Help in identifying components (SMD fuses?)

Merlin

Ministry of Retr0bright and Street Judge
VIP
Joined
Nov 24, 2007
Posts
15,597
Country
UK
Region
Manchester
This is one for the SMD component gurus.

I need some help with this one please. I am looking to repair a relatively new motherboard that has suddenly stopped working and can't be RMA'd back to where it came from, as it's second hand. It was working on the bench and stopped working when it was installed into a case - a short-out could have occurred, but so far I've only seen the board for a quick look-over.

The symptoms are a brief 1/4 turn of the fans at power on then nothing. The motherboard is a Gigabyte GA890FXA UD-7 board (socket AM3, DDR3) and the board LEDS remain on '0 0' when powered up.

I've narrowed the search down to some SMD fuses that I think might be along the lines of 1210 or 1206 types, however, they are marked with an italic capital P and L 19 or 1517 and I am struggling to positively identify these components, as they don't correspond with information on the 1206 or 1210 data sheets, particularly the L marking. I've added a picture of the area of the board I'm talking about, so that you see the components I'm on about.

The component references on the board (e.g. F15, IF1) tells me that these are fuses, but which type exactly are they?

Thanks
 

Attachments

  • fuses.jpg
    fuses.jpg
    28.7 KB · Views: 0
Hmm... I thought as much. I probably don't have to struggle replacing SMD fuses then - good.

His CPU and memory stick check out AOK in other machines, so I can rule those out. The issue I have right now is that I don't have a 24-pin plus 8-pin ATX PSU to hand, all of my spare ones are 24 plus 4 ATX. It could be a fault with his PSU.

Just as a 'quick and dirty test', I replaced the 4-pin CPU fan with a 3-pin fan and connected my PSU, knowing that 4 pins aren't connected for the CPU lines (I think). The 3-pin fan fires up but the 4-pin fan doesn't - strange.....more investigation needed...the 4th pin is the control line, so a voltage may have gone AWOL...

Any hints and tips would be appreciated for troubleshooting this motherboard.
 
Put an ordinary 3-pin fan on the CPU and try again.

Oh, the fuse is a 1A part, but as Hikey mentioned they are self-reset (the types that acts more like a disconnector). They may go faulty, but they never shorts to ground.
 
A 3-pin fan works, a 4-pin doesn't, however, I only have a 24 plus 4-pin ATX PSU and I really need a 24 plus 8-pin PSU for the extra 12v lines that feed the extra CPU cores, as the 4th pin (Fan Control) takes it's cue from timing pulses from the CPU cores. I discovered that last night from a lot of reading around....

They are PTC fuses and they 'blow' by heating up, disconnecting the current as they do so. As they cool, they 'reset' and normal service is resumed - clever!

I am now leaning towards either a faulty PSU, or a possible motherboard stand-off in the wrong place on his case, creating a short and causing the PTC fuse to blow. He is sending me his PSU, so more diagnostics and bench tests are on the cards for the weekend.
 
A bent PTC will pop off the board and remains open even when cold.

A solder reflux is needed on this case.
 
OK, here's an update...

The board light up and shows 00 on the LED display, but still no boot, even with a 24+8 pin PSU connected. The CPU and memory are known to be good, as is the graphics card.

It's looking like the main bios is corrupt or knackered (hence the 00 as no data is being run) and I've tried the 'power button in - switch on - then release power button and switch off' method of trying to invoke the backup bios, with no success.

The bios chips are Micronix 25LV8006E flash EEPROMS and although I have a JTAG interface and could probably work out the connections, I have no software for a parallel interface JTAg that supports the 25LV800 series EEPROMS.

It's looking like £200-ish worth of motherboard is going down the gurgler unless I can reprogram the main bios - any ideas, anyone?
 
A much belated update on this motherboard.

After LOTS of discussions on various Gigabyte support and other forums and trying it with various brands of memory, it appears that the issue is power related.

Apparently the symptoms I am seeing are down to insufficient power :blink:, causing the 600W PSU I have used so far to keep resetting and the recommendations of the Gigabyte gurus is to replace the Corsair Stealth Stream II 600W PSU I have been testing the board with with.. get this...<drum roll>

A 1000W PSU minimum! :wooha:

Seriously, they honestly reckon that this board is so power-hungry due to the quad Crossfire support, that it will struggle to boot using anything less than an 850W PSU..:eek: I know that this is an ultimate gaming board, but 1000 Watts? That's nearly a single-bar electric fire's worth before you factor in graphics cards, hard drives etc. I may end up needing 1200W and at that point I don't think I'll need the central heating on...

I must admit the reviews I have read about this board have been in uber test rigs with mahoosive PSUs and none of them that I read commented about the huge power requirements. That's because review testers always have the best kit, don't they...?

It seems to have become a recent thing with a lot of newer Gigabyte boards too, from searches i have done.

My gasted is well and truly flabbered at this and it means I now have to save up over a lot of moolah to be able to even use this motherboard.
 
@Merlin, are you connecting just the motherboard and nothing else connected to the power?

I had a PSU that would cut out after a short spin of the fans. Turned out that the PSU needed something to drive, i.e. graphics cards and a hard drive etc.
 
@Merlin
Here they state they used just a 750 Watt one. 1000 watts just to power up is laughable to say the least anyway, next thing will be requiring direct connection to 3-phase AC current to power them up :p
I'd say a 450 watt one would be more than enough for testing.

Are the regulators known to be ok?

If you suspect the BIOS, you could always desolder and swap the BIOS chips as it has dual BIOS (boring but...).

Also there's an ancient trick of removing the CPU, power-cycling the board with no CPU attached, then powerup again with the CPU in. For some odd reason it worked in some seemingly DOA PII/PIII systems, no reason why it would work in this case, just another thing to try.
 
Last edited:
I dug out the manual and it states 500W minimum, hence my thinking that a 600W quality PSU would have enough overhead to fire it up.

@ Ed.D

Motherboard, memory, CPU and either a GTX260 NVvidia Card, a Radeon 4770 or a Radeon 4250, which as far as I can tell are in decreasing order of power requirements.

The board does spin up a 3-pin fan on both the CPU and Case headers, so the Power Good line must be coming up, even if the fan management side of things isn't doing anything.

@ BLTCON0

The board has absolutely no sign of any damage at all, it looks pristine.

I've tried invoking the backup bios by shorting the primary bios out as recommended but that doesn't help.

I'm going to get at least an 800W PSU and try that; there's no way I can afford a 1000W+ PSU at the moment as they are uber-silly prices.
 
@Merlin
Yes the prices are ridiculous plus it's definite 1000 watt should be surplus to requirements in this case.

Also maybe there's a problem with the 5VSB line? A minimum amperage capacity is required from the PSU for the motherboard to power up - it used to be a mere 100 mA in the first ATX revision but I have no idea if the requirements have gone higher nowadays.
 
Yep - been there, tried that, nothing....:(

That's the forum I've been talking to; my username on there is 633Squadron, as the name Merlin was already taken.

There are lots of 'no boot' threads on that forum - it's scary how much Gigabyte's quality and reliability has slipped in recent years. Phenom X6 and Sandy Bridge CPUs seem to have caused lots of problems for Gigabyte boards.

My youngest son has a 1kW PSU that is is bringing around on Sunday, so we will know for certain one way or the other by then.
 
Back
Top Bottom