Should I risk repairing this Gigabyte GA-8IPE1000?

[valuequest]

Hello badcaps.net experts ~
I am not a computer hardware tech. I don't mind taking a project on if I need to though. This board seems to run okay for the time being but I guess the This is why you NEVER Ignore Bad Capacitors thread has got me nervous about what might happen if I don't replace the capacitors that appear to be leaking.
Does this look bad enough to risk my damaging this board while trying to replace these capacitors?

[yyonline]

Re: Should I risk repairing this Gigabyte GA-8IPE1000?

Well, if you do nothing, it will get worse. It may happen tomorrow, or it may happen a year from now.

Those look to be Rubycon MBZ capacitors, which are usually very reliable. Given that they failed, I would check your power supply for issues. If the power supply isn't doing its job properly, it overworks the caps on the board. If you do open up the power supply, do so only after letting it sit unplugged for several minutes. This will give any stored power a chance to drain so that you don't get zapped. Exposure to high heat shortens the lifespan of capacitors as well, so check your case to make sure it's ventilating properly.

[PCBONEZ]

Re: Should I risk repairing this Gigabyte GA-8IPE1000?

They are only using 5 caps in Vcore.
Probably barely enough to handle the Ripple.
.
If you want it to last a while use 5x 1200uF Solid Polymer there.
[1000uF minimum, 1200uF preferred.]
Other caps on the board can be left as Lytics.
.
.

[Th3_uN1Qu3]

Re: Should I risk repairing this Gigabyte GA-8IPE1000?

That is a good board but it runs VERY VERY hot especially if you use a Prescott core processor. The reason the (good quality) Rubycon capacitors got damaged is mainly because the coils near them get very hot. I know this because i was testing one while it was sitting on an antistatic bag - and the bag melted in the CPU area.

[bigbeark]

Re: Should I risk repairing this Gigabyte GA-8IPE1000?

Hello badcaps.net experts ~
Does this look bad enough to risk my damaging this board while trying to replace these capacitors?

I would say replace then NOW. If you have the right equipment and are patient, replacing these caps is not difficult. Leaving them as is could damage other components. I've seen fried toroidal inductors and it's not pretty.

[valuequest]

Re: Should I risk repairing this Gigabyte GA-8IPE1000?

Okay I've got it... I need to replace all 5 of these caps now and check the power supply and case ventilation. I appreciate the your prompt and thorough replies yyonline, PCBONEZ,Th3_uN1Qu3 and bigbeark. Who knows... maybe a success on this project will grow some hair on my chest

[severach]

Re: Should I risk repairing this Gigabyte GA-8IPE1000?

The reason it still works is because those 5 caps are in parallel. The two that are still good are taking the entire load.

The 5 didn't last because of too much external heat. The two that remain have more external heat than ever from a clogged fan and by carrying the entire load are making much more internal heat than before. Chances are the two that look good are on the way out.

It isn't just the toroid coils burning the caps. The caps are blocking an air outlet for the processor and the dirt in the heat sink is restricting air flow. I prefer the smooth and rounded Intel heat sinks made for the old slow Pentium 4 processors. What these Prescott heat sinks gain in radiating area they lose in dirt clog resistance.

[Th3_uN1Qu3]

Re: Should I risk repairing this Gigabyte GA-8IPE1000?

They don't gain in radiating area - they have that funky fan with built-in temperature sensor and boost converter. It goes to 5200 rpm, and sounds like you would expect a 5200 rpm fan to sound - something close to a vacuum cleaner.

[c_hegge]

Re: Should I risk repairing this Gigabyte GA-8IPE1000?

Poly mod the VRM. That way, you won't have to worry about the coils overheating the caps. That's what I do when I see bad rubycon.

[Agent24]

Re: Should I risk repairing this Gigabyte GA-8IPE1000?

i was testing one while it was sitting on an antistatic bag

Not a great idea as far as I know, remember antistatic bags are designed to be conductive...

[PCBONEZ]

Re: Should I risk repairing this Gigabyte GA-8IPE1000?

They don't gain in radiating area - they have that funky fan with built-in temperature sensor and boost converter. It goes to 5200 rpm, and sounds like you would expect a 5200 rpm fan to sound - something close to a vacuum cleaner.

It does gain in radiating area because the thinner fins allows -more- fins which gives more surface area.

That board [as most of that vintage] has a BIOS based fan speed control.
.

