Gateway (MPC) E4620

[rollme]

This system was purchased in Aug. of 09, just months before MPC folded. MB is built by ECS, Intel Q35T chipset. Board is populated with a lot of OST caps, two of which are bulging, (See attached photos) both are 6.3v 1800uF

History: The PC one day would not boot. First thing I did was check the PSU which was OK. Bought another board and got the system back up and running.

Gonna check with the boss to see if he wants to try and have it repaired. If not I'm going to ask if I can have it. My question is about which caps to update? I seem to remember reading so time ago that only certain caps need to be replaced, and that it revolved around the size (voltage and capacitance). Any help provided would be much appreciated.

[severach]

All those OST caps were made at the same time so should be equally bad. The earliest failures were subjected to the highest heat. You could:

* Replace just the failed caps and see what fails next.
* Replace all OST blue/gold caps in high heat areas since some might on the verge of failure. This doesn't seem practical since I only see 4 blue/gold caps that aren't at high heat risk.
* Replace all OST blue/gold caps.
* Replace all off brand and at risk good brand caps.

If it was for a server I'd choose the last option. If it is an experimental home machine I'd choose the first option.

[rollme]

sererach,

The boss told me I could have the board, so I'll replace the bulging caps and see if it will come back from the dead. It means I'll have to buy a CPU and Ram.

[yyonline]

OST capacitors do not always bulge when they fail. Even if they look good, they may have failed internally. I would replace all the OST. The little caps (>100uF) are not likely to fail, so those can usually be skipped.

[rollme]

Originally Posted by yyonline

OST capacitors do not always bulge when they fail. Even if they look good, they may have failed internally. I would replace all the OST. The little caps (>100uF) are not likely to fail, so those can usually be skipped.

yyonline, you mean =<100uF right?

Here are the caps that meet that spec...

OST 6.3v 1000uF x14
OST 6.3v 470uF x1
G-Luxon 6.3v 220uF x5

Here are the other lytics on the board:

Evercon 16v 100uF x10
G-Luxon 25v 10uF x6
OST 50v 22uF x2

[goodpsusearch]

Evercon 16v 100uF x10

Oh no! Those caps are so unreliable that they could have failed even though they are 100uF..

[yyonline]

Originally Posted by rollme

yyonline, you mean =<100uF right?

D'oh. Yes, that's what I meant.

Originally Posted by rollme

Here are the caps that meet that spec...

OST 6.3v 1000uF x14
OST 6.3v 470uF x1
G-Luxon 6.3v 220uF x5

Yeah, I'd probably change all those. Some people wouldn't do the 220uF...

Originally Posted by rollme

Here are the other lytics on the board:

Evercon 16v 100uF x10
G-Luxon 25v 10uF x6
OST 50v 22uF x2

Changing them wouldn't hurt, but it probably isn't necessary either. The caps this small tend to run in low-stress areas. Evercon (GSC) is bad enough that I probably would change them out after confirming the board works...

[weirdlookinguy]

Note that this is a BTX board. It won't fit in a regular case.

Hypermicro had a lot of cheap BTX cases last time I checked, if you are planning on building a system around this board.

[bzyzny]

Do you have any updated info on how this project turned out? I have decided to learn the recapping skill, and will be using a few gateway e4620s to practice on. My main question now is that my mobo has no evercon caps, they are all taepo or ost. Should the 100uf 6.3v caps be replaced even though they are not evercon? The computers I have keep power cycling before posting, is this a sign of the bad psu? I did open the psu and saw 1 noticeably swallen cap. Thanks for any help

[bzyzny]

Ok so here is what I've done so far. So far I've gotten 2 of 3 computers working that I've tried so far. All that was needed was for me to replace the 6.3v 1800uF ost caps with 10v 1800uF nichicon caps, and replace 1 cap in the power supply. I can't say much about the ps because I had my friend do that for me, but I can tell you about the mobo. The nichicons I used were a little bigger and the leads were spaced apart further so that was a pain to deal with. After replacing those two caps I reset the cmos battery and then it posted fine. I noticed that the only one that didn't work after doing that was one that had evercon caps where the other two had teapo. Does anyone think that is the cause? Later I will try replacing more of the caps on that mobo and see if it helps.

