Help with 2 x fried Intel Little Falls 2 mobos!

Status
Not open for further replies.

Thread Starter

davcom1

Joined May 24, 2009
32
Hi everyone and sorry to jump straight in with a request for help, but...

Having built a nice car PC, my D945GCLF2 motherboard suddenly 'died'. I bought a replacement and that went the same way in approx. 60 seconds!!

I have the ATX power indicator on the motherboard glowing but no life, no fans etc. Have stripped out all peripherals.

Both boards have the same symptoms - no life, ATX power led lit, same very hot transistor!

Unfortunately I'm not an electronics wizard but want to at least try to get one of the boards going again!

The transistor is marked with 1117A 822176 which I cannot find details of on the Internet. It measures 8mm x 4mm across the main body. I have taken resistance readings across the pins with the component in situ, using my crappy meter on the 'diode' setting and got 171 ohms across pins 1 and 2 and 18 ohms across pins 2 and 3. Also get 186ohms from pins 1 and 3. Reversing the probes has no effect on readings. (I've no idea if this means anything!!)

Here's a photo showing the 'hot' transistor: (the feint curved line above the marked component is a hair, not a crack!)



And here's a shot of the motherboard:



Any helpful advice would be gratefully received - i.e. Transistor details, common replacements, idea of what to replace etc.

Thanks!
 

Thread Starter

davcom1

Joined May 24, 2009
32
Hi studiot,

The transistor is too hot to touch for more than half a second. Everything else on the board seems to be ambient temp. The transistor is getting hot like this without the board running - just with ATX and P4 power connected.

I'm using a M2-ATX PSU which is designed to work with car electrics. Can cope with supply voltage variations and engine cranking etc. Someone mentioned the car's alternator could be at fault but the second board died with the ignition set to ACC.

I've obviously tried both boards with a standard ATX PSU and got the same no-working results.

I've also tried the M2-ATX with a standard desktop board using a 12V supply and checked the voltages which were all spot-on. I understand that I will need to find out what is causing the Intel boards to burn out in my car - must be one of the other components or the PSU is doing something strange when powered by the car's electrical circuit.

Here's a stock shot of the PSU:

 

studiot

Joined Nov 9, 2007
4,998
What happens about earthing the boards?

Many boards create problems if the earthing or insulation is incorrect at the mounting hole locations.

i.e. some are designed to be bolted to (earthed) metal standoffs, others to insulating types.
 

Thread Starter

davcom1

Joined May 24, 2009
32
Interesting point! I have to say I don't know. The mobo looks like it should be earthed through the mounts and the PSU possibly not...

With the first install I had everything mounted to wooden board so that it fit behind the dash. This setup ran faultlessly for 6 months. For the replacement motherboard, I bought a proper mini ITX case and had to drill a couple of holes to mount brass standoffs for the PSU...

I guess I'm going to have to start again :)
 

studiot

Joined Nov 9, 2007
4,998
Does sound like a wiring/earthing problem with the MKII.

Oh and welcome to AAC. If you are capable of building this level of stuff I'm sure you will have things to contribute.
 

Thread Starter

davcom1

Joined May 24, 2009
32
Hi,

I have a digital multi meter but no hot gun - not sure what this is.. hot air gun?

I do have standard and mini soldering irons.

Thanks for the link - it's a SOT-223 then.
 

studiot

Joined Nov 9, 2007
4,998
That one looks accessible to other SM soldering techniques. I thought it didn't look like a transistor. Silicon can get too hot to touch so it may still be OK.

I would read the mobo manual and review your mounting techniques and then try it once more on the bench, before changing components.
 

R!f@@

Joined Apr 2, 2009
9,918
Don't mean to prove any one wrong but as I have seen in mobo's nay component besides CPU and chip set's that is too hot to touch means something is wrong.
as for mounting holes, all of them are connected to the GND plane of the board
Hot gun is used to remove SMD components, mini iron cannot heat the board copper layer quickly enuf to melt the solder since the copper layer inside the mobo will absorb heat much faster rate than the iron's heat up time. If you don't wanna damage the board I suggest to get a hot gun and some experience in the SMD component removal techniques, other wise it is pretty hard to repair.
 

R!f@@

Joined Apr 2, 2009
9,918
I like you take a high resolution pic of the board so I can study it closely, or take parts of it separately at high res and close, it would help
 

studiot

Joined Nov 9, 2007
4,998
as for mounting holes, all of them are connected to the GND plane of the board
On some boards, yes.

