Gigabyte GA-8SIML Motherboard

[gobbler]

Hi,

Got hold of this motherboard, Gigabyte GA-8SIML Rev 1.1. Using the onboard graphic, it does POST but only 4-5s, then it powers off. Saw 5 bloated caps,
1500uF 16V(GSC)*3, 3300uF 6.3V(CHOYO)*2. After changing these 5 caps, the problem still persist. I guessed not the 5 caps giving problem.

Any guys encounter this before? Thanks.

[Toasty]

Re: Gigabyte GA-8SIML Motherboard

Just because there are "un-bloated" ones, does not mean that they aren't bad also... Many fail with no signs.

Pics of your board?

What brand & series did you replace with? What brand/series remains?

Toast

[ratdude747]

Re: Gigabyte GA-8SIML Motherboard

yes, replace all bad brand caps. ensures problem removal and takes care of problems down the road

[gobbler]

Re: Gigabyte GA-8SIML Motherboard



1500uF 16V replaced with United Chemicon KZG
3300uF 6.3V replaced with OST (cannot find any better ones at local stores)

Left 330uF 25V GSC*9
3300uF 6.3V CHOYO *2
1500uF 6.3V CFJ*6
1000uF 6.3V Sanyo*2

[lucky13]

Re: Gigabyte GA-8SIML Motherboard

You still have bad brand caps. You do know OST is bad, right?

The other caps over 470uF all need to be replaced if you don't have a ESR meter to test them. You definitenly should try to replace the other 2 3300uF you left behind. They are probably for Vcore.

Can you take a close up shot so we can see exactly which ones you have changed?

[Toasty]

Re: Gigabyte GA-8SIML Motherboard

>>3300uF 6.3V replaced with OST (cannot find any better ones at local stores)<<

Where are you located?
UCC has a 3300uF @ 6.3v in the KZG line...? If you can't get them locally or online, Topcat has MBZ Rubycon's that will work nicely there: https://www.badcaps.net/pages.php?vid=22

OST is not recommended due to their history of failures. What series did you use? Are they "brand new" or new old stock? Date code?

You can't replace just 2 or 3 caps in a "bank" of them here. You need to replace all of the same type in the same "bank".

i.e.- 4 Choyo, 2 bloated, you only replaced those 2. Replace ALL 4.

I've not heard of CFJ. Can you get a pic of those? If there are any of them between the CPU and the adjacent peripheral ports, I'd be replacing them if I had no way to test them. Being that they are an unknown cap and their location near the CPU (heat), which will quickly kill them.

The Sanyo's are likely just fine.

GSC's are junk too, but their values (330uF) don't warrant a replacement at this time. Are they the ones out around the PCI/video and RAM slots?

LMK

Toast

[gobbler]

Re: Gigabyte GA-8SIML Motherboard

lucky13,

i know OST is bad caps brand. no choice, this is the only brand i can find in local store, one and only store selling. however, i have order some on ebay(Sanyo & Panasonic), it will take 2 weeks to reach me, I dont think i can wait so long. so i will use OST for this round while wating.

i have desoldered all 6 of the 1500uF 6.3V and use my EVB ESR meter, the readings were 0.04/0.05. There is no guideline reading for 6.3V on my ESR meter, is this reading ok? as for the remaining 2 3300uF 6.3V, i also desoldered them and the readings were 0.05/0.06.

[gobbler]

Re: Gigabyte GA-8SIML Motherboard

toast,

the 3300uF 6.3V i used was OST RLX/O418 series. i got some from ebay, i will use sanyo on next round. i bought it as new. can you teach me how to see date code?

ok, i will replace the other 2 Choyo once my sanyo caps arrive.

apologised, typo error, it should be FJ. however, there is a word 1NDEL there. Please see picture.



yes, you are right. those GSC's are around PCI/video and ram slots.

i have one latest finding, after desoldering all the 6 1500uF 6.3V caps, i tried to power up the motherboard again. surprisingly, the same symptom still there. it can display but after 4-5s, it power off again. so can i conclude the 6 1500uF 6.3V caps are not the culprit?

[Toasty]

Re: Gigabyte GA-8SIML Motherboard

>>...can i conclude the 6 1500uF 6.3V caps are not the culprit?<<
Perhaps. You really shouldn't run the mobo with out at least a few of those caps in place. You haven't checked the PSU so swap out your PSU for another. You don't mention checking that out and if it's having "issues" it could be shutting itself down.

FJ - These are Panasonic. The "M" in the square is Matsushita - aka Panasonic.

FJ's are decent caps (many better types exist). However, they're out of spec. They should be reading (depending on size) between .025 & .018 per the datasheet:

8x20 - .016
10x12.5 -.025
10x16 - .018

OST - RLX - .014 ohms @ 2780 mA ripple - okay spec values, but I don't trust them. The new ones are also out of spec and likely the culprits as the 6.3v caps are handling VCore. If that 0418 is the date code, these are "new" old stock. That number would indicate to me 2004, week 18. The time when the cap plague was peaking. These will pop and sizzle before too long, especially right up against the CPU.

VCore is critical on obtaining the absolute lowest ripple. Capacitance is secondary, but ripple handling is the main target here. With poor ESR, you're allowing ripple into the CPU power source. Not good at all.