[asdfvsafderf234]

I have the same computer, and the motherboard had the same capacitor blown like the original poster, RollMe, uploaded. In my case, only one OST capacitor blown.
Please refer to the pictures I have uploaded. The symptoms I kept getting was
the frequent blue screens of death saying there are problems with system files or the system memory. Then one day, the computer completely died. It would power on with fan blowing, but no POST. So, I read RollMe's solution in this thread, tried to replace the blown cap. However, I ONLY replaced the blown cap with a Rubycon 2200uf 6.3v which has larger capacity than the blown OST 1800uf 6.3v, and the result is excellent! It immediately solved the problem!
I was worried that replacing with a capacitor with higher capacitance would
end up badly unlike the voltage which is rather recommended to be replaced with higher one than the blown cap, but I read "Topcats" reply in this thread

http://badcaps.net/forum/showthread.php?p=25370

and took his advice and just replaced it. Now the computer is working fine with no issues.

Huge Thanks to RollMe, TopCat, the Rubycon capacitor!

[sys-eng]

I have 4 Gateway E-4620S systems that start for about 4 seconds then turn off. I can tell that two of them have swollen capacitors - especially the 1800uF one near the memory slots under the optical drive.

It appears that they all have ECS motherboards marked Q35T-GB v1.1
I sent an e-mail message to [email protected] to check on a capacitor kit for this board. I have not replaced capacitors on a motherboard before but it looks like this is a good reason to do it.

[sys-eng]

I never received a reply to my e-mail message so I had to buy elsewhere.

[lcdman]

Hello All,

I have two of the boards that have cap issues. I changed the two 1800uf/6.3v OST caps on one board and still have a restart issue? I had one by the memory slots and one by the PCI-E slot that were bloated. I replaced the two OST caps with Nichicon HZ that I purchased from Topcat. I thought I'd better get some pointers before tinkering with the second mobo! Could one of the members of this site please tell me which caps besides the two that I replaced already should be replaced next? I haven't worked on too many mobos and lack the experience of the other members of this site, so please be kind.

Thanks In Advance,

[lcdman]

Hello Again All,

Can anyone please point out which other caps should be replaced. I don't have a ESR meter. There must be more that are going bad or are bad! Anyone?

Take Care,
lcdman

[sys-eng]

I do not know your computer background so I will ask - - have you taken all the steps to rule out other possible causes such as:

disk drive
memory
power supply
optical drives
card readers
keyboard
mouse

I have seen each of the above cause a computer to not post and restart. My own family PC did this last week when one of the optical drives failed. It would not post anything.

[larrymoencurly]

Originally Posted by lcdman

Can anyone please point out which other caps should be replaced. I don't have a ESR meter. There must be more that are going bad or are bad!

I'm no expert, but I generally replace any questionable brands of caps located around coils, like L6 (near the PCI-e slot -- caps EC13, EC22, EC24) and L24 (next the DIMM slots), around large transistors (anything labelled "Q" and look like like Q16 & Q21 and Q49 & Q51). Sometimes the caps near the north bridge chip (the one with the heatsink in the picture) go bad because of heat from that chip. In the case of the pictured motherboard, the silver-red and silver-blue caps around the CPU don't have to be changed because they're the solid polymer type.

However if the memory is questionable, that is, it has heatsinks on it or is made of chips that can't easily be identified by actual manufacturer (Samsung, Hynix, Micron, Elpida, Nanya, ProMOS, PowerChip), then I'd first run a couple of self-booting memory diagnostics overnight, including MemTest86/86+ and Gold Memory.

[ebayironman]

We have so far replaced the 2 1800uF 6.3V caps at position EC53 and EC13 on many of these boards from Gateway E4620 and other MPC machines.
The original brand caps just fail, almost always.
We use Nichicon 105C 1800uF 6.3V caps and have had nearly 100% success in resolving no post, or spontaneous reboot issues.
When replacing you can use up to 100% greater picofarad values and higher voltages with no problems with the circuit integrity.
One would think higher voltage like 10 or 16V would be more substantial and have higher heat dissipation characteristics.
But replacing with equivalents and good quality high temp caps will suffice.

[rollme]

Originally Posted by rollme

Bought another board and got the system back up and running.

Ok, 2+ years and here is the latest. The replacement board (bosses decision) finally failed just as I expected. So I had my purchasing department get me some samples of the 1800uf 6.3v caps. I'm writing this post on the fixed MB.