[PCBONEZ]

Re: Should I risk repairing this Gigabyte GA-8IPE1000?

The reason it still works is because those 5 caps are in parallel. The two that are still good are taking the entire load.

Yes, exactly.

The 5 didn't last because of too much external heat.

Sort of, and that definitely contributes to the problem.
External heat does reduce cooling which drives up the internal heat.
-
Caps actually die from the -internal- heat.
The act of passing Ripple to ground creates heat -inside- the cap.
The Ripple rating tells you how much Ripple the cap can handle without overheating internally.
-
Most socket 478 boards [especially Prescott capable] use 8 to 10 [or more] Vcore caps to spread the heat load around and reduce ESR.
With only 5 caps paralleled in Vcore the Vcore caps are already loaded [heat wise] heavier than typical.
If even one fails open [or it's ESR goes up as it usually does when caps bloat] the load on the other caps will likely go up enough to cause a cascade of failures. [Domino effect.]

I prefer the smooth and rounded Intel heat sinks made for the old slow Pentium 4 processors. What these Prescott heat sinks gain in radiating area they lose in dirt clog resistance.

Be careful with that one. [Check temps often until you know it's good enough.]
Those heat-sinks have fewer fins and therefore less surface area to start with.
They might be harder to clog but they are a less effective cooler even when completely clean and so it will take less dirt to make cooling inadequate.

I've had no serious problems with stock Intel coolers cooling, but I don't like the stupid lever clips.
-
The kind I like and look for are shown in the pics.
Clips are easier to deal with, replacement [or upgrade] fans are easy to find, and they are easy to clean.
- BUT look closely at them. They are NOT created equal.
-
The harder to find 'good one' shown on left has 33 fins. [Anything over about 30 fins is good.]
That is what I prefer and use if the CPU is over about 70 watts.
[One in photo is a 'stock' Compaq/HP unit.]
-
The typical generic ones you see of that style have around 20 fins. [One shown has 23.]
Those are fine up to about 65-70 watts TDP. [If the case is properly cooled too.]
-
They don't -look- that much different but the 'good one' has roughly 50% more cooling surface.
- That is a BIG difference.
.

[Th3_uN1Qu3]

Re: Should I risk repairing this Gigabyte GA-8IPE1000?

That board [as most of that vintage] has a BIOS based fan speed control.

Yes but that type of cooler does *not* care about the BIOS at all. I had two of those boards, both ran Prescott CPUs with the type of cooler severach mentioned and behaved exactly the same.

[PCBONEZ]

Re: Should I risk repairing this Gigabyte GA-8IPE1000?

Yes but that type of cooler does *not* care about the BIOS at all.

Not true.
The fan can't run faster than the power sent to it by the mobo and the BIOS controls that if you set it up correctly.
If you enabled the fan control in the BIOS then it is in charge.

I had two of those boards, both ran Prescott CPUs with the type of cooler severach mentioned and behaved exactly the same.

So have I.
But I enabled the fan control [and it worked].
.

[Th3_uN1Qu3]

Re: Should I risk repairing this Gigabyte GA-8IPE1000?

Not true.
The fan can't run faster than the power sent to it by the mobo and the BIOS controls that if you set it up correctly.

I thought the same back when i had it but yet it did. It has a built-in temperature sensor and BOOST CONVERTER thus it CAN run faster than the mobo (or you for that matter) would like it to. Try this: Hook one to a spare power supply and hold a lighter close to the fan hub. You'll see what i mean. If this never happened to you it means that it's either very cold where you live, or you never really loaded those CPUs up.

[PCBONEZ]

Re: Should I risk repairing this Gigabyte GA-8IPE1000?

I thought the same back when i had it but yet it did. It has a built-in temperature sensor and BOOST CONVERTER thus it CAN run faster than the mobo (or you for that matter) would like it to. Try this: Hook one to a spare power supply and hold a lighter close to the fan hub. You'll see what i mean. If this never happened to you it means that it's either very cold where you live, or you never really loaded those CPUs up.

Sorry, it can not run faster than the power sent to it.
Mobo BIOS cuts power at fan's source.

Normal day time temp here is 110-115F for a good part of the year.
At 70F people put on winter coats.

[Th3_uN1Qu3]

Re: Should I risk repairing this Gigabyte GA-8IPE1000?

You're mistaking power for voltage here.

[PCBONEZ]

Re: Should I risk repairing this Gigabyte GA-8IPE1000?

No.
I'm not.
.