Toast

[PCBONEZ]

Re: Gigabyte GA-8SIML Motherboard

1NDEL is the date code.
'1N' part of code = 2001, November.
The meaning of the rest of the code changes now and then.
-
Those caps are old enough they should be reformed.
[Term meaning to rebuild the oxide layer on the aluminum foil.]
Given they will be used on +5v or less, using a +5v power supply should be adequate.
It's simple.
Put them on +5v DC [with correct polarity] for at least 5 minutes + 1 minute for each month of storage.
You have 91 months of storage so 91+5 = 96 minutes minimum.
.
Technically you 'should' raise the voltage in steps initially.
Something like 1v for 5 minutes, 2 volts for 5 minutes, ect,,, until you get to 5v -then- start the 91 minutes.
That is to cover the rare case where the oxide layer is so degraded that the cap shorts when you put the reforming voltage on it.
The idea is to rebuild the layer slowly to avoid that shorting.
- Most of the time that's not really necessary, especially at 5v max.
.
If you have scrap parts around you could use some resistors as current limiters and reduce their values until you have no resistor at 5v.
Or, if you have a 3v-ish DC [or less] wall wart around you could use that for 5-10 minutes before going on to 5v.
.

[gobbler]

Re: Gigabyte GA-8SIML Motherboard

toast,

in fact, every time i desoldered one 1500uF 6.3V cap, i connect up to test. same problem. till the 6th cap, also same problem. i tried using another PSU, also same symptom so i would rule out the PSU is a problem.

the FJ caps are 8x20mm, so should be 0.16. btw, my ESR meter only 2 decimal places, so i should get 0.02 reading if according to spec?

in the meantime, i will have to wait for my sanyo caps and will swap out all the remaining CHOYO and OST. if still fails, i will swap out the panasonic FJ.

pcbonez,

thanks for the tips. mabe i will change the panasonic caps, your recommendation will take up alot of time as there are 6 caps. mabe i will try out on one.

[PCBONEZ]

Re: Gigabyte GA-8SIML Motherboard

You can do all the caps at once.
Once they charge up there isn't much current flowing so if you don't kill your power source in the first few seconds [when the caps initially charge up] you're good till the end.
.

[PCBONEZ]

Re: Gigabyte GA-8SIML Motherboard

You could minimize the initial current surge by adding one cap at a time to the power source until they are all connected at once.

[gobbler]

Re: Gigabyte GA-8SIML Motherboard

where can i get 5V supply source?

[Toasty]

Re: Gigabyte GA-8SIML Motherboard

>>same symptom so i would rule out the PSU is a problem.<<
Agreed.

I think something is crossed here. The FJ caps were on the board originally. The OST caps are the new/old stock. Since they've been unused, they are the candidates for reforming.

So that begs the question - Why would you even mess with OST's from 2004?

Temporary - okay. Permanent - No!

>>the FJ caps are 8x20mm, so should be 0.16.<<
No - .016 or 16 milliohms

>>btw, my ESR meter only 2 decimal places, so i should get 0.02 reading if according to spec?<<
I do not know your meter, but I would think that is correct: .01 to .02
Brand new caps might even read .00

Toast

[PCBONEZ]

Re: Gigabyte GA-8SIML Motherboard

In that case [the FJ] you have pulled good caps.
Put them back and be happy.
.

[gobbler]

Re: Gigabyte GA-8SIML Motherboard

Hi

my sanyo 3300uF 6.3V caps came today and i replaced the CHOYO & OST with sanyo.
the symptom still persist.

i noticed, if i do not connect the 4 pins 12V connector to the motherboard, the cpu fan will spin non-stop without stopping.

any tips? thanks.

[lucky13]

Re: Gigabyte GA-8SIML Motherboard

If there is the 4-pin 12v connector, you must connect that otherwise the CPU won't have enough juice to run.

BTW, have you checked the MOSFETs in the VRM area?

[Toasty]

Re: Gigabyte GA-8SIML Motherboard

Start with the basics and provide answers please...

**Make sure you discharge the board and PSU correctly before inserting or removing ANYTHING including power cables:
- Unplug mains power cable from PSU or wall socket
- push power switch on comp front
- fans and lights may operate for a second or two
- discharged

-CPU+heatsink & RAM are in place when you try this, correct?

-Any CPU pins bent or broken?

-Put a video card in and see if that changes anything.

-Mobo speaker connected?
-Any beeps when POSTing?
-Looks like it has an onboard speaker; any noises?

-CPU and/or RAM tested in another computer?

-Is the BIOS reset jumper in the "run/normal" position? Not "reset".

-Check the 330uF caps for very high ESR or shorted.

As lucky13 said, have you tested the MOSFET's for shorts? Do the power up cycle and touch each one with your fingertip. Are any very warm/hot? Further testing will require an ohmmeter.

Let us know.

Toast

[gobbler]

Re: Gigabyte GA-8SIML Motherboard

toasty,

**Make sure you discharge the board and PSU correctly before inserting or removing ANYTHING including power cables:
- Unplug mains power cable from PSU or wall socket
- push power switch on comp front
- fans and lights may operate for a second or two
- discharged

done

-CPU+heatsink & RAM are in place when you try this, correct?

yes

-Any CPU pins bent or broken?

no. cpu working on another mobo.

-Put a video card in and see if that changes anything.

same symptom.

-Mobo speaker connected?

mobo has an onboard buzzer. if needed i will go buy a speaker.

-Any beeps when POSTing?

there is a weak beep and then the power cut off.
i inserted a pci POST card, last error code was 2524.

-Looks like it has an onboard speaker; any noises?

there is a weak beep and then the power cut off.

-CPU and/or RAM tested in another computer?

yes. both working well.

-Is the BIOS reset jumper in the "run/normal" position? Not "reset".

in open mode/normal operation.

-Check the 330uF caps for very high ESR or shorted.

i found one seems to be shorted with a reading of 00 on my ESR meter. i need to desolder out to confirm. will revert.

as for the MOSFETs, will test as my last resort. im reading the FAQ on Testing VRMs, can someone tell me where is the input VRM and output VRM? thanks.