• Welcome to Overclockers Forums! Join us to reply in threads, receive reduced ads, and to customize your site experience!

Capacitor replacement on my old S754 board

Overclockers is supported by our readers. When you click a link to make a purchase, we may earn a commission. Learn More.

Tech Tweaker

Member
Joined
Dec 14, 2010
Forgive me if this is in the wrong section, I wasn't entirely sure where to put a thread about capacitor replacement.

Long ago I decommissioned this motherboard due to failing capacitors and voltage-regulation issues. Stopped using it in 2010 after two caps failed around the same time, just so happened I got lucky when I found a deal on a board and picked up a CPU and some RAM for cheap not long after I started noticing these odd behaviors from my board and so I stopped using this one after upgrading to a new platform.

So, I stashed it away in the original box and forgot about it for a few years.

After searching for another board of this same model and being unsuccessful I decided to start looking for someone that could repair it (and others). Found four different people that offered to do the work, but one of them wanted way too much to repair it and the others stopped communicating with me once I had decided I wanted them to do the work and asked where to ship it. I didn't find all of these people at one time though, I'd find one person and then they'd go MIA, then I'd find another person and that person would go MIA as well, and so on.

Finally, I managed to find someone who agreed to show me how to recap boards myself. After watching them do one or two and helping them out with them I got up the courage to attempt this myself on my own board.

Why did I want to recap this old board in the first place? It's a socket 754 board with a PCIe x16 slot (which is rare in and of itself), but it also has an nVidia nForce 4 chipset (which is next to impossible to find). Motherboard is an MSI MS-7135, K8N Neo3 H (nForce 4-4x). Also, I had someone that gave me some socket 754 CPUs a while back, but without a working board I've been unable to test them and want to get this board working so that I can.

After opening the box and taking out the board I was surprised to find that during its long storage another of the capacitors had its top push up and had begun to leak. Luckily I was prepared though, because by chance I had ordered extra capacitors in case I had to replace more than expected.

Only ran into two minor problems, (1) A pin from one of the old capacitors broke off and got stuck in the hole when I was removing it, which then required an additional half-hour to figure out a way to get it out. (2) I bought the wrong physical size of capacitors to replace the two that were side-by-side, so I had to lean one at an angle to attach it since they are too wide (which is why it looks so weird).

Before:
DSCN2737.JPG

DSCN2733.JPG
After:
DSCN2747.JPG

DSCN2745.JPG
 
Last edited:
Thanks for sharing. I had the same board, but parted from it before it began showing visual signs of failing caps. I did notice a bit instability on some voltages, but ran fine at stock. I was mostly interested in overclocking it :)
Have you tested the board ?

I have recapped several boards with good success. Only one board did not wake up, do not know why, but it had a large amount if failing caps replaced.
 
Thanks for sharing. I had the same board, but parted from it before it began showing visual signs of failing caps. I did notice a bit instability on some voltages, but ran fine at stock. I was mostly interested in overclocking it :)
Have you tested the board ?

I have recapped several boards with good success. Only one board did not wake up, do not know why, but it had a large amount if failing caps replaced.

Hmm, first time I've run across someone else that had the same board. Must have been a rare board or limited production run since I've seen so few of them since purchasing it in 2006.

Yes, I tested it later the same day. Working again.

Do have a minor gripe though since the CPU vCore is still unstable and gives me everything from 1.4-1.5V when I manually set it to 1.4V in the bios.
 
Did you install those Chemicon model KZG capacitors? Because they're one of the few types of Japanese caps considered bad, the others being Chemicon model KZJ and Toshin Kogyo brand (marked TK). OTOH the Chemicon KZE caps are considered good.

If a lead is stuck in the board, the safest thing to do is add more 60% tin, 40% lead solder (or 63% tin, 37% lead) and try again, preferrably with a 50W iron. But if it's really stuck, cut off the capacitor from the top side of the circuit board so each of its leads can be removed individually.

Panasonic capacitors (black ones with the "T" stamps on top) are considered the best and shouldn't need replacement.
 
You will have stability issues until you get the caps flush with the board...any lead stickout will cause spurious signals in the board, and instability...one of my slot one boards is like that...I forced the wrong size in there, and they have considerable stickout (naked leads between the cap and board) and it is unstable, unlike all the rest of my refurb'ed boards.

All my replacement caps were purchased from www.badcaps.net and I have never been dissapointed...they also have a tutorial for DYI newbs.

I recommend Rubycon caps...they have never failed me...and they look cool too.
 
Did you install those Chemicon model KZG capacitors? Because they're one of the few types of Japanese caps considered bad, the others being Chemicon model KZJ and Toshin Kogyo brand (marked TK). OTOH the Chemicon KZE caps are considered good.

If a lead is stuck in the board, the safest thing to do is add more 60% tin, 40% lead solder (or 63% tin, 37% lead) and try again, preferably with a 50W iron. But if it's really stuck, cut off the capacitor from the top side of the circuit board so each of its leads can be removed individually.

Panasonic capacitors (black ones with the "T" stamps on top) are considered the best and shouldn't need replacement.

No, those were the original caps (the KZG's).

Yep, that's exactly what I did, removed the leads individually that is. Removing both at the same time proved too difficult on these caps for whatever reason.

Darn lead got stuck in the backside of the board though.

I don't have a 50W iron though, think mine is either a 25W or 30W, it seems to do okay though. It's my terrible soldering skills that are the main problem.

You will have stability issues until you get the caps flush with the board...any lead stickout will cause spurious signals in the board, and instability...one of my slot one boards is like that...I forced the wrong size in there, and they have considerable stickout (naked leads between the cap and board) and it is unstable, unlike all the rest of my refurb'ed boards.

All my replacement caps were purchased from www.badcaps.net and I have never been dissapointed...they also have a tutorial for DYI newbs.

I recommend Rubycon caps...they have never failed me...and they look cool too.
Crap, looks like I may have to try to find the right size then.

I got my replacement caps through www.digikey.com.

I was between Nippon Chemi-con/United Chemi-con and Rubycon, I went with the Nippon Chemi-con caps since that was the brand that most of the ones I was replacing were and I just felt like making it look semi-original.
 
I would amend that to "can cause instability". It depends on the situation.
In some situations it absolutely can, Doc's seen it, I've seen it, it happens.
In others, you can get away with it.

It depends on the specific situation, and since you can't really tell easily without a scope the safe thing to do is minimize lead length.
 
It's interesting to note that (if I got the math right) with a 4GHz processor, those cute little electrons move just about 75 mm. in one clock cycle. Amazing, isn't it, how a mm. or two of lead on a cap can make such a difference. Parasitic this and that. MB designers gotta live on Maalox. I ain't even going to try and imagine what chip heads must go through...
 
Back