VOGONS


First post, by Sly_Botts

User metadata
Rank Member
Rank
Member

Greetings and Salutations!

I recently re-obtained my old motherboard from my in-laws. They've finally decided to stop using that old Athlon XP 2800+ system.

Anyway here is my issue. I noticed that the capacitors around the CPU socket were all bulging and some had leaked. I cleaned some of it up first and tested the motherboard to make sure it was still working. Sure enough, everything was workihg great, however I knew I had to replace those caps. The board had some green/gold coloured caps that were 3300UF 6.3V so I ordered:

Caps:
Panasonic CAP, 3300UF, 6.3V, ALU ELEC, RADIAL;
Capacitance:3300F;
Voltage(DC):6.3V;
Capacitance Tolerance:ñ 20%;
Capacitor Terminals:PC Pin;
Lifetime @ Temperature:10000 hours @ 105øC;
Polarity:Polar;
Capacitor Case / Package:Radial Leaded

I will attach the layout and highlight the caps I replaced.

Anyway, after a successful recap I decided to remove the Northbridge heatsink, clean it and put new paste there. Once this was done I started up my system only to get beep error codes for RAM not being inserted properly. I switched from 2 chips to one and noted that the 3rd lone RAM slot is now the only slot that is working properly. If I add any RAM to slot 1/2 the system will not post and give me the RAM error code.

QUERY

Is it possible I damaged the memory controller on the Northbridge chip by applying too much force to remove the heat sink? Or can the Capacitors be doing this somehow? (I'm thinking not).

The Motherboard is an ASUS A7N8X-X. Socket A (462) .

Attachments

  • mobo1.jpg
    Filename
    mobo1.jpg
    File size
    86.77 KiB
    Views
    619 views
    File license
    Public domain
  • mobo2.jpg
    Filename
    mobo2.jpg
    File size
    63.98 KiB
    Views
    619 views
    File license
    Public domain
  • mobo3.jpg
    Filename
    mobo3.jpg
    File size
    53.6 KiB
    Views
    619 views
    File license
    Public domain
  • mobo1a.jpg
    Filename
    mobo1a.jpg
    File size
    86.19 KiB
    Views
    619 views
    File license
    Public domain
Last edited by Sly_Botts on 2023-01-24, 02:19. Edited 1 time in total.

It is possible to commit no errors and still lose. That is not a weakness, that is life.

Reply 1 of 8, by Ozzuneoj

User metadata
Rank l33t
Rank
l33t

How much force would you say you applied to pull the heatsink off? I would think that you'd have to pry pretty hard to damage traces or solder joints under the chip, or you'd have pry with a metal tool that gouged\scraped something.

Without knowing how experienced you are at soldering, I would say it's more likely that one of your solder joints isn't good enough. Getting a good solid joint on through-hole caps on nice multi-layer boards like that can be surprisingly tough unless you do a lot of soldering or have a really nice setup. Preheating the board helps a lot since they can absorb a ton of heat compared to older boards.

If you have a really good magnifier, check each solder joint, and apply a tiny amount of force to the cap to rock it back and forth (tiny amount!) to see if the joint moves.

Also, check inside all of the slots, and brush off the front and back of the board thoroughly to make sure there aren't any conductive contaminants causing problems.

EDIT: Also, it's worth mentioning that if a heatsink is stuck on with any kind of thermal epoxy, it really shouldn't need to be repasted because the curing\drying process happened before it even left the factory, so there are no concerns of the paste drying out over time or thinning out and pumping\flowing out from the contact area.

In this case, since you were able to actually get it off, it was probably just paste. But even if it's just stuck on really well with paste, I think the benefits of pulling it off and reapplying thermal paste on most hardware are greatly exaggerated. In most cases I wouldn't bother doing it unless it's a part that's known for getting heat damaged due to dried out paste, or if it's a part that's really hard to access normally and you've already got it partially disassembled.

Last edited by Ozzuneoj on 2023-01-24, 08:45. Edited 2 times in total.

Now for some blitting from the back buffer.

Reply 2 of 8, by majestyk

User metadata
Rank Oldbie
Rank
Oldbie

I cannot find the Panasonic "CAP" series/type...
https://industrial.panasonic.com/ww/products/ … cap-lead/models

Sometimes the old thermal paste gets hard so the heatink is glued to the BGA chip. If any solder balls have deteriorated due to age, heat or a non-optimal production process it´s indeed possible to loosen them when pulling the heatsink away from the chip/mainboard.
You can try if the error persists when you press the heatsink and chip down firmly during startup.

Reply 3 of 8, by dionb

User metadata
Rank l33t++
Rank
l33t++

There are more caps than just those 8 you replaced on that board. It's entirely possible other caps have now become the weakest link, such as the medium-sized ones near the DIMM slots.

That said, checking your solder joints and pressing down on the chipset heatsink as already advised are definitely steps to try before soldering any more.

Reply 4 of 8, by Sly_Botts

User metadata
Rank Member
Rank
Member

I have enough experience to competently solder through hole capacitors and I did check and reflow the solder joints. I think maybe I did something to the North Bridge because it was glued. It's ok though, this was a more of a fun side project. Everything else still works fine.

It is possible to commit no errors and still lose. That is not a weakness, that is life.

Reply 5 of 8, by Roman555

User metadata
Rank Oldbie
Rank
Oldbie
Sly_Botts wrote on 2023-01-24, 15:11:

I have enough experience to competently solder through hole capacitors and I did check and reflow the solder joints. I think maybe I did something to the North Bridge because it was glued. It's ok though, this was a more of a fun side project. Everything else still works fine.

It's better to heat a glued heatsink before removing using a usual hair dryer. After heating you won't have to apply too much force to remove the heatsink.
Though there's glue that doesn't get softer after heating. In this case I push a knife with a thin blade between a chip (not a bare die FCPGA !) and a heatsink.

[ MS6168/PII-350/YMF754/98SE ]
[ 775i65G/E5500/9800Pro/Vortex2/ME ]

Reply 6 of 8, by majestyk

User metadata
Rank Oldbie
Rank
Oldbie

Every now and then, when I´m in a hurry, small sprinkels of liquid solder drop down and slip under an ISA-, PCI- DIL- or a memory socket so they cannot be seen.
Each time it´s quite an effort to find them...

Reply 7 of 8, by Hoping

User metadata
Rank Oldbie
Rank
Oldbie

My hands are getting worse as time passes, like if they were bad enough, I damaged an A7N8X deluxe 2.0 because I damaged on trace on one of the two capacitors on the left side of the memory slots, and now, the board only recognizes the memory on the first slot.
So maybe other caps on your board are bad, and it may be interesting to try to recap the rest of the board, after all, at least my board was plagued with OST capacitors.

Reply 8 of 8, by Sly_Botts

User metadata
Rank Member
Rank
Member
Hoping wrote on 2023-01-24, 16:50:

My hands are getting worse as time passes, like if they were bad enough, I damaged an A7N8X deluxe 2.0 because I damaged on trace on one of the two capacitors on the left side of the memory slots, and now, the board only recognizes the memory on the first slot.
So maybe other caps on your board are bad, and it may be interesting to try to recap the rest of the board, after all, at least my board was plagued with OST capacitors.

You know... I guess I could try that, just replace the caps next to the memory slots.

It is possible to commit no errors and still lose. That is not a weakness, that is life.