2014 (close to 7 years ago now ), so these TMV caps have not seen too much use since then... and 7 years for caps mostly sitting around is not so good for them. Nonetheless, it is also worth mentioning I have a lot of other hardware (with other cap brands) that I use just as infrequently without getting such failures. So UCC TMV probably aren't the most stable caps out there... but far from the worst, either. With that said, if your motherboard has some inexplicable issues, it's probably worth a shot to replace these.
                        Last edited by momaka; 06-11-2021, 02:30 PM.

                        Comment

                        Working...
                        X