Some boards are fixed with insulating washers or mounted on plastic pillars because this is not so.

Every so often I come across someone has either removed (lost) these washers, or tried to mount a board on metal standoff or directly onto a metal case.

This prevents the board functioning correctly allows and excess curents overheating certain components.

However the board is not necessarily terminally damaged and functions normally when corrrectly mounted.

I am not familiar with this particular board but it is cost free and sensible to make this check before more drastic action is taken.

I take heart from my experience of components damaged by overheating; they usually show more signs than I can see from the photo, often the type number is burned off at the very least. Even then silicon can become hot enough to boil water and still function.

As regards for changing an SM components I agree the proper equipment is best but if davcom does not have access to these he might have to improvise. Component leads can be cut as well as unsoldered and a replacement mounted above with flying leads.

R!f since you are familiar with this regulator it would be more useful to state its value in case an substitute has to be found.

If the regulator has overheated through excess currrent then it is likely that other components have also suffered. That is assuming that this is indeed the faulty component and the fault does not actually lie elsewhere.

It is true that there some motherboards which used an SM transistor to power the USB ports and DIN mouse/keyboard ports. This could blow due to overcurrent. However the usual symptom here was non functioning mouse/keyboard, as opposed tothe whole board. However this chip is in the wrong location for that particular fault.
 

Wendy

Joined Mar 24, 2008
23,415
I have been known to (carefully) superglue fiber washers in place. Normally they aren't to insulate the board from the case, but the tiny traces that might overlap the mounting scheme. Shorts on expensive components like this are bad news, so a little extra precaution is always good. The screw heads will still ground the motherboard where it needs to be grounded.
 

R!f@@

Joined Apr 2, 2009
9,918
I agree with you studiot.
When mounting mobo's care should be taken to see whether the metal case standoff's don't touch other than the board's ground surrounding the hole.
I have seen ppl using soft cushions between the board and the case, this is not recommended since it will thermally insulate the board.
Normally a well designed casing standoff are perfectly aligned and do not need any washers, but now a days crappy casing are been made to make a few bucks.

Daccom1.
If you are using a mobo ina car I suggest that you use prober sheilding, cause to my knowledge, a vehicle is one hell of a EMI generator.
Motherboard should be mounted on a metad base plate with proper standoff's. I hope you are using something similar. It will help to protect mobo against ignition noise.

As for the LM117 regulator that is currently heating up means it is overloading, the regulator itself could be good, since it has thermal protection. A quick check with the DMM in diode check range cross the pins will tell u if it is not shorted. U might get 0.7V reading but a dead short should not be there.
My guess is a shorted MOSFET is the culprit, since I have replaced these quite a few times.
do this and post back
 

Thread Starter

davcom1

Joined May 24, 2009
32
R!f@@ and others, thanks for the tips!

Here goes... you will have to be patient please!

The transistor in question has 3 pins at its base. With pins down and starting from the left, I get the following in diode check on my DMM: Pins 1+2=173 2+3=34 1+3=192 (I take it the figures are ohms?). Swapping the probes (red and black) makes no difference to the results.

With ATX power connected (and no signs of life apart from the mobo LED) I get the following:
1+2 = approx 900
2+3 = 1
1+3 = 1

Swapping probes gives:
1+2 = approx -300
1+3 = approx -400
2+3 = approx -400

Voltage readings with ATX power plugged in:
1+2 = 0.5 - 0.6
2+3 = 3.1 - 3.3
1+3 = 3.9

I'll upload or provide a link to an extra large mobo image tomorrow.

Thanks again!! :D
 

R!f@@

Joined Apr 2, 2009
9,918
OK. :rolleyes:
Give me a fine pic and I'll label it, point out the things you should check.

PS..I'm a patient guy. In fact I'm so patient I am up for like 36 hrs and counting :D
 

R!f@@

Joined Apr 2, 2009
9,918
I know dav, :p I'm on gear with a project, have to meet the dead line.. big money :)
Anyways I can go for like 48hrs straight. Don't be surprised. or......u'll be surprised. :D
Give me a few mins or hour or so to study the pics. In the mean time, think of any thing I should know like, did smoke came out from the board, or did you accidentally shorted anything.. or whatever , any thing and everything, cause it will be easy or me to begin with.
I'll get back to u.
 

R!f@@

Joined Apr 2, 2009
9,918
dav tell me is the CPU soldered to the board or in a ZIF socket.
and can you show me a pic of the OLD mobo that died.
 
Last edited:
Status
Not open for further replies.